LASSO-Based Machine Learning Algorithm for Prediction of Lymph Node Metastasis in T1 Colorectal Cancer

Jeonghyun Kang; Yoon Jung Choi; Im-kyung Kim; Hye Sun Lee; Hogeun Kim; Seung Hyuk Baik; Nam Kyu Kim; Kang Young Lee

doi:10.4143/crt.2020.974

Articles

Page Path: HOME > Cancer Res Treat > Volume 53(3); 2021 > Article

Original Article Gastrointestinal Cancer LASSO-Based Machine Learning Algorithm for Prediction of Lymph Node Metastasis in T1 Colorectal Cancer: Jeonghyun Kang¹, Yoon Jung Choi², Im-kyung Kim³, Hye Sun Lee⁴, Hogeun Kim⁵, Seung Hyuk Baik¹, Nam Kyu Kim³, Kang Young Lee³; Cancer Research and Treatment : Official Journal of Korean Cancer Association 2021;53(3):773-783.
DOI: https://doi.org/10.4143/crt.2020.974
Published online: December 29, 2020

¹Department of Surgery, Gangnam Severance Hospital, Yonsei University College of Medicine, Seoul, Korea

²Department of Pathology, Yongin Severance Hospital, Yonsei University College of Medicine, Seoul, Korea

³Department of Surgery, Severance Hospital, Yonsei University College of Medicine, Seoul, Korea

⁴Biostatistics Collaboration Unit, Yonsei University College of Medicine, Seoul, Korea

⁵Department of Pathology, Severance Hospital, Yonsei University College of Medicine, Seoul, Korea

Correspondence: Kang Young Lee, Department of Surgery, Severance Hospital, Yonsei University College of Medicine, 50-1 Yonsei-ro, Seodaemun-gu, Seoul 03722, Korea,
Tel: 82-2-2228-2096, Fax: 82-2-313-8289, E-mail: kylee117@yuhs.ac

^* Jeonghyun Kang and Yoon Jung Choi contributed equally to this work.

• Received: September 25, 2020 • Accepted: December 28, 2020

This is an Open-Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

23,520 Views
427 Download
221 Web of Science
215 Crossref

prev next

Full Article

Download PDF

Abstract

Purpose
The role of tumor-infiltrating lymphocytes (TILs) in predicting lymph node metastasis (LNM) in patients with T1 colorectal cancer (CRC) remains unclear. Furthermore, clinical utility of a machine learning–based approach has not been widely studied.
Materials and Methods
Immunohistochemistry for TILs against CD3, CD8, and forkhead box P3 in both center and invasive margin of the tumor were performed using surgically resected T1 CRC slides. Three hundred and sixteen patients were enrolled and categorized into training (n=221) and validation (n=95) sets via random sampling. Using clinicopathologic variables including TILs, the least absolute shrinkage and selection operator (LASSO) regression model was applied for variable selection and predictive signature building in the training set. The predictive accuracy of our model and the Japanese criteria were compared using area under the receiver operating characteristic (AUROC), net reclassification improvement (NRI)/integrated discrimination improvement (IDI), and decision curve analysis (DCA) in the validation set.
Results
LNM was detected in 29 (13.1%) and 12 (12.6%) patients in training and validation sets, respectively. Nine variables were selected and used to generate the LASSO model. Its performance was similar in training and validation sets (AUROC, 0.795 vs. 0.765; p=0.747). In the validation set, the LASSO model showed better outcomes in predicting LNM than Japanese criteria, as measured by AUROC (0.765 vs. 0.518, p=0.003) and NRI (0.447, p=0.039)/IDI (0.121, p=0.034). DCA showed positive net benefits in using our model.
Conclusion
Our LASSO model incorporating histopathologic parameters and TILs showed superior performance compared to conventional Japanese criteria in predicting LNM in patients with T1 CRC.
Key words: Tumor-infiltrating lymphocytes, Lymph node, T1 colorectal cancer, Machine learning, LASSO

Introduction

Colorectal cancers (CRCs) confined to the submucosa (pT1) are found in approximately 5%–13% of overall CRC patients. [1,2] Although curative resection anticipates good clinical outcomes in T1 CRC, non-surgical treatment is of interest due to reduced additional burden. Accurate prediction of lymph node metastasis (LNM) after endoscopic resection or local excision is essential for deciding if additional curative resection is needed. Japanese Society for Cancer of the Colon and Rectum (JSCCR) guidelines (“Japanese criteria”) recommends an additional surgical resection if one or more of the following is applicable: (1) positive vertical resection margin; (2) depth of submucosa invasion ≥ 1,000 μm; (3) lymphovascular invasion (LVI) positive; (4) poorly differentiated adenocarcinoma, signet ring cell carcinoma, or mucinous carcinoma; or (5) grade 2/3 budding at the site of deepest invasion [3]. This criterion is based mainly on pathologic features of the obtained specimens. In fact, most of the patients who underwent such additional surgeries harbor no metastatic lymph nodes [4–6], resulting in overtreatment. The probability of LNM is known to increase with more risk factors [7], but these additional risks do not seem to be taken into account because the current guidelines only make all-or-nothing decisions based on each included variable. Given the lack of a reliable tool that can predict LNM, additional biomarkers have been investigated to raise the predictability using different approaches [8,9].

Tumor-infiltrating lymphocytes (TILs) mediate local host anti-tumor immunity of human malignancies and are known to be important prognostic factors in various cancers [10,11]. TILs are composed of various regulatory subtypes. CD8⁺ T lymphocytes are critical effectors of anti-tumor immunity and CD4⁺ T lymphocytes help induce, maintain, and memorize CD8⁺ T cells [12]. A subset of T cells, called regulatory T lymphocytes, inhibit anti-tumor immune reaction and are considered a key factor in immune escape in cancer patients. The transcription factor forkhead box P3 (FOXP3) functions as a master regulator of the development and function of regulatory T cells and is an important developmental factor for CD4⁺CD25⁺ T regulatory cells that distinguishes “regulatory cells” from conventional “helper cells” among the CD4⁺ helper T cells [13,14]. Previous studies have demonstrated that immunoscores (defined as summation of specific number imposed according to the densities of CD3⁺ or CD8⁺ positive cells in tumor center [TC] and invasive margin [IM]) are better at estimating patient’s prognosis compared to the TNM stage system in colon and rectal cancers, respectively [15]. Nevertheless, the respective prognostic role of spatial distribution of CD3⁺ and CD8⁺ in stage II and stage III CRC showed different results [16,17]. Moreover, the role of TILs in predicting LNM in T1 CRC has not been investigated sufficiently.

Machine learning–based algorithms have been widely used in clinical decision making [18]. Of them, the least absolute shrinkage and selection operator (LASSO) is one of the most commonly used algorithms and its clinical efficacy has been demonstrated previously [17,19]. Such situations could yield many insights for the adoption of machine learning algorithms in predicting LNM in T1 CRC.

The aim of the present study was to investigate the predictive value of immunohistochemistry (IHC) results of TILs in addition to routinely reported pathologic parameters. In addition, we hypothesized that a machine learning–based algorithm that incorporated TILs and conventional histopathologic parameters could be used to improve the predictive accuracy of detecting LNM in T1 CRC patients undergoing curative resection.

Materials and Methods

1. Patients and samples

Patients were identified from a prospectively collected database of CRC resections in a tertiary referral center. This retrospective, single center based study initially considered a total of 381 patients who had undergone potentially curative resection of pathologic T1 CRC between April 2004 and December 2011. In case of CRCs treated with endoscopic procedures such as endoscopic mucosal resection or endoscopic submucosal dissection, surgeons usually recommended additional curative surgical resection with lymph node dissection to the patients when invasion to submucosa or deeper was suspected, resection margin was either vertical or lateral involvement was detected, LVI was noted, or poorly differentiated or mucinous/signet ring cell carcinoma were reported. In this study, we included surgically resected specimens only, which allowed us to evaluate the correct pathologic LNM status.

The inclusion criteria for this study was the availability of formalin-fixed, paraffin-embedded tissue. Exclusion criteria were as follows: those who underwent neoadjuvant therapy, such as preoperative chemoradiotherapy for rectal cancer or neoadjuvant chemotherapy for colon cancer; those who underwent emergency operations; those who had a history of hereditary nonpolyposis CRC, ulcerative colitis or Crohn’s disease; those who was diagnosed as stage IV; and patients with missing data i.e., total retrieved lymph node numbers.

Finally, 316 patients met the inclusion criteria and constitute the study cohort. This study conforms to the TRIPOD (Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis or Diagnosis) statement for studies developing a prediction model.

2. Pathologic evaluation

For this retrospective study, the examinations of pathologic features, including depth of gross morphology, histologic grade, depth of submucosa invasion, LVI, tumor budding, presence of background adenoma (BGA), and immunochemical examinations were conducted by a single experienced pathologist (Y.J.C.) who did not know the patient’s clinical outcomes or LNM status.

Level 2 of Haggitt’s classification for pedunculated type and muscularis mucosae in other non-pedunculated type, such as sessile and superficial cancers, were used as baseline. The vertical distance from baseline to the deepest invasive lesion was used as the depth of invasion in our study [4]. Tumors were classified histologically as: well, moderately, and poorly differentiated adenocarcinomas or as mucinous or signet ring cell type, based on the most predominant histologic feature [8]. Tumor budding is a cancer cell cluster composed of fewer than five cancer or undifferentiated cells, isolated from large cancer portions without forming a unique structure [20]. After selecting the area where budding foci was most intense, the number of budding foci were counted using a magnification of ×200 in hematoxylin and eosin stained (H&E) slides. Tumor budding was defined as high in case of presence of more than five tumor budding, otherwise it was defined as low [21]. LVI was evaluated using H&E slides.

3. IHC and evaluation of IHC results using image analysis software

Representative sections of primary tumors were processed for additional immunohistochemical staining with antibodies against mismatch repair (MMR) status (MLH1 [1:400, clone ES05, Leica Biosystems, Newcastle upon Tyne, UK] and MSH2 [1:800, clone G219-1129, Cell Marque, Rocklin, CA]) and TILs such as CD3 (1:200, clone F7.2.38, Dako, Glostrup, Denmark), CD8 (1:100, clone C8/144B, Dako), and FOXP3 (1:150, clone 236A/E7, Abcam, Cambridge, UK) using a BOND-MAX automatic stainer (Leica Biosystems, Melbourne, Australia) with Bond Polymer Refine Detection (Leica Biosystems, Newcastle upon Tyne, UK) kit.

To evaluate the MMR status, adjacent normal colon tissue served as an internal control for positive staining and a negative control staining was carried out without the primary antibody. Expression was reported as either MMR proficient (tumor cell nuclear expression with positive immune cell expression) or MMR deficient (tumor nuclear expression absent with normal immune cell expression).

An automated imaging software program (Image Pro Plus ver. 7.0, Media Cybernetics, Rockville, MD) was used to evaluate the percentage of each TILs subtype among the detected cells. The density of TILs was expressed by measuring the area occupied by mononuclear cells over the stromal area captured by NIS-Element F (ver. 4.30, Nikon, Tokyo, Japan) under ×400 magnification. Immunoreactivity was measured at two points: in the center of the tumor (TC) and in the IM (S1 Fig.). Spatial location of TC and IM was selected by the pathologist. The mean percentage of the three areas occupied by TILs in TC and IM per section was reported as the density of TILs.

4. Development of LASSO-derived prediction model in the training set

A total of 316 patients were allocated to the training and validation sets using computer-generated random sampling at a fixed ratio; 70% of the patients were assigned to the training set and the remaining 30% were assigned to the validation set.

Using clinicopathologic variables, including TILs information dichotomized into high and low subgroups according to the spatial distributions, LASSO regression was used to generate the predictive model in the training set. LASSO regression is known to be able to remove unimportant variables via the regression coefficients penalizing the size of the parameters. Applying the LASSO regression method, feature selection and predictive signature building was done. LASSO regression shrinks the coefficient estimates toward zero, with the degree of shrinkage dependent on an additional parameter, λ. To determine the optimal values for λ, a 10-time cross-validation was used, and we chose λ via the minimum criteria. The LASSO model was designed to predict the presence of LNM in patients with T1 CRC and is the linear predictor of the binary model built on the training set with selected variables via LASSO algorithm.

5. Validation of LASSO model using area under the receiver operating characteristic, net reclassification improvement, integrated discrimination improvement, and decision curve analysis in the validation set

The performance of the LASSO model, in comparison to Japanese criteria, was measured via area under the receiver operating characteristic (AUROC) analysis, net reclassification improvement (NRI), integrated discrimination improvement (IDI) calculation, and decision curve analysis (DCA) in the validation set. NRI quantifies the net proportion of patients with and without event of interest who are reclassified as higher and lower risk, respectively.

6. Statistical analysis

The clinicopathologic characteristics were analyzed using a variance test where appropriate. The chi-square test or Fisher exact test was used for comparison of categorical variables. Continuous variables were analyzed via Student t test or Mann-Whitney U test. The cutoff values for all immunohistochemical markers were determined at the maximum of Youden’s index and the maximum of accuracy. A univariable analysis was performed to calculate the odds ratio of the single variables in the logistic regression (LR) model after entering one of the variables under investigation. Univariable analysis denoted the association between LNM and the parameter through a 1:1 matching.

A two-sided p-value less than 0.05 was considered statistically significant. All statistical analyses were performed using SPSS software ver. 23.0 (IBM Corp., Armonk, NY) and R ver. 3.5.1 (R-project, Institute for Statistics and Mathematics, Vienna, Austria).

Results

1. Patient characteristics

Three hundred and sixteen patients were included in our analysis. Our initial cohort was categorized into 221 patients in the training set and the remaining 95 patients in the validation set. LNM was detected in 29 (13.1%) and 12 (12.6%) patients after a curative resection in the training and validation sets, respectively.

Table 1 demonstrates the characteristics of the training and the validation set. No significant differences were observed with respect to sex, age, tumor location, depth of invasion, LVI, tumor budding, BGA, LNM, and the rate of MMR between the two groups. In contrast, proportion of more than five of carcinoembryonic antigen (CEA); proportion of G2, G3, and mucinous type; and proportion of pedunculated morphology were significantly higher in the validation set (all p < 0.05).

2. Distribution of TILs and defining cut-off value of TILs in the training set

Distribution of TILs according to LNM based on the anatomical location are illustrated in S2 Fig. We found no significant difference in the mean numbers between the LNM status. Receiver operating characteristic curve analysis to predict LNM based on each TILs subtype’s densities and AUROC values are shown in S3 Fig. and S4 Table. Area under the curve (AUC) values are 0.55, 0.50, 0.50, 0.59, 0.52, and 0.53 in CD3_invasive margin (CD3_IM), CD3_tumor center (CD3_TC), CD8_invasive margin (CD8_IM), CD8_ tumor center (CD8_TC), FOXP3_invasive margin (FOXP3_IM), and FOXP3_tumor center (FOXP3_TC), respectively. We calculated the cut-off value of each TILs subtype using Youden’s index to be 11.3, 9.2, 28.2, 18.5, 15.1, and 20.3 in CD3_IM, CD3_TC, CD8_IM, CD8_TC, FOXP3_IM, and FOXP3_TC, respectively (S4 Table). Each TILs subtype was categorized as high or low according to these cutoff values.

3. Univariable analysis using clinicopathologic parameters and TILs for predicting LNM in the training set

There was no difference in incidence of LNM due to sex, age, CEA level, tumor location, depth of invasion, and gross morphology (pedunculated versus others). Mean depth of invasion did not differ according to LNM (2,947 μm in lymph node [LN]–positive group vs. 2,550 μm in LN-negative group, p=0.099). Among clinicopathologic parameters, histologic grade (grade 1 vs. grade 2, grade 3 and etc., p=0.040), LVI (positive vs. negative, p=0.001), tumor budding (presence vs. absence, p=0.003), and BGA (absence vs. presence, p=0.024) were associated with LNM in the training set. With respect to TILs, CD8 densities in TC dichotomized as 18.5 were identified as a significant parameter for predicting LNM in the training set (p=0.011) (Table 2).

4. Generating LASSO model to predict LNM in the training set

The binomial deviance curve was plotted versus log (λ), where λ is a tuning hyperparameter. Solid vertical lines represent binomial deviance±standard error (SE). The dotted vertical lines are drawn at optimal values by using the minimum criteria and 1-SE criteria. An optimal λ value was selected for the LASSO model by using 10-fold cross-validation via minimum criteria (Fig. 1A). A value of λ=0.02469015 with log (λ)=–3.701351 was chosen. A coefficient profile plot was produced against the log (λ) sequence (Fig. 1B). Training the optimized model with the training set resulted in nine non-zero coefficients for histology grade (grade 1 vs. others), LVI (absence vs. presence), tumor budding (absence vs. presence), BGA (presence vs. absence), CD3_IM (high vs. low), CD3_TC (high vs. low), CD8_IM (high vs. low), CD8_TC (high vs. low), and FOXP3_TC (high vs. low) with coefficients 0.28665580, 0.84002838, 0.60610111, 0.33894896, −0.07225639, 0.34769731, −0.35012378, −0.66414982, and −0.03525056, respectively. The linear predictor was defined as (–2.51312270)+histology grade (grade 1 vs. others)×(0.28665580)+LVI (absence vs. presence)×(0.84002838)+tumor budding (absence vs. presence)×(0.60610111)+BGA (presence vs. absence)×(0.3389–4896)+CD3_IM (high vs. low)×(–0.07225639)+CD3_TC (high vs. low)×(0.34769731)+CD8_IM (high vs. low)×(−0.35012378)+ CD8_TC (high vs. low)×(–0.66414982)+FOXP3_TC (high vs. low)×(–0.03525056).

5. Performance of the LASSO model in the validation set

AUROC comparison of the LASSO model in the training and validation sets showed similar results. (AUROC, 0.795 vs. 0.765; p=0.747) (Fig. 2). In Japanese criteria, patients with either one of the following results were classified as positive for the surgery group (coded as 1) and other patients were denoted as negative for the surgery group (coded as 0): positive vertical resection margin, depth of submucosa invasion ≥ 1,000 μm, LVI positive, poorly differentiated adenocarcinoma, signet ring cell carcinoma, or mucinous carcinoma or grade 2/3 budding at the site of deepest invasion.

In the validation set, AUROC demonstrated better predictive accuracy of LASSO model compared to Japanese criteria (0.765 in the LASSO model vs. 0.518 in the Japanese criteria, p=0.003) (Fig. 2). In addition, the LASSO model showed significantly improved prediction for LNM compared to Japanese criteria, as measured by NRI (0.447, p=0.039) and IDI (0.121, p=0.034) (Table 3). DCA showed positive net benefit in using LASSO model in the training set and validation set, respectively (Fig. 3). Sensitivity, specificity, positive predictive value, and negative predictive value using the LASSO model at each cut-off point in overall set are illustrated in S5 Table.

6. Clinical significance of immunoscore-like score in predicting LNM in T1 CRC

Immunoscore-like score was calculated based on the recommendation [15]. In our cohort, there was no difference in immunoscore-like score between the LN positive and the negative group (p=0.630) (S6 Fig.).

Discussion

Previous studies investigating predictors of LNM in T1 CRC mainly focused on histopathologic parameters [4,22]. Although we expect the combination of pathologic risk factors could help in discriminating the risk of LNM in most patients with T1 CRC, more than 80% of the patients who were candidates for additional surgery were node-negative in final pathologic examinations. This study demonstrated that a machine learning–based algorithm that incorporated clinicopathologic parameters and TILs showed high predictive accuracy of LNM in T1 CRC and had superior discriminatory performance when compared with the conventional Japanese criteria.

Although prognostic value of the inflammatory cell reaction on survival in CRC has been thoroughly studied, the role of lymphocytic infiltration on LNM in early CRC has not been studied as extensively. A previous study considering peritumoral lymphocytic infiltration (PLI) as one of the candidates to predict LNM did not reveal any difference of PLI rate between the node-positive and node-negative groups. [20] Although statistically insignificant, the rate of PLI was conspicuous—at 8.3% in the node-positive group and 22.6% in the node-negative group (p=0.1). Considering the relatively low number of enrolled patients (n=111), the lack of association might be due to a type-II error [20]. Compared to the H&E slide–based evaluation of PLI in the previous study, we evaluated the role of IHC-stained TILs at different anatomical subsites, such as IM and TC. Univariable analysis demonstrated that lower densities of CD8⁺ T cells in TC were significantly correlated with higher rate of LNM in T1 CRC. In the equation that formed the linear predictor of LASSO model, the absolute value of CD8_TC (0.66414982) was larger than that of other IHC values. This result indicates that CD8_TC acted as one of the most obvious factors in predicting LNM, which was predictable from the results of univariable analysis.

A recent meta-analysis reported that the clinical significance of various TILs in CRC showed definite differences according to anatomical subsites [23]. High densities of CD3⁺ T cells indicated poor disease-free survival (DFS) in the invasive margin, whereas high densities of CD8⁺ T cells were a favorable prognosticator with respect to DFS in tumor center. Furthermore, several studies demonstrated that the significance of specific TILs according to spatial distribution might be different depending on the stage. One study revealed that both CD3⁺ T cells in both TC and IM are favorable prognosticators in stage III colon cancer, whereas another study observed the same trends only in stage II colon cancer but not in stage III colon cancer [16,17]. As far as we know, the clinical significance of spatial distribution of TILs as a predictor of LNM in T1 CRC has not been investigated sufficiently. When applying LR-based univariable or multivariable analysis, only CD8_TC would be left as an independent factor. In contrast, using LASSO model, we can estimate the relative role of each spatial TILs subtype using the coefficients imposed on respective variables. Along with CD8_TC, CD3_TC, and CD8_IM are thought to have a certain role in predicting LNM, although the direction of CD3_TC and CD8_IM would be different. Considering the different roles of each TILs subtype depending on the anatomical subsites or stages in non-metastatic CRC patients, it appears that our LASSO model is more reflective of the complex interactions of each TILs subtype. Further research is required to confirm our hypothesis.

Several recent studies have introduced machine learning algorithms for predicting LNM in T1 CRC [18,24,25]. It was reported that an artificial intelligence model using support vector machine (SVM) showed better discrimination than the LR-based model (AUC; 0.821 in SVM vs. 0.789 in LR, p=not stated) among 100 validation sets out of 690 T1 CRC patients [18]. The other study developed a random forest classifier–based prediction model using cytokeratin immunostaining images of resected whole slide images and reported similar predictive ability (AUC, p=0.1) and fewer false-negative cases compared to the LR-based clinical model [24]. However, one should bear in mind that the LR-based clinical models that were used to compare with the aforementioned machine learning algorithms (SVM and random forest classifier) were not the same as the Japanese criteria–based decision model. The risk assessment LR-based model generated in the prior two studies constructed prediction probabilities by combining each conventional histopathologic risk factor, whereas the Japanese criteria classified patients solely into “surgery” or “no surgery” subgroups without considering the imposed risk of each clinicopathologic variable. It is in this context that we developed a LASSO model and compared it with the Japanese criteria used in current clinical practice. The LASSO-based prediction model is a classifier that can accumulate relative risks of meaningful variables and maximize its predictive power. This fundamental characteristic might be one possible reason for the better diagnostic accuracy of the LASSO model. In a similar way, a recent Dutch T1 CRC working group demonstrated that a LASSO-derived prediction model (AUC, 0.83; 95% confidence interval [CI], 0.76 to 0.90) showed better AUC value than the conventional prediction models suggested by American Society for Gastrointestinal Endoscopy/European Society of Gastrointestinal Endoscopy (AUC, 0.67; 95% CI, 0.60 to 0.74; p=0.002) and JSCCR guidelines (AUC, 0.64; 95% CI, 0.58 to 0.70; p < 0.001), respectively, in 708 T1 pedunculated colorectal carcinoma [25].

Previously, the depth of invasion of T1 CRC was classified as SM1 (submucosa 1), SM2, and SM3, respectively [26]. The rate of LNM according to this definition was reported as 0%–3% in SM1, 8%–10% in SM2, and 23%–25% in SM3 [27,28], and SM3 was regarded as a strong indicator for an additional surgery. To clarify the depth of invasion using SM categories, the whole submucosa layer should be evaluated, which is not always possible after endoscopic treatment. In a large retrospective study, it was recommended that a 1,000 μm depth of invasion could be used as a cut-off value for predicting LNM [4]. In our study, however, there was no difference of LNM rate according to depth of invasion dichotomized as 1,000 μm and, therefore, depth of invasion was not selected as a parameter in the LASSO model. Although it is very difficult to explain why depth of invasion did not predict LNM in our cohort, there are several other studies that also reported no association between depth of invasion with LNM [8,22,29]. Selection bias might be one possible reason for this dissociation [29]. In addition, long-span retrospective studies undergoing histopathologic examinations by various pathologists inevitably showed some heterogeneities of dealing with pathologic slides. In our study, depth of invasion as well as other histopathologic parameters were re-evaluated for this study by one pathologist, which is a merit of this study. Large scale prospective studies are required to reveal the discrepancy of the role of depth of invasion on LNM in T1 CRC.

The main limitation of our study is that it was based on a single center, retrospective data collection. Our hospital is one of the largest referral centers in our nation, which skews cases toward more advanced stage patients. The accuracy of Japanese criteria measured by AUC in this study was around 0.52, which is approximately the same probability as tossing a coin. Such a situation would be very embarrassing given its firm status as a predictive model. Although this might reflect the real situation, some unmeasurable biases, such as inclusion of only patients who underwent surgeries, might have acted in reducing the overall performance. Nevertheless, it is impossible to identify the actual incidence of LNM among those patients who did not undergo surgeries, and such a situation would be the main dilemma in dealing with this issue. Also, IHC on paraffin sections might be semiquantitative although we applied image analyzer–based measurements. Artificial intelligence–guided whole slide image analysis has been actively adopted in medicine [17]. It would be an interesting issue that computer-aided detection could minimize the subjectivity of measuring TILs using pathologic manual estimation or image analyzer–based measurements [30]. Despite the internal validation, a larger prospective external validation is required before the LASSO model can be used as a clinical decision maker. It is expected that endoscopic resection will be performed more frequently as the incidence of early CRC increases. Therefore, it is considered to be very important clinically whether the predictive power can be increased through such TILs measurement and machine learning adoption in the specimen after endoscopic procedure. Our study has a limitation in that it cannot provide a direct answer to such a point because it is not a study using a tissue from endoscopic procedure. However, if the center and IM of the tumor can be identified pathologically even in specimens using endoscopic resection, the model developed by us can be used when the density of each location can be measured after IHC. Additional research is needed to confirm the possibility in practical application.

In conclusion, we developed a LASSO-based prediction model for predicting LNM in patients with T1 CRC, and it outperformed the current Japanese guidelines. This finding suggests that machine learning can potentially improve the accuracy of prediction, thereby minimizing unnecessary surgical resections. As the rate of CRCs detected early is growing, incidence of endoscopic treatments for early stage CRC may also be increasing. Evaluation of TILs after endoscopic treatments and application of a new prediction model might be beneficial in deciding for a subsequent radical surgery in T1 CRC.

Electronic Supplementary Material

Supplementary materials are available at Cancer Research and Treatment website (https://www.e-crt.org).

crt-2020-974-suppl1.pdf

crt-2020-974-suppl2.pdf

crt-2020-974-suppl3.pdf

crt-2020-974-suppl4.pdf

crt-2020-974-suppl5.pdf

crt-2020-974-suppl6.pdf

NOTES

Ethical Statement

The protocol of the present study was approved by the institutional review board of the Severance Hospital, Yonsei University College of Medicine (Seoul, Republic of Korea) (approval no. 4-2012-0900). Informed consent was waived by the institutional review board due to the retrospective study design.

Author Contributions

Conceived and designed the analysis: Kang J, Choi YJ, Kim IK, Lee KY.

Collected the data: Kang J, Choi YJ, Kim H, Baik SH, Kim NK, Lee KY.

Contributed data or analysis tools: Kang J, Choi YJ, Kim IK, Lee HS, Lee KY.

Performed the analysis: Kang J, Lee KY.

Wrote the paper: Kang J, Choi YJ, Lee KY.

Conflicts of Interest

Conflict of interest relevant to this article was not reported.

Acknowledgements

We would like to thank Editage (www.editage.co.kr) for English language editing. This study was supported by a CMB-Yuhan research grant of Yonsei University College of Medicine for (6-2014-0069).

Fig. 1

Selection of significant parameters in clinicopathologic variables in the training set and definition of linear predictor. (A) Ten time cross-validation for tuning parameter selection in the LASSO model. (B) LASSO coefficient profiles. The LASSO was used for regression of high dimensional predictors. The method uses an L1 penalty to shrink some regression coefficients to exactly zero. The binomial deviance curve was plotted versus log (λ), where λ is the tuning parameter (A). LASSO coefficient profiles of clinicopathologic variables (B). LASSO, least absolute shrinkage and selection operator.

Fig. 2

Comparison of AUROC between LASSO model in the training and validation sets and Japanese criteria in the validation set. AUC, area under the curve; AUROC, area under the receiver operating characteristic; CI, confidence interval; LASSO, least absolute shrinkage and selection operator.

Fig. 3

Decision curve analysis of Japanese criteria and LASSO model in the training (A) and validation (B) set. The y-axis measures the net benefit. The green line represents the LASSO model. The red line represents the Japanese criteria. The gray line represents the assumption that all patients underwent surgeries. The black line represents the assumption that patients underwent no surgeries. The net benefit was calculated by subtracting the proportion of all patients who are false positive from the proportion who are true positive, weighting by the relative harm of forgoing treatment compared with the negative consequences of an unnecessary treatment. The decision curve showed that if the threshold probability of a patient or doctor is >10%, using the LASSO model in the current study to predict LNM adds more benefit than the treat-all-patients scheme or the treat-none scheme. For example, if the personal threshold probability of a patient is 20% (i.e., the patient would opt for surgery if his/her probability of LNM was > 20%), then the net benefit is 0.35 when using the LASSO model to make the decision of whether to undergo surgery, with added benefit than the treat-all scheme or the treat-none scheme. This decision curve analysis showed that the net benefit was comparable on the basis of the Japanese criteria and the treat-all or treat-none strategies. LASSO, least absolute shrinkage and selection operator; LNM, lymph node metastasis.

Table 1

Comparison of clinicopathological characteristics between the training and validation set

	Training set (n=221)	Validation set (n=95)	p-value
Sex
Male	123 (55.7)	59 (62.1)	0.347
Female	98 (44.3)	36 (37.9)
Age (yr)
< 70	177 (80.1)	77 (81.1)	0.966
≥ 70	44 (19.9)	18 (18.9)
Preoperative CEA (ng/mL)
< 5.0	208 (94.1)	73 (76.8)	< 0.001
≥ 5.0	13 (5.9)	22 (23.2)
Tumor location
Colon	129 (58.4)	60 (63.2)	0.502
Rectum	92 (41.6)	35 (36.8)
Retrieved lymph node numbers	12 (7–17)	14 (12–21)	< 0.001
Depth of invasion (μm)
< 1,000	20 (9.0)	6 (6.3)	0.557
≥ 1,000	201 (91.0)	89 (93.7)
Histologic grade^a)
G1	93 (42.1)	28 (29.5)	0.047
G2, G3, etc.	128 (57.9)	67 (70.5)
LVI
Present	27 (12.2)	14 (14.7)	0.668
Absent	194 (87.8)	81 (85.3)
Gross morphology
Pedunculated	17 (7.7)	16 (16.8)	0.025
Flat and sessile	204 (92.3)	79 (83.2)
Tumor budding^b)
Low grade	88 (39.8)	38 (40.0)	> 0.99
High grade	133 (60.2)	57 (60.0)
Background adenoma
Present	196 (88.7)	84 (88.4)	> 0.99
Absent	25 (11.3)	11 (11.6)
LNM
Negative	192 (86.9)	83 (87.4)	> 0.99
Positive	29 (13.1)	12 (12.6)
MMR
MMR-deficient	14 (6.3)	8 (8.4)	0.669
MMR-proficient	207 (93.7)	87 (91.6)

Values are presented as number (%) or median (IQR). CEA, carcinoembryonic antigen; IQR, interquartile range; LNM, lymph node metastasis; LVI, lymphovascular invasion; MMR: mismatch repair.

^a) Histologic grade: G1, well differentiated; G2, moderately differentiated; G3, poorly differentiated,

^b) Tumor budding grade was as follows: grade 1, 0–4/grade 2, 5–9/grade 3, 10 or more. Grade 1 was defined as low grade and grade 2/3 as high grade.

Table 2

Univariable analysis for predicting lymph node metastasis in the training set (n=221)

	LN positive, n (%)	Univariable analysis
	LN positive, n (%)	OR (95% CI)	p-value
Sex
Female	14/98 (14.3)	Reference
Male	15/123 (12.2)	0.833 (0.37–1.83)	0.647
Age (yr)
< 70	24/177 (13.6)	Reference
≥ 70	5/33 (11.4)	0.817 (0.26–2.12)	0.699
Preoperative CEA (ng/mL)
< 5.0	27/208 (13.0)	Reference
≥ 5.0	2/13 (15.4)	1.218 (0.18–4.86)	0.803
Tumor location
Colon	14/129 (10.9)	Reference
Rectum	15/92 (16.3)	1.6 (0.72–3.53)	0.239
Depth of invasion (μm)
< 1,000	3/20 (15.0)	Reference
≥ 1,000	26/201 (12.9)	0.841 (0.25–3.78)	0.794
Histologic grade^a)
G1	7/93 (7.5)	Reference
G2, G3, etc.	22/128 (17.2)	2.549 (1.08–6.70)	0.040
LVI
Absent	20/194 (10.3)	Reference
Present	9/27 (33.3)	4.35 (1.67–10.84)	0.001
Gross morphology
Pedunculated	2/17 (11.8)	Reference
Flat and sessile	27/204 (13.2)	1.144 (0.29–7.51)	0.863
Tumor budding^b)
Low grade	10/133 (7.5)	Reference
High grade	19/88 (21.6)	3.387 (1.52–7.97)	0.003
Background adenoma
Present	22/196 (11.2)	Reference
Absent	7/25 (20.0)	3.075 (1.09–7.96)	0.024
MMR
MMR-deficient	1/14 (7.1)	Reference
MMR-proficient	28/207 (13.5)	2.033 (0.38–37.65)	0.502
CD3_IM
Low (< 11.3)	10/55 (18.2)	Reference
High (≥ 11.3)	19/166 (11.4)	0.581 (0.25–1.38)	0.203
CD3_TC
Low (< 9.2)	2/39 (5.1)	Reference
High (≥ 9.2)	27/182 (14.8)	3.222 (0.90–20.53)	0.121
CD8_IM
Low (< 28.2)	29/199 (14.6)	Reference
High (≥ 28.2)	0/27 (0)	0 (NA-1.68e+26)	0.990
CD8_TC
Low (< 18.5)	24/133 (18.0)	Reference
High (≥ 18.5)	5/88 (5.7)	0.273 (0.08–0.69)	0.011
FOXP3_IM
Low (< 15.1)	16/99 (16.2)	Reference
High (≥ 15.1)	13/122 (10.7)	0.618 (0.27–1.35)	0.230
FOXP3_TC
Low (< 20.3)	23/157 (14.6)	Reference
High (≥ 20.3)	6/64 (9.4)	0.602 (0.21–1.47)	0.290

CEA, carcinoembryonic antigen; CI, confidence interval; IM, invasive margin; LN, lymph node; LVI, lymphovascular invasion; MMR, mismatch repair; NA, not available; OR, odds ratio; TC, tumor center.

^a) Histologic grade: G1, well differentiated; G2, moderately differentiated; G3, poorly differentiated,

^b) Tumor budding grade was as follows: grade 1, 0–4/grade 2, 5–9/grade 3, 10 or more. Grade 1 was defined as low grade and grade 2/3 as high grade.

Table 3

NRI and IDI in the training set and the validation set

	Training set (n=221)		Validation set (n=95)

	Japanese criteria vs. LASSO model	p-value	Japanese criteria vs. LASSO model	p-value
NRI (95% CI)^a)	0.722 (0.402–1.041)	< 0.001	0.447 (0.041–0.854)	0.039

IDI (95% CI)	0.187 (0.100–0.274)	< 0.001	0.121 (0.008–0.234)	0.034

CI, confidence interval; LASSO, least absolute shrinkage and selection operator; NRI, net reclassification improvement.

^a) Cutoff of “0, 0.1, 0.2, 1” was used in this analysis.

REFERENCES

1. Dai W, Mo S, Xiang W, Han L, Li Q, Wang R, et al. The critical role of tumor size in predicting prognosis for T1 colon cancer. Oncologist. 2020;25:244–51. Article PubMed
2. Gunderson LL, Jessup JM, Sargent DJ, Greene FL, Stewart AK. Revised TN categorization for colon cancer based on national survival outcomes data. J Clin Oncol. 2010;28:264–71. Article PubMed
3. Hashiguchi Y, Muro K, Saito Y, Ito Y, Ajioka Y, Hamaguchi T, et al. Japanese Society for Cancer of the Colon and Rectum (JSCCR) guidelines 2019 for the treatment of colorectal cancer. Int J Clin Oncol. 2020;25:1–42. Article PubMed
4. Kitajima K, Fujimori T, Fujii S, Takeda J, Ohkura Y, Kawamata H, et al. Correlations between lymph node metastasis and depth of submucosal invasion in submucosal invasive colorectal carcinoma: a Japanese collaborative study. J Gastroenterol. 2004;39:534–43. Article PubMed
5. Kobayashi H, Mochizuki H, Morita T, Kotake K, Teramoto T, Kameoka S, et al. Characteristics of recurrence after curative resection for T1 colorectal cancer: Japanese multicenter study. J Gastroenterol. 2011;46:203–11. Article PubMed
6. Oh JR, Park B, Lee S, Han KS, Youk EG, Lee DH, et al. Nomogram development and external validation for predicting the risk of lymph node metastasis in T1 colorectal cancer. Cancer Res Treat. 2019;51:1275–84. Article PubMed PMC
7. Ueno H, Mochizuki H, Hashiguchi Y, Shimazaki H, Aida S, Hase K, et al. Risk factors for an adverse outcome in early invasive colorectal carcinoma. Gastroenterology. 2004;127:385–94. Article PubMed
8. Nishida T, Egashira Y, Akutagawa H, Fujii M, Uchiyama K, Shibayama Y, et al. Predictors of lymph node metastasis in T1 colorectal carcinoma: an immunophenotypic analysis of 265 patients. Dis Colon Rectum. 2014;57:905–15. PubMed
9. Jung CK, Jung SH, Yim SH, Jung JH, Choi HJ, Kang WK, et al. Predictive microRNAs for lymph node metastasis in endoscopically resectable submucosal colorectal cancer. Oncotarget. 2016;7:32902–15. Article PubMed PMC
10. Geng Y, Shao Y, He W, Hu W, Xu Y, Chen J, et al. Prognostic role of tumor-infiltrating lymphocytes in lung cancer: a meta-analysis. Cell Physiol Biochem. 2015;37:1560–71. Article PubMed
11. Cha YJ, Park EJ, Baik SH, Lee KY, Kang J. Clinical significance of tumor-infiltrating lymphocytes and neutrophil-to-lymphocyte ratio in patients with stage III colon cancer who underwent surgery followed by FOLFOX chemotherapy. Sci Rep. 2019;9:11617.Article PubMed PMC
12. Yu P, Fu YX. Tumor-infiltrating T lymphocytes: friends or foes? Lab Invest. 2006;86:231–45. Article PubMed
13. Hori S, Nomura T, Sakaguchi S. Control of regulatory T cell development by the transcription factor Foxp3. Science. 2003;299:1057–61. Article PubMed
14. deLeeuw RJ, Kost SE, Kakal JA, Nelson BH. The prognostic value of FoxP3+ tumor-infiltrating lymphocytes in cancer: a critical review of the literature. Clin Cancer Res. 2012;18:3022–9. Article PubMed
15. Pages F, Mlecnik B, Marliot F, Bindea G, Ou FS, Bifulco C, et al. International validation of the consensus Immunoscore for the classification of colon cancer: a prognostic and accuracy study. Lancet. 2018;391:2128–39. PubMed
16. Cavalleri T, Bianchi P, Basso G, Celesti G, Grizzi F, Bossi P, et al. Combined low densities of FoxP3(+) and CD3(+) tumor-infiltrating lymphocytes identify stage II colorectal cancer at high risk of progression. Cancer Immunol Res. 2019;7:751–8. Article PubMed
17. Reichling C, Taieb J, Derangere V, Klopfenstein Q, Le Malicot K, Gornet JM, et al. Artificial intelligence-guided tissue analysis combined with immune infiltrate assessment predicts stage III colon cancer outcomes in PETACC08 study. Gut. 2020;69:681–90. Article PubMed
18. Ichimasa K, Kudo SE, Mori Y, Misawa M, Matsudaira S, Kouyama Y, et al. Artificial intelligence may help in predicting the need for additional surgery after endoscopic resection of T1 colorectal cancer. Endoscopy. 2018;50:230–40. Article PubMed
19. Huang YQ, Liang CH, He L, Tian J, Liang CS, Chen X, et al. Development and validation of a radiomics nomogram for preoperative prediction of lymph node metastasis in colorectal cancer. J Clin Oncol. 2016;34:2157–64. Article PubMed
20. Akishima-Fukasawa Y, Ishikawa Y, Akasaka Y, Uzuki M, Inomata N, Yokoo T, et al. Histopathological predictors of regional lymph node metastasis at the invasive front in early colorectal cancer. Histopathology. 2011;59:470–81. Article PubMed
21. Watanabe T, Itabashi M, Shimada Y, Tanaka S, Ito Y, Ajioka Y, et al. Japanese Society for Cancer of the Colon and Rectum (JSCCR) guidelines 2010 for the treatment of colorectal cancer. Int J Clin Oncol. 2012;17:1–29. Article PubMed
22. Wada H, Shiozawa M, Katayama K, Okamoto N, Miyagi Y, Rino Y, et al. Systematic review and meta-analysis of histopathological predictive factors for lymph node metastasis in T1 colorectal cancer. J Gastroenterol. 2015;50:727–34. Article PubMed
23. Zhao Y, Ge X, He J, Cheng Y, Wang Z, Wang J, et al. The prognostic value of tumor-infiltrating lymphocytes in colorectal cancer differs by anatomical subsite: a systematic review and meta-analysis. World J Surg Oncol. 2019;17:85.Article PubMed PMC
24. Takamatsu M, Yamamoto N, Kawachi H, Chino A, Saito S, Ueno M, et al. Prediction of early colorectal cancer metastasis by machine learning using digital slide images. Comput Methods Programs Biomed. 2019;178:155–61. Article PubMed
25. Backes Y, Elias SG, Groen JN, Schwartz MP, Wolfhagen FH, Geesing JM, et al. Histologic factors associated with need for surgery in patients with pedunculated T1 colorectal carcinomas. Gastroenterology. 2018;154:1647–59. Article PubMed
26. Kudo S. Endoscopic mucosal resection of flat and depressed types of early colorectal cancer. Endoscopy. 1993;25:455–61. Article PubMed
27. Kikuchi R, Takano M, Takagi K, Fujimoto N, Nozaki R, Fujiyoshi T, et al. Management of early invasive colorectal cancer: risk of recurrence and clinical guidelines. Dis Colon Rectum. 1995;38:1286–95. PubMed
28. Nascimbeni R, Burgart LJ, Nivatvongs S, Larson DR. Risk of lymph node metastasis in T1 carcinoma of the colon and rectum. Dis Colon Rectum. 2002;45:200–6. Article PubMed
29. Choi DH, Sohn DK, Chang HJ, Lim SB, Choi HS, Jeong SY. Indications for subsequent surgery after endoscopic resection of submucosally invasive colorectal carcinomas: a prospective cohort study. Dis Colon Rectum. 2009;52:438–45. PubMed
30. Klauschen F, Muller KR, Binder A, Bockmayr M, Hagele M, Seegerer P, et al. Scoring of tumor-infiltrating lymphocytes: from visual estimation to machine learning. Semin Cancer Biol. 2018;52:151–7. Article PubMed

Figure & Data

REFERENCES

Citations

Citations to this article as recorded by

Identification of nitric oxide-related key genes in pulmonary hypertension via bioinformatics and in vitro validation for therapeutic target discovery
Zhuyang Chu, Yiming Li, Jianjun Ge
Computer Methods in Biomechanics and Biomedical Engineering.2026; 29(9): 1991. CrossRef
WGCNA-Based Identification of Hub Genes and Key Pathways Involved in Obesity
Yin Yuan, Shujiao Yue, Zixuan Wu, Xuan Sun, Hongwu Wang
Molecular Biotechnology.2026; 68(5): 2459. CrossRef
Elucidating anti-triple-negative breast cancer mechanisms and mitigating toxicity of Fritillaria thunbergii Miq.: A multi-omics and network pharmacology approach
Yubin Zhu, Yuqing Zhang, Xinni Li, Luyao Zhang, Zhixin Shen
Journal of Ethnopharmacology.2026; 355: 120600. CrossRef
POLR2C, HIF1A, CD4, and CREB1 as the identified key regulators in geriatric insomnia: A comprehensive approach using systems biology and machine learning methods
Tejaswini B, Sajitha Lulu S
Computational Biology and Chemistry.2026; 120: 108777. CrossRef
Immune profiling may improve risk stratification in early-stage colorectal carcinoma
Ruben Oganesyan, Berk Kaan Aktas, Soo Hyun Lee, Amaya Pankaj, Omer Yilmaz, Deepa Patil, Vikram Deshpande, Osman Yilmaz
Human Pathology.2026; 167: 105987. CrossRef
HPCSMN: A Classification Method of Chemotherapy Sensitivity of Hypopharyngeal Cancer Based on Multimodal Network
Weiqi Fu, Haiyan Li, Xiongwen Quan, Xudong Wang, Wanwan Huang, Han Zhang
Interdisciplinary Sciences: Computational Life Sciences.2026; 18(2): 515. CrossRef
Interpretable machine learning-based predictive model for malnutrition in subacute post-stroke patients: an internal and external validation study
Ping Sun, Junqi Luan, Guotao Duan, Qingqing Sun, Genli Liu
Frontiers in Nutrition.2026;[Epub] CrossRef
Electrochemical biosensors application in diagnosis of lung Cancer and adenocarcinoma and discovery of autophagy gene characteristics: Molecular mechanisms of CFLAR and ERBB2
Jiashuo Wang, Kaiming Ren, Jungang Zhao, Xiwen Wang
Microchemical Journal.2026; 221: 117072. CrossRef
Optimal Lymph Node Count for Colorectal Cancer Surgery: A Cohort Study Utilizing Real-World Data
Xu Sun, Rui Li, Wen Zhao, Sizhe Wang, Hao Liu, Wenxing Gao, Xianqiang Liu, Dingchang Li, Guanglong Dong
Surgical Laparoscopy, Endoscopy & Percutaneous Techniques.2026;[Epub] CrossRef
Single-cell sequencing analysis and machine learning model reveal aberrant TIM-3 expression in microglia during Alzheimer’s disease progression
Zongtang Xu, Minshan Chen, Fengchu Liang, Siyuan Song, Jiawen Lei, Xingting Huang, Di Hu, Ling Tang, Pingyi Xu, Lin Lu
Journal of Translational Medicine.2026;[Epub] CrossRef
Development and Validation of a Novel Thrombosis Prediction Model for Adult Immune Thrombocytopenia (ITP-THROMBO)
Huiling Yan, Xing Hu, Lijun Zhu, Yuhan Jiang, Chen Luo, Mengya Lv, Yan Wang, Juan Tong, Changcheng Zheng
Journal of Blood Medicine.2026; Volume 17: 1. CrossRef
Machine Learning-Driven Personalized Risk Prediction: Developing an Explainable Sarcopenia Model for Older European Adults with Arthritis
Xiao Xu
Journal of Clinical Medicine.2026; 15(3): 1022. CrossRef
Childhood dyslexia risk elevated by heavy metal mixtures from e-waste: A machine learning–driven mixture modeling study
Xinle Yu, Xuanzhi Zhang, Wanyi Wen, Xiaoqi Lin, Xuanzi Xia, Dinghui Wang, Kusheng Wu, Yanhong Huang
Environmental Pollution.2026; 394: 127745. CrossRef
Artificial intelligence and multi-omics nominate TAZ as an insomnia-related diagnostic and druggable target for Parkinson’s disease patients
Wenjing Ma
Frontiers in Aging Neuroscience.2026;[Epub] CrossRef
Increased Erythrocyte Osmotic Fragility as a Risk Factor for Predicting Glaucoma
Jialiang Yang, Fang Yang, Kecheng Li, Junming Gu, Yilian Cheng, Qian Luo, Fang Hao, Bo Gong, Houbin Zhang
Translational Vision Science & Technology.2026; 15(2): 4. CrossRef
Identification and validation of biomarkers of Shenggu Zaizao Wan in the treatment of steroid-induced osteonecrosis of the femoral head by integrating network pharmacology and bulk transcriptomic
Tao Ma, Duoxian Wang, Qingsheng Xie, Yin Li, Jinpeng Wang, Xiaogang Zhang, Lingwei Yuan, Hairong He, Xianfu Han, Xuerui Liu, Jianjun Liu, Haiyang Yu, Jinqiu Wu
Frontiers in Medicine.2026;[Epub] CrossRef
Identification and verification of diagnostic biomarkers related to matrisome in patients with knee osteoarthritis based on machine learning algorithms
Jianqiao Sun, Gongchang Yu, Yingjie Zhao, Jun Zhang, Bin Shi
Electronic Journal of Biotechnology.2026; 81: 100710. CrossRef
Potential mechanisms of Si-Wu-Tang against esophageal squamous cell carcinoma: A machine learning pharmacological study
Xinbo Li, Chenchen Zhang, Jing Li, Jianfeng Yuan, Tiantian Xing, Heyuan Yang, Huijuan Shi
Medicine.2026; 105(7): e47629. CrossRef
m5C: Novel Diagnostic and Drug Repurposing Targets for Nonalcoholic Steatohepatitis
Shuxian Chen, Renquan Duan, Jingyi Qiu, Zhiyu Lei, Wei Chen, Xiumei Li, Zhenyu Gong
International Journal of Genomics.2026;[Epub] CrossRef
Prediction of lymph node metastasis in T1 colorectal cancer based on machine learning
Suyujie Shi, Xiongwu Li, Linjun Li, Haowen Zhong, Ruoyan Wang, Zhenyu Zhang, Chuyi Liao, Yun Mao, Meijie Yang, Yaying Yang
PeerJ.2026; 14: e20500. CrossRef
Development and interpretable machine learning models for classification of pancreatic pseudocyst risk in acute pancreatitis
Hailong Feng, Ping Wang, Weihan He, Liwei Shang, Mingrui Cui, Keyang Wang, Kejia An, Yingjian Zhang
Frontiers in Digital Health.2026;[Epub] CrossRef
Silencing of HAVCR2 Attenuates Skeletal Muscle Ischemia-Reperfusion Injury by Inhibiting Oxidative Stress
Huan Wang, Pang Li, Xifeng Xiong, Yijing Gao
Biochemical Genetics.2026; 64(5): 6448. CrossRef
Machine learning frameworks to accurately compute CO2 capture in porous liquids
Farag M. A. Altalbawy, Soumya V. Menon, Anupam Yadav, Subhashree Ray, Kapil Ghai, Atreyi Pramanik, Shirin Shomurotova, Gunjan Garg, Ahmad Alkhayyat, Ahmad Abumalek
Chemical Papers.2026; 80(5): 5131. CrossRef
Bioinformatics approach to identifying molecular targets of Danlou tablet against atherosclerosis: a machine learning pharmacology study
Li Wang, Miaomiao Zhang, Hongli Wang, Sihan Peng, Xiaolun Liang, Guiyu Li, Danping Xu
Naunyn-Schmiedeberg's Archives of Pharmacology.2026; 399(7): 10691. CrossRef
CX3CL1/CX3CR1 axis dysregulation contributes to epileptogenic mechanisms in focal cortical dysplasia
Lei Lei, Shengyu Yang, Yinchao Li, Jiahao Tian, Yubao Fang, Tiancai Huang, Shuda Chen, Liemin Zhou
Acta Neuropathologica Communications.2026;[Epub] CrossRef
An interpretable machine learning model for predicting distant organ metastasis after radical resection of colorectal cancer
Lin Weibin, Ni Weixiang, Zhou Junwei, Hong Weixuan, Fang Junwei, Wang Meiping, Xiao Chunhong, Huang Guoliang
Frontiers in Oncology.2026;[Epub] CrossRef
Bioinformatics combined with machine learning for potential biomarker screening and immune infiltration analysis in neonatal sepsis
Na Liu, Zhiyong Tan
Frontiers in Pediatrics.2026;[Epub] CrossRef
Development and validation of an online machine-learning tool for predicting delirium risk in older adults with type 2 diabetes: A retrospective cohort study based on the MIMIC-IV database
Lang Gao, Guangdong Wang, Xingyi Yang, Yuanshuo Ge, Shijun Tong, Xia Xiang, Chunyan Zhang, Yun Huang
Science Progress.2026;[Epub] CrossRef
Exploring histone acetylation in ischemic stroke: CREBBP and CKAP4 as candidate biomarkers linked to histone acetylation networks
Yanni Wang, Fan Yang, Jiaojiao Tan, Zhentao Guo
Frontiers in Pharmacology.2026;[Epub] CrossRef
Transcriptomics and AI-driven approaches to the diagnosis and treatment of rheumatoid arthritis
Marzena Ciechomska, Maciej Oldak, Magdalena Massalska
Frontiers in Immunology.2026;[Epub] CrossRef
Integrating single-cell RNA-seq and machine learning to dissect polyamine metabolism in metabolic dysfunction-associated steatotic liver disease
Peng Zou, Lin Sun, Zhibin Lin
Frontiers in Medicine.2026;[Epub] CrossRef
Integrated multi-omics analysis identifies TBC1D10C as an NK- and T cell-associated diagnostic biomarker in childhood obesity
Houyan Huang, Yang Huang, Yuping Zhang, Ye Jiang, Pei Shi, Yingying Wang, Sujuan Liu, Ping Xu, Juanjuan Wang, Taocheng Zhou
3 Biotech.2026;[Epub] CrossRef
Structure characteristics of a polysaccharide isolated from safflower and its neuroprotective effects against cerebral ischemia/reperfusion injury by LASSO regression analysis of transcriptomic and metabolomic data
Wentao Li, Ruoqi Li, Yihang Lu, Weihao Chen, Jianjun Wu, Li Shen, Haitong Wan, Yu He
Carbohydrate Polymers.2026; 386: 125425. CrossRef
Identification of Mitochondrial Dysfunction-Related Candidate Biomarkers and Analysis of the Immune Cell Infiltration in Epilepsy
Zhenghan Wu, Honglong Pei, Rongzhu Huang, Wenqiang Huang, Anbo Xiong, Zhuoqi Dong, Shipeng Li, Yuyun Chen, Wei Wang, Hualin Yu, Renli Qi, Jinghui Li
Molecular Neurobiology.2026;[Epub] CrossRef
Identification and validation of cholesterol homeostasis and NK cell-associated predictive and therapeutic models for Coronary atherosclerotic heart disease patients with ST-elevated myocardial infarction: Insights from artificial intelligence and multi-o
Xuemei Chen, Yifan Zhao, Xixi Guo, Qiushan Man, Ming Wang, Zhou Wang, Jun Cheng, Shijian Chen
Computational Biology and Chemistry.2026; 124: 109103. CrossRef
Construction and validation of a predictive model for the risk of emergence delirium in older adult patients
Yi Xin, Bin He, Xiao-Hui Wei, Ya-Ling Yan, Chen Huang, Chun-Yan Gao, Shuo Wang, Guang-Ming Zhang, Rui Li, Ying Wu
World Journal of Psychiatry.2026;[Epub] CrossRef
Bioinformatics reveals the prognostic potential of manganese metabolism-related genes in lung adenocarcinoma
Jianhua Zha, Xiaoming Liu, Shansong Gao, Jiasi He, Zhi Zhang
Functional & Integrative Genomics.2026;[Epub] CrossRef
Construction and internal temporal validation of a LASSO regression-based risk assessment model for non-suicidal self-injury addiction-like features in adolescents and young adults with depression
Houqin Huang, Fengmei Lu, Shuiqin Cao, Chuan Jiang, Jianyan Peng, Yuhao Tong, Pengyuan Wang, Yi Liu, Zixiang Ye
Frontiers in Psychology.2026;[Epub] CrossRef
AI-Guided Ferroptosis Biomarker Discovery and Routine Laboratory–Based Machine Learning for Predicting Nodal Metastasis in Colon Cancer
Nail Besli, Talar Vartanoglu Aktokmakyan, Adil Koyuncu, Bahar Sarikamis Johnson, Ulkan Celik
Metabolites.2026; 16(8): 557. CrossRef
Multiparametric Radiomics for Characterization and Outcome Prediction in Colorectal Cancer: The Central Role of Diagnostic Imaging
David Farkas, József Baracs, Zsombor Ritter, David Sipos
Cancers.2026; 18(16): 2578. CrossRef
A machine learning–based approach to establish predictors of recurrent tuberculosis in the Guelmim region of Morocco
Ahmed Oubaasri, Fatima Zahra El Mskini, Asmae Labyad, Rachida Hassikou, Aicha Madkour, Slimane Mehdad, Souad Benaich
Biostatistics & Epidemiology.2026;[Epub] CrossRef
Assessing prospective molecular biomarkers and functional pathways in severe asthma based on a machine learning method and bioinformatics analyses
Ya-Da Zhang, Yi-Ren Chen, Wei Zhang, Bin-Qing Tang
Journal of Asthma.2025; 62(3): 465. CrossRef
A web-based tool for cancer risk prediction for middle-aged and elderly adults using machine learning algorithms and self-reported questions
Xingjian Xiao, Xiaohan Yi, Nyi Nyi Soe, Phyu Mon Latt, Luotao Lin, Xuefen Chen, Hualing Song, Bo Sun, Hailei Zhao, Xianglong Xu
Annals of Epidemiology.2025; 101: 27. CrossRef
Risk stratification scores for lymph node metastases in T1 colorectal cancer—A systematic review
Rakesh Quinn, Giuleta Jamsari, Ewan MacDermid
Colorectal Disease.2025;[Epub] CrossRef
Artificial Intelligence in Lymph Node Metastasis Prediction for T1 Colorectal Cancer: Promise and Challenges
Jung Ho Bae
Gut and Liver.2025; 19(1): 3. CrossRef
Utilizing Ensemble Learning and Dimension Reduction in Predicting Stock Prices: A Transparent Methodology with Insights from Explainable AI
Nabanita Das, Bikash Sadhukhan, Chayan Ghosh, Avigyan Chowdhury, Satyajit Chakrabarti
SN Computer Science.2025;[Epub] CrossRef
Bioinformatics approach reveals the critical role of inflammation-related genes in age-related hearing loss
Xi Gu, Chenyu Chen, Yuqing Chen, Chaojun Zeng, Yanchun Lin, Ruosi Guo, Shujin Xu, Chang Lin
Scientific Reports.2025;[Epub] CrossRef
Genome-wide identification and expression analysis of phytochrome gene family in Aikang58 wheat (Triticum aestivum L.)
Zhu Yang, Wenjie Kan, Ziqi Wang, Caiguo Tang, Yuan Cheng, Dacheng Wang, Yameng Gao, Lifang Wu
Frontiers in Plant Science.2025;[Epub] CrossRef
Following intravenous thrombolysis, the outcome of diabetes mellitus associated with acute ischemic stroke was predicted via machine learning
Xiaoqing Liu, Miaoran Wang, Rui Wen, Haoyue Zhu, Ying Xiao, Qian He, Yangdi Shi, Zhe Hong, Bing Xu
Frontiers in Pharmacology.2025;[Epub] CrossRef
An early lung cancer diagnosis model for non-smokers incorporating ct imaging analysis and circulating genetically abnormal cells (CACs)
Ran Ni, Yongjie Huang, Lei Wang, Hongjie Chen, Guorui Zhang, Yali Yu, Yinglan Kuang, Yuyan Tang, Xing Lu, Hong Liu
BMC Cancer.2025;[Epub] CrossRef
Identification and validation of immune-related biomarkers and polarization types of macrophages in keloid based on bulk RNA-seq and single-cell RNA-seq analysis
Yuzhu Zhang, Chenglong Fang, Lizhong Zhang, Fengyu Ma, Meihong Sun, Ning Zhang, Nan Bai, Jun Wu
Burns.2025; 51(3): 107413. CrossRef
Comprehensive analysis of mitochondrial solute carrier family 25 (SLC25) identifies member 19 (SLC25A19) as a regulatory factor in hepatocellular carcinoma
Xueke Gao, Yangtao Xu, Xinyao Hu, Jiayu Chen, Daoming Zhang, Ximing Xu
Gene.2025; 944: 149299. CrossRef
A dynamic nomogram predicting nosocomial infections in patients after colon cancer surgery
Xue Yao, Shuhui Wang, Anning Lu, Yun Xu, Na Li
Frontiers in Oncology.2025;[Epub] CrossRef
Machine learning models to predict 6-month mortality risk in home-based hospice patients with advanced cancer
Wan Cheng, Jianwei Zheng, Yuanfeng Lu, Guojuan Chen, Zheng Zhu, Hong Wu, Yitao Wei, Huimin Xiao
Asia-Pacific Journal of Oncology Nursing.2025; 12: 100679. CrossRef
Interpretable machine learning model for predicting depression in middle-aged and elderly Chinese arthritis patients: A nationwide prospective cohort study
Xiao Xu, Hai-Yan Huang, Shu-yun Wang, Shen-Yee Tan, Hong-Hui Chen, Ming-Ming Zhou, Mei-juan Qian
Archives of Gerontology and Geriatrics.2025; 134: 105810. CrossRef
Prediction of Lymph Node Metastasis in Colorectal Cancer Using Intraoperative Fluorescence Multi-Modal Imaging
Xiaobo Zhu, He Sun, Yuhan Wang, Gang Hu, Lizhi Shao, Song Zhang, Fucheng Liu, Chongwei Chi, Kunshan He, Jianqiang Tang, Yu An, Jie Tian, Zhenyu Liu
IEEE Transactions on Medical Imaging.2025; 44(3): 1568. CrossRef
Tertiary Lymphoid Structures Are Associated with Enhanced Macrophage Activation and Immune Checkpoint Expression and Predict Outcome in Cervical Cancer
Laurent Gorvel, Marylou Panouillot, Marie-Sarah Rouvière, Emilien Billon, Stéphane Fattori, Jumaporn Sonongbua, Nicolas Boucherit, Amira Ben Amara, Olivia Quilichini, Samuel Granjeaud, Clara Degos, Jacques A. Nunès, Xavier Carcopino, Eric Lambaudie, Anne-
Cancer Immunology Research.2025; 13(5): 712. CrossRef
A predictive model for the transformation from cervical inflammation to cancer based on tumor immune-related factors
Wenjie Wang, Chuntao Huang, Shiwen Bi, Huiting Liang, Songlin Li, Tingting Lu, Ben Liu, Yong Tang, Qi Wang
Frontiers in Immunology.2025;[Epub] CrossRef
Application of an interpretable machine learning method to predict the risk of death during hospitalization in patients with acute myocardial infarction combined with diabetes mellitus
Zhijun Bu, Siyu Bai, Chan Yang, Guanhang Lu, Enze Lei, Youzhu Su, Zhaoge Han, Muyan Liu, Jingge Li, Linyan Wang, Jianping Liu, Yao Chen, Zhaolan Liu
Acta Cardiologica.2025; 80(4): 358. CrossRef
Ferroptosis regulation by traditional chinese medicine for ischemic stroke intervention based on network pharmacology and data mining
Jun Lei, Wei Chen, Yaodong Gu, Xueyan Lv, Xingyu Kang, Xicheng Jiang, Ruo Wang
PLOS ONE.2025; 20(4): e0321751. CrossRef
Establishment and validation of a nomogram model for surgical site infections after posterior lumbar interbody fusion: a retrospective observational study
Zongke Long, Xiaole Hu, Jian Liu, Peiyun Zhou, Bingyan Zhang, Simeng Zhang, Huimin Wei, Wenran Qu, Xiaorong Luan
Neurosurgical Review.2025;[Epub] CrossRef
Identification of m5C RNA modification-related gene signature for predicting prognosis and immune microenvironment-related characteristics of heart failure
Zirui Liu, Rui Feng, Ying Xu, Meili Liu, Haocheng Wang, Yu Lu, Weiqi Wang, Jikai Wang, Cao Zou
Hereditas.2025;[Epub] CrossRef
CLDN9 and hsa-miR-4496 as non-invasive biomarkers for gastric cancer detection
Qiongxia Hu, Lu Han, Jinglin Wang, Fei Li, Hongfei Pu, Yue Shi
Discover Oncology.2025;[Epub] CrossRef
Development and validation of a nomogram model of lung metastasis in breast cancer based on machine learning algorithm and cytokines
Zhaoyi Li, Hao Miao, Wei Bao, Lansheng Zhang
BMC Cancer.2025;[Epub] CrossRef
Acute Myeloid Leukemia Genome Characterization Study and Subtype Classification Employing Feature Selection and Bayesian Networks
Zhenzhen Li, Jingwen Li, Sifan Li, Yangyang Wang, Jihan Wang
Biomedicines.2025; 13(5): 1067. CrossRef
Clinical potential and experimental validation of prognostic genes in hepatocellular carcinoma revealed by risk modeling utilizing single cell and transcriptome constructs
Hang Deng, Xu Wang, Zi-Ang Jiang, Jian Xu, Yu Zhang, Yao Zhou, Jun Gong, Xiang-Yu Lu, Yi-Fu Hou, Hao Zhang
Frontiers in Immunology.2025;[Epub] CrossRef
Identifying GAP43, NMU, and TEX29 as Potential Prognostic Biomarkers for COPD Combined With Lung Cancer Patients Using Machine Learning
Zhilong Xu, Kaiyao Zhang, Ao Zeng, Yanze Yin, KeYi Chen, Chao Wang, Xinyun Fang, Abudumijiti Abuduwayiti, JiaRui Wang, Jie Dai, Gening Jiang
The Journal of Gene Medicine.2025;[Epub] CrossRef
Identification of DSC2 as a Key Biomarker for Induction Chemotherapy Sensitivity in Locally Advanced Laryngeal and Hypopharyngeal Squamous Cell Carcinoma
Qianyue Yang, Jiayi Liu, Zhiwei Lin, Shuang Liu, Zhaoming Hu, Xiaowen Zhang, Baoqing Sun
Journal of Chemical Information and Modeling.2025; 65(12): 5899. CrossRef
The Construction of a New Prognostic Model of Breast Cancer and the Exploration of Drug Sensitivity Based on Machine Learning for Glycosylation-Related Genes
Jia-Ning Zhang, Xi-Rui Zhou, Zi-Lu Yi, Xin-Yu Tian, Hong Liu
Clinical Breast Cancer.2025; 25(6): e756. CrossRef
Multi-Omics-Based Analysis of the Effect of Longevity Genes on the Immune Relevance of Colorectal Cancer
Yichu Huang, Guangtao Min, Hongpeng Wang, Lei Jiang
Biomedicines.2025; 13(5): 1085. CrossRef
EM-DeepSD: A Deep Neural Network Model Based on Cell-Free DNA End-Motif Signal Decomposition for Cancer Diagnosis
Zhi-Yang Zhao, Chang-Ling Huang, Tong-Min Wang, Shi-Hao Zhou, Lu Pei, Wen-Hui Jia, Wei-Hua Jia
Diagnostics.2025; 15(9): 1156. CrossRef
Identification of thyroid cancer biomarkers using WGCNA and machine learning
Gaofeng Hu, Wenyuan Niu, Jiaming Ge, Jie Xuan, Yanyang Liu, Mengjia Li, Huize Shen, Shang Ma, Yuanqiang Li, Qinglin Li
European Journal of Medical Research.2025;[Epub] CrossRef
Identification of neutrophil extracellular trap-related biomarkers in ulcerative colitis based on bioinformatics and machine learning
Jiao Li, Yupei Liu, Zhiyi Sun, Suqi Zeng, Caisong Zheng
Frontiers in Genetics.2025;[Epub] CrossRef
Correlation Analysis Between Low-Density Lipoprotein and Clinical Prognosis in Aneurysmal Subarachnoid Hemorrhage: A Single-Center Study
Zhixing Wang, Yongxing Li, Mengdi Gao, Yifan Xu, Zhe Wang
Journal of Craniofacial Surgery.2025; 36(6): e757. CrossRef
Integrated analysis of single-cell RNA-seq and spatial transcriptomics to identify the lactylation-related protein TUBB2A as a potential biomarker for glioblastoma in cancer cells by machine learning
Yifan Xu, Chonghui Zhang, Jinpeng Wu, Pin Guo, Nan Jiang, Chao Wang, Yugong Feng
Frontiers in Immunology.2025;[Epub] CrossRef
Machine learning assists regulated cell death crucial biomarker selection in adenocarcinoma of the lung: biological data testing and cell assay determination
Han Ning, Ying Jiang, Mengli Zheng, Gao Yang, Lianjun Ma, Yachao Zhao
Discover Oncology.2025;[Epub] CrossRef
Predicting the risk of lean non-alcoholic fatty liver disease based on interpretable machine models in a Chinese T2DM population
Shixue Bao, Qiankai Jin, Tieqiao Wang, Yushan Mao, Guoqing Huang
Frontiers in Endocrinology.2025;[Epub] CrossRef
Investigation of the mechanism of Xiaoyin Jiedu Yin in the treatment of psoriasis based on bioinformatics, machine learning
Ru-Nan Fang, Yang Zhou, Yang Shen, Yuan Sun, Jian-Hong Li
Frontiers in Chemistry.2025;[Epub] CrossRef
Hybrid Intelligence-Driven Nanopolymeric Sensor for Precise Electrochemical Vitamin C Analysis, Free from LoD: Application in Real Lemon Juice
Emre Dokuzparmak, Emine Sezer, Timuçin Güner, Esra Yaşar, Hilal Özçeli̇k, Sinan Akgöl
ACS Applied Electronic Materials.2025; 7(15): 6980. CrossRef
A risk model based on signature genes predicts prognosis and associates with tumor immunity, drug sensitivity in breast cancer
Yuan Li, Hao Li, Jichuan Quan, Ping Bi, Xuemei Liu, Yanwei Yao, Yanqin Peng, Congrui Wang, Xiaofang Gao, Junfang Duan, Xiaoru Wang, Jian Peng
Cancer Biomarkers.2025;[Epub] CrossRef
Tumor Budding as a Risk Factor for Lymph Node Metastasis and Local Recurrence in pT1 Colorectal Cancer: A Systematic Review and Meta-Analysis
Heng Zhang, Femke Simmer, Alessandro Lugli, Iris D. Nagtegaal
Gastro Hep Advances.2025; 4(9): 100713. CrossRef
Interpretable machine learning approaches for predicting prostate cancer by using multiple heavy metal exposures based on the data from NHANES 2003–2018
Zu-Ming You, Yuan-Sheng Li, Fan-Shuo Meng, Rui-Xiang Zhang, Chen-Xi Xie, Zhijiang Liang, Ji-Yuan Zhou
Ecotoxicology and Environmental Safety.2025; 302: 118730. CrossRef
CT-based intratumoral and peritumoral radiomics to predict the treatment response to hepatic arterial infusion chemotherapy plus lenvatinib and PD-1 in high-risk hepatocellular carcinoma cases: a multi-center study
Zihao Liu, Xinge Li, Yong Huang, Xu Chang, Hong Zhang, Xiaodong Wu, Yanzhao Diao, Fengling He, Junyong Sun, Baomin Feng, Hexin Liang
Hepatology International.2025; 19(6): 1397. CrossRef
Paternally Expressed Gene 10 Promoter Methylation Level as a Predictor of HBeAg Seroconversion in Chronic Hepatitis B Patients
Pengyu Luo, Nan Chen, Yuna Tang, Jing Wang, Pei Liu, Yuchen Fan, Huihui Liu, Kai Wang
Journal of Medical Virology.2025;[Epub] CrossRef
The role of mitochondria-related genes and immune infiltration in carotid atherosclerosis: identification of hub targets through bioinformatics and machine learning approaches
Dan Liu, Kun Guo, Min Li, Xiaochen Yu, Xue Guan, Xiuru Guan
Frontiers in Genetics.2025;[Epub] CrossRef
Machine learning integration of bulk and single-cell RNA-seq data reveals glycolytic heterogeneity in colorectal cancer
Yuanyuan Du, Zefeng Miao, Peng Li, Dan Feng, Mulin Liu, Aifang Ji, Shijun Li
Medical Oncology.2025;[Epub] CrossRef
Identification and Experimental Validation of OS-Related Gene Sets Based on Integrated Analysis of Single-Cell and Bulk RNA Sequencing Data with Machine Learning in Patients with Sepsis
Linfeng Tao, Wei Tian, Ping Li, Yan Chen, Jun Liu
Inflammation.2025; 48(6): 4549. CrossRef
Prognostic Significance of R-Loops in Lung Adenocarcinoma: Implications for Immune Response and Drug Sensitivity
Jianjun Jiang, Yingxin Zhang, Biying Men, Yujing He, Liang Yun, Fangfang Li, Xuguang Rao, Kaican Cai, Shuan Rao
Cancer Management and Research.2025; Volume 17: 1625. CrossRef
Significant adverse prognostic events in patients with urosepsis: a machine learning based model development and validation study
Yiqu Wei, Wanqing Xu, Shuo Yang, Congfeng Zhang, Jia Wang, Xianyao Wan
Frontiers in Cellular and Infection Microbiology.2025;[Epub] CrossRef
Dendrobin A inhibits gastric cancer: Mechanistic insights supported by integrated evidence
Yonghao Fan, Yan Chen, Wenyan Lu, Kaijia Shi, Yangyang Zhao, Cheng Zhang, Zhihua Shen, Shaojiang Zheng, Wei Jie
Phytomedicine.2025; 147: 157215. CrossRef
Analysis of the toxicity and mechanisms of osteoporosis caused by cigarette toxicants using network toxicology and molecular docking techniques
Wenbo Xie, Chao Song, Pandeng Hao, Feilong Li, Xianghan Hou, Jingwen Chen, Zongchao Liu, Zhenlong Wang
Science of The Total Environment.2025; 1000: 180414. CrossRef
Machine Learning Integration of Bulk and Single-Cell RNA-Seq Data Reveals Cathepsin B as a Central PANoptosis Regulator in Influenza
Bin Liu, Lin Zhu, Caijuan Zhang, Dunfang Wang, Haifan Liu, Jianyao Liu, Jingwei Sun, Xue Feng, Weipeng Yang
International Journal of Molecular Sciences.2025; 26(17): 8533. CrossRef
Prediction Models of Microinvasive Cervical Cancer in High-Grade Squamous Intraepithelial Lesion Treatment by Loop Electrosurgical Excision Procedure
Maodan Huang, Xiaohong Chen, Xin Lin, Yuxiang Yang, Lu Liu, Youzhong Zhang, Ronglong Wang, Wei Chen
Risk Management and Healthcare Policy.2025; Volume 18: 2921. CrossRef
Interpretable Machine Learning Models for Predicting Lateral Pelvic Lymph Node Metastasis in Rectal Cancer: A Chinese Multicenter Retrospective Study
Tixian Xiao, Wei Zhao, Zhen Sun, Fangze Wei, Fuqiang Zhao, Fei Huang, Zeyu Wu, Junge Bai, Xin Wang, Qian Liu
JCO Precision Oncology.2025;[Epub] CrossRef
Predicting intraoperative blood loss risk in severe lumbar disc herniation patients undergoing PLIF: a multicenter cohort study using ensemble learning
Ning Shen, Yusi Zhang, Zhendong Ding, Runmin Li, Xubin Quan, Xiaozhu Liu, Yang Zhang, Tianyu Xiang, Yingang Zhang, Chengliang Yin, Wenle Li
International Journal of Surgery.2025; 111(9): 5904. CrossRef
Polybrominated diphenyl ether profiles in adipose tissues of breast cancer patients and their carcinogenic potential investigation based on network toxicology and molecular docking
Qihao Zhao, Xi Liu, Haoyi Chen, Yingming Jin, Qian Chen, Yiteng Huang, Lin Peng
Frontiers in Chemistry.2025;[Epub] CrossRef
Role of SAA1 in mediating renal fibrosis through TLR4-dependent NF-κB activation
Xiao Wei, Lixiang Feng, Yuan Yuan, Qihui Kuang, Hong Yu, Jun Yang, Xiong Wang, Pengcheng Luo
BMC Nephrology.2025;[Epub] CrossRef
Mendelian Randomization and Machine Learning Reveal Immune Cell and Gene Drivers in Systemic Lupus Erythematosus
Luofei Huang, Jian shi, Han Li, Quanzhi Lin
Brain and Behavior.2025;[Epub] CrossRef
Explainable machine learning model for predicting the outcome of acute ischemic stroke after intravenous thrombolysis
Fanhai Bu, Runlu Cai, Wei Zhang, Xiaohong Tang, Guiyun Cui, Xinxin Yang
Frontiers in Neurology.2025;[Epub] CrossRef
Relationships Among Mobile Internet Use, Social Support, and Depressive Symptoms: Prospective Cohort Study Among Community Residents
Yingyue Xu, Meiqi Wang, Qixiu Li, Xiaoying Su, Long Sun
Journal of Medical Internet Research.2025; 27: e76567. CrossRef
Optimizing methane catalytic hydrogen production via a hybrid Boruta-XGB and stacking ensemble machine learning framework
Xinyi Liu, Yaxuan Heng, Linmeng Zhou, Huiru Gao, Yanyan Ji, Wu Zhang
International Journal of Hydrogen Energy.2025; 181: 151812. CrossRef
Prediction of prognosis in T4 or N3 locally advanced nasopharyngeal carcinoma receiving chemoradiotherapy using machine learning methods
Zheng Ma, Weijie Liu, Xiaoya Luo, Xinran Niu, Yanmei Li, Yuanling Ma, Li Hou
Frontiers in Oncology.2025;[Epub] CrossRef
Development and validation of an endoscopic diagnostic model for sessile serrated lesions based on machine learning algorithms
Xinying Yu, Lianyu Li, Qiang He
Frontiers in Medicine.2025;[Epub] CrossRef
Applying machine learning for perioperative adverse event prediction: a narrative review toward better clinical efficacy and usability
Xuechao Hao, Yaqiang Wang, Ke Li, Tao Zhu, Vitaly Herasevich
Anesthesiology and Perioperative Science.2025;[Epub] CrossRef
The diagnostic accuracy of deep learning-based AI models in predicting lymph node metastasis in T1 and T2 colorectal cancer: A systematic review and meta-analysis
Qihong Guo, Ruiping Wang, Yichen Guo
Medicine.2025; 104(45): e45172. CrossRef
Identification of mitochondrial-related genes to evaluate the immune infiltration and prognosis of lung adenocarcinoma
Yutong Ge, Ao Sun, Tao Yu, Shaokun Yu, Kaihua Lu
PeerJ.2025; 13: e20262. CrossRef
The toxicological impact of PET-MPs exposure on atherosclerosis: insights from network toxicology, molecular docking, and machine learning
Qiang Wang, Hu Liang, Lang Li
Scientific Reports.2025;[Epub] CrossRef
Predicting Age-Related Hearing Loss in Community-Dwelling Older Adults: Multicenter Retrospective Cohort Study
Jing Li, Shuai Jin, Liu Sun, Jun-E Liu, Qiang Shen, Miao Shang, Hanting Wang, Yuanyuan Zhao
Interactive Journal of Medical Research.2025; 14: e81135. CrossRef
Development and explanation of a machine learning model for identifying non-localized early-onset T1 colorectal cancer
Yin Zhang, Fuzhou Han, Mingyu Zheng, Duo Xu, Nan Yao, Wenqiang Li, Jun Qu
Discover Oncology.2025;[Epub] CrossRef
Development and validation of an AI-augmented deep learning model for survival prediction in de novo metastatic colorectal cancer
Merih Yalçıner, Efe Cem Erdat, Engin Eren Kavak, Güngör Utkan
Discover Oncology.2025;[Epub] CrossRef
Inflammatory markers partially mediate the association between volatile organic compounds exposure and hyperlipidemia: a nationally representative cross-sectional study from NHANES
Yaxiong Nie, Zining He, Bei Liu, Jiaai Li, Yanyu Liu, Xin Su, Zhiqiang Yan, Zheng Li, Chang Yan, Qian Lu, Yanfang Fu, Wanyu Yang, Yutong He
Frontiers in Public Health.2025;[Epub] CrossRef
Clinical applications of artificial intelligence in the histopathology of lymphoma: diagnosis, treatment and prognosis
Mengyao Kang, Zibo Yang, Tian Yu, Dongyu Li, Zhiqiong Wang, Liting Chen
Discover Oncology.2025;[Epub] CrossRef
Targeting PSMB5-induced PANoptosis in bladder cancer: multi-omics insights and TCM candidate discovery
Zhe Chang, Jirong Wang, Jiajia Cao, Xinpeng Fan, Kunpeng Li, Chenyang Wang, Yalong Zhang, Li Wang, Jianwei Yang, Siyu Chen, Li Yang
Frontiers in Immunology.2025;[Epub] CrossRef
Development of robust machine learning models to estimate hydrochar higher heating value and yield based upon biomass proximate analysis
Guoliang Hou, Ahmad Alkhayyat, Ahmad Almalkawi, Anupam Yadav, H. S. Shreenidhi, Vishnu Saini, Shirin Shomurotova, Devendra Singh, Vatsal Jain, Aseel Smerat, Ahmad Khalid
Bioresources and Bioprocessing.2025;[Epub] CrossRef
Comprehensive molecular characterization of high-stemness gastric cancer cells using single-cell transcriptomics, spatial mapping, and machine learning
Ziyi Wang, Xuehao Li, Jin Wang, Huidong Yu, Defeng Zhao, Yan Xu, Siyu Zhou, Wanfu Men
npj Precision Oncology.2025;[Epub] CrossRef
Machine learning-based survival prediction in colorectal cancer combining clinical and biological features
Lucas M. Vieira, Natasha A.N. Jorge, João B. Sousa, João C. Setubal, Peter F. Stadler, Maria E.M.T. Walter
Oncotarget.2025; 16(1): 834. CrossRef
Analysis of CSF2RA Expression and Its Functional and Clinical Significance in Gastric Cancer Based on Bioinformatics
Rong He, Yongle Li, Lin Jiang, Mingtao Luo, Shiquan Wei, Lihe Jiang
Journal of Biosciences and Medicines.2025; 13(09): 405. CrossRef
Additional staining for lymphovascular invasion is associated with increased estimation of lymph node metastasis in patients with T1 colorectal cancer: Systematic review and meta‐analysis
Jun Watanabe, Katsuro Ichimasa, Yuki Kataoka, Atsushi Miki, Hidehiro Someko, Munenori Honda, Makiko Tahara, Takeshi Yamashina, Khay Guan Yeoh, Shigeo Kawai, Kazuhiko Kotani, Naohiro Sata
Digestive Endoscopy.2024; 36(5): 533. CrossRef
Risk assessment in pT1 colorectal cancer
Emma Jane Norton, Adrian C Bateman
Journal of Clinical Pathology.2024; 77(4): 225. CrossRef
Identification and Analysis of Immune Microenvironment-Related Genes for Keloid Risk Prediction and Their Effects on Keloid Proliferation and Migration
Yongyan Pei, Yikai Wu, Mengqi Zhang, Xuemin Su, Hua Cao, Jiaji Zhao
Biochemical Genetics.2024; 62(4): 3174. CrossRef
Multiple machine-learning tools identifying prognostic biomarkers for acute Myeloid Leukemia
Yujing Cheng, Xin Yang, Ying Wang, Qi Li, Wanlu Chen, Run Dai, Chan Zhang
BMC Medical Informatics and Decision Making.2024;[Epub] CrossRef
Artificial Intelligence Applications in the Treatment of Colorectal Cancer: A Narrative Review
Jiaqing Yang, Jing Huang, Deqian Han, Xuelei Ma
Clinical Medicine Insights: Oncology.2024;[Epub] CrossRef
Transcriptomic and machine learning analyses identify hub genes of metabolism and host immune response that are associated with the progression of breast capsular contracture
Yukun Mao, Xueying Hou, Su Fu, Jie Luan
Genes & Diseases.2024; 11(3): 101087. CrossRef
Bioinformatics and Machine Learning Methods Identified MGST1 and QPCT as Novel Biomarkers for Severe Acute Pancreatitis
Yang Sun, Jingjun Xie, Jun Zhu, Yadong Yuan
Molecular Biotechnology.2024; 66(5): 1246. CrossRef
Multi-omics reveals the role of ENO1 in bladder cancer and constructs an epithelial-related prognostic model to predict prognosis and efficacy
Zhixiong Su, Lijie You, Yufang He, Jingbo Chen, Guifeng Zhang, Zhenhua Liu
Scientific Reports.2024;[Epub] CrossRef
Use of artificial intelligence in the management of T1 colorectal cancer: a new tool in the arsenal or is deep learning out of its depth?
James Weiquan Li, Lai Mun Wang, Katsuro Ichimasa, Kenneth Weicong Lin, James Chi-Yong Ngu, Tiing Leong Ang
Clinical Endoscopy.2024; 57(1): 24. CrossRef
SLCO4A1, as a novel prognostic biomarker of non‑small cell lung cancer, promotes cell proliferation and migration
Shihao Li, Zihao Li, Lan Huang, Zhenyang Geng, Feng Li, Bin Wu, Yinliang Sheng, Yifan Xu, Bowen Li, Yiming Xu, Zhuoyu Gu, Yu Qi
International Journal of Oncology.2024;[Epub] CrossRef
Overexpression of circulating CD38+ NK cells in colorectal cancer was associated with lymph node metastasis and poor prognosis
Xueling Wang, Haoran Li, Huixian Chen, Kehua Fang, Xiaotian Chang
Frontiers in Oncology.2024;[Epub] CrossRef
Management after non-curative endoscopic resection of T1 rectal cancer
Hao Dang, Daan A. Verhoeven, Jurjen J. Boonstra, Monique E. van Leerdam
Best Practice & Research Clinical Gastroenterology.2024; 68: 101895. CrossRef
Prediction of Lymph Node Metastasis in T1 Colorectal Cancer Using Artificial Intelligence with Hematoxylin and Eosin-Stained Whole-Slide-Images of Endoscopic and Surgical Resection Specimens
Joo Hye Song, Eun Ran Kim, Yiyu Hong, Insuk Sohn, Soomin Ahn, Seok-Hyung Kim, Kee-Taek Jang
Cancers.2024; 16(10): 1900. CrossRef
Identification of immune characteristic biomarkers and therapeutic targets in cuproptosis for sepsis by integrated bioinformatics analysis and single-cell RNA sequencing analysis
Tianfeng Wang, Xiaowei Fang, Ximei Sheng, Meng Li, Yulin Mei, Qing Mei, Aijun Pan
Heliyon.2024; 10(5): e27379. CrossRef
Possibilities and prospects of artificial intelligence in the treatment of colorectal cancer (review)
A. Yu. Kravchenko, E. V. Semina, V. V. Kakotkin, M. A. Agapov
Koloproktologia.2024; 23(2): 184. CrossRef
A risk prediction nomogram for resistant hypertension in patients with obstructive sleep apnea
Hongze Lin, Chen Zhou, Jiaying Li, Xiuqin Ma, Yan Yang, Taofeng Zhu
Scientific Reports.2024;[Epub] CrossRef
Unveiling the best predictive models for early‑onset metastatic cancer: Insights and innovations (Review)
Liqing Yu, Zhenjun Huang, Ziqi Xiao, Xiaofu Tang, Ziqiang Zeng, Xiaoli Tang, Wenhao Ouyang
Oncology Reports.2024;[Epub] CrossRef
Exploration of the molecular biological mechanisms and review of postoperative radiotherapy cases in tenosynovial giant cell tumors
Tianwei Zhang, Bin Zeng, Ke Liu, Qin Zeng, Na Wang, Ling Peng, Hongbo Qiu, Xiaomei Chen, Lin Wang
Frontiers in Oncology.2024;[Epub] CrossRef
Radiomics-based machine learning in the differentiation of benign and malignant bowel wall thickening
Hande Melike Bülbül, Gülen Burakgazi, Uğur Kesimal, Esat Kaba
Japanese Journal of Radiology.2024; 42(8): 872. CrossRef
Potential Mechanism of Tibetan Medicine Liuwei Muxiang Pills against Colorectal Cancer: Network Pharmacology and Bioinformatics Analyses
Shaochong Qi, Xinyu Liang, Zijing Wang, Haoran Jin, Liqun Zou, Jinlin Yang
Pharmaceuticals.2024; 17(4): 429. CrossRef
A new clinical model for predicting lymph node metastasis in T1 colorectal cancer
Kai Wang, Hui He, Yanyun Lin, Yanhong Zhang, Junguo Chen, Jiancong Hu, Xiaosheng He
International Journal of Colorectal Disease.2024;[Epub] CrossRef
Comprehensive analysis of the interaction of antigen presentation during anti‐tumour immunity and establishment of AIDPS systems in ovarian cancer
Wenhuizi Sun, Ping Xu, Kefei Gao, Wenqin Lian, Xiang Sun
Journal of Cellular and Molecular Medicine.2024;[Epub] CrossRef
Use of artificial intelligence for the prediction of lymph node metastases in early-stage colorectal cancer: systematic review
Nasya Thompson, Arthur Morley-Bunker, Jared McLauchlan, Tamara Glyn, Tim Eglinton
BJS Open.2024;[Epub] CrossRef
Machine learning for predicting colon cancer recurrence
Erkan Kayikcioglu, Arif Hakan Onder, Burcu Bacak, Tekin Ahmet Serel
Surgical Oncology.2024; 54: 102079. CrossRef
A machine learning screening model for identifying the risk of high-frequency hearing impairment in a general population
Yi Wang, Xinmeng Yao, Dahui Wang, Chengyin Ye, Liangwen Xu
BMC Public Health.2024;[Epub] CrossRef
Potential clinical value of fibrinogen-like protein 1 as a serum biomarker for the identification of diabetic cardiomyopathy
Yao Liu, Min Wang, Jia-Bao Su, Xiao Fu, Guan-Li Zheng, Shan Guo, Li-Juan Zhang, Qing-Bo Lu
Scientific Reports.2024;[Epub] CrossRef
CXCR4-mediated neutrophil dynamics in periodontitis
Xuanwen Xu, Tiange Li, Jingqi Tang, Danlei Wang, Yi Zhou, Huiqing Gou, Lu Li, Yan Xu
Cellular Signalling.2024; 120: 111212. CrossRef
Hspb1 and Lgals3 in spinal neurons are closely associated with autophagy following excitotoxicity based on machine learning algorithms
Lei Yan, Zihao Li, Chuanbo Li, Jingyu Chen, Xun Zhou, Jiaming Cui, Peng Liu, Chong Shen, Chu Chen, Hongxiang Hong, Guanhua Xu, Zhiming Cui, Suyan Tian
PLOS ONE.2024; 19(5): e0303235. CrossRef
Impact of artificial intelligence in the management of esophageal, gastric and colorectal malignancies
Ayrton Bangolo, Nikita Wadhwani, Vignesh K Nagesh, Shraboni Dey, Hadrian Hoang-Vu Tran, Izage Kianifar Aguilar, Auda Auda, Aman Sidiqui, Aiswarya Menon, Deborah Daoud, James Liu, Sai Priyanka Pulipaka, Blessy George, Flor Furman, Nareeman Khan, Adewale Pl
Artificial Intelligence in Gastrointestinal Endoscopy.2024;[Epub] CrossRef
Identification and Validation of Nicotinamide Metabolism-Related Gene Signatures as a Novel Prognostic Model for Hepatocellular Carcinoma
Sijia Yang, Ang Li, Lihong Lv, Jinxin Duan, Zhihua Zheng, Wenfeng Zhuo, Jun Min, Jinxing Wei
OncoTargets and Therapy.2024; Volume 17: 423. CrossRef
Machine learning algorithms integrate bulk and single-cell RNA data to unveil oxidative stress following intracerebral hemorrhage
Chaonan Du, Cong Wang, Zhiwei Liu, Wenxuan Xin, Qizhe Zhang, Alleyar Ali, Xinrui Zeng, Zhenxing Li, Chiyuan Ma
International Immunopharmacology.2024; 137: 112449. CrossRef
Single-cell analysis and machine learning identify psoriasis-associated CD8+ T cells serve as biomarker for psoriasis
Sijia He, Lyuye Liu, Xiaoyan Long, Man Ge, Menghan Cai, Junling Zhang
Frontiers in Genetics.2024;[Epub] CrossRef
DcR3-associated risk score: correlating better prognosis and enhanced predictive power in colorectal cancer
Ying Duan, Hangrong Fang, Juanhong Wang, Banlai Ruan, Juan Yang, Jie Liu, Siqi Gou, Yijie Li, Zhengyi Cheng
Discover Oncology.2024;[Epub] CrossRef
Lymph node metastasis detection using artificial intelligence in T1 colorectal cancer: A comprehensive systematic review
Xiaoyan Yao, Zhiyong Zhou, Shengxun Mao, Jiaqing Cao, Huizi Li
Journal of Surgical Oncology.2024; 130(3): 637. CrossRef
Development and Validation of an Interpretable Machine Learning Model for Early Prognosis Prediction in ICU Patients with Malignant Tumors and Hyperkalemia
Zhi-Jun Bu, Nan Jiang, Ke-Cheng Li, Zhi-Lin Lu, Nan Zhang, Shao-Shuai Yan, Zhi-Lin Chen, Yu-Han Hao, Yu-Huan Zhang, Run-Bing Xu, Han-Wei Chi, Zu-Yi Chen, Jian-Ping Liu, Dan Wang, Feng Xu, Zhao-Lan Liu
Medicine.2024; 103(30): e38747. CrossRef
Machine learning-based model to predict composite thromboembolic events among Chinese elderly patients with atrial fibrillation
Jiefeng Ren, Haijun Wang, Song Lai, Yi Shao, Hebin Che, Zaiyao Xue, Xinlian Qi, Sha Zhang, Jinkun Dai, Sai Wang, Kunlian Li, Wei Gan, Quanjin Si
BMC Cardiovascular Disorders.2024;[Epub] CrossRef
Identifying and validating necroptosis‐associated features to predict clinical outcome and immunotherapy response in patients with glioblastoma
Qinghua Yuan, Weida Gao, Mian Guo, Bo Liu
Environmental Toxicology.2024; 39(10): 4729. CrossRef
CuPCA: a web server for pan-cancer association analysis of large-scale cuproptosis-related genes
Yishu Xu, Zhenshu Ma, Yajie Wang, Long Zhang, Jiaming Ye, Yuan Chen, Zhengrong Yuan
Database.2024;[Epub] CrossRef
Transcriptomic analysis reveals oxidative stress-related signature and molecular subtypes in cholangio carcinoma
Zichao Wu
Molecular Genetics and Genomics.2024;[Epub] CrossRef
The application value of support vector machine model based on multimodal MRI in predicting IDH-1mutation and Ki-67 expression in glioma
He-Xin Liang, Zong-Ying Wang, Yao Li, An-Ning Ren, Zhi-Feng Chen, Xi-Zhen Wang, Xi-Ming Wang, Zhen-Guo Yuan
BMC Medical Imaging.2024;[Epub] CrossRef
The AC010247.2/miR-125b-5p axis triggers the malignant progression of acute myelocytic leukemia by IL-6R
Fang Xie, Jialu Xu, Lina Yan, Xia Xiao, Liang Liu
Heliyon.2024; 10(18): e37715. CrossRef
Application of machine learning for predicting lymph node metastasis in T1 colorectal cancer: a systematic review and meta-analysis
Chinock Cheong, Na Won Kim, Hye Sun Lee, Jeonghyun Kang
Langenbeck's Archives of Surgery.2024;[Epub] CrossRef
Short-term PV energy yield predictions within city neighborhoods for optimum grid management
Stefani Peratikou, Alexandros G. Charalambides
Energy and Buildings.2024; 323: 114773. CrossRef
Integrating microarray-based spatial transcriptomics and RNA-seq reveals tissue architecture in colorectal cancer
Zheng Li, Xiaojie Zhang, Chongyuan Sun, Zefeng Li, He Fei, Dongbing Zhao
Journal of Big Data.2024;[Epub] CrossRef
Construction and validation of a nomogram prediction model for the catheter-related thrombosis risk of central venous access devices in patients with cancer: a prospective machine learning study
Guiyuan Ma, Shujie Chen, Sha Peng, Nian Yao, Jiaji Hu, Letian Xu, Tingyin Chen, Jiaan Wang, Xin Huang, Jinghui Zhang
Journal of Thrombosis and Thrombolysis.2024; 58(2): 220. CrossRef
Association of coexposure to perfluoroalkyl and polyfluoroalkyl compounds and heavy metals with pregnancy loss and reproductive lifespan: The mediating role of cholesterol
Hua Fang, Dai Lin, Ziqi Zhang, Haoting Chen, Zixin Zheng, Dongdong Jiang, Wenxiang Wang
Ecotoxicology and Environmental Safety.2024; 286: 117160. CrossRef
Identification of novel diagnostic biomarkers associated with liver metastasis in colon adenocarcinoma by machine learning
Long Yang, Ye Tian, Xiaofei Cao, Jiawei Wang, Baoyang Luo
Discover Oncology.2024;[Epub] CrossRef
Identification of a disulfidptosis-related genes signature for diagnostic and immune infiltration characteristics in endometriosis
Xiangyu Chang, Jinwei Miao
Scientific Reports.2024;[Epub] CrossRef
Machine learning model based on SERPING1, C1QB, and C1QC: A novel diagnostic approach for latent tuberculosis infection
Linsheng Li, Li Zhuang, Ling Yang, Zhaoyang Ye, Ruizi Ni, Yajing An, Weiguo Zhao, Wenping Gong
iLABMED.2024; 2(4): 248. CrossRef
Development of machine learning models for predicting depressive symptoms in knee osteoarthritis patients
Dan Li, Han Lu, Junhui Wu, Hongbo Chen, Meidi Shen, Beibei Tong, Wen Zeng, Weixuan Wang, Shaomei Shang
Scientific Reports.2024;[Epub] CrossRef
Integrated analysis of single-cell, spatial and bulk RNA-sequencing identifies a cell-death signature for predicting the outcomes of head and neck cancer
Yue Pan, Lei Fei, Shihua Wang, Hua Chen, Changqing Jiang, Hong Li, Changsong Wang, Yao Yang, Qinggao Zhang, Yongwen Chen
Frontiers in Immunology.2024;[Epub] CrossRef
Screening the Best Risk Model and Susceptibility SNPs for Chronic Obstructive Pulmonary Disease (COPD) Based on Machine Learning Algorithms
Zehua Yang, Yamei Zheng, Lei Zhang, Jie Zhao, Wenya Xu, Haihong Wu, Tian Xie, Yipeng Ding
International Journal of Chronic Obstructive Pulmonary Disease.2024; Volume 19: 2397. CrossRef
Identification of biomarkers related to Escherichia coli infection for the diagnosis of gastrointestinal tumors applying machine learning methods
Tingting Ge, Wei Wang, Dandan Zhang, Xubo Le, Lumei Shi
Heliyon.2024; 10(23): e40491. CrossRef
Immune Microenvironment Alterations and Identification of Key Diagnostic Biomarkers in Chronic Kidney Disease Using Integrated Bioinformatics and Machine Learning
Jinbao Shi, Aliang Xu, Liuying Huang, Shaojie Liu, Binxuan Wu, Zuhong Zhang
Pharmacogenomics and Personalized Medicine.2024; Volume 17: 497. CrossRef
Identification of alternative lengthening of telomeres-related genes prognosis model in hepatocellular carcinoma
FanLin Zeng, YuLiang Chen, Jie Lin
BMC Cancer.2024;[Epub] CrossRef
Bioinformatics Analysis Identifies PLA2G7 as a Key Antigen-Presenting Prognostic Related Gene Promoting Hepatocellular Carcinoma through the STAT1/PD-L1 Axis
Sihang Guo, Qinhe Yang
Frontiers in Bioscience-Landmark.2024;[Epub] CrossRef
Risk of intraoperative hemorrhage during cesarean scar ectopic pregnancy surgery: development and validation of an interpretable machine learning prediction model
Xinli Chen, Huan Zhang, Dongxia Guo, Siyuan Yang, Bao Liu, Yiping Hao, Qingqing Liu, Teng Zhang, Fanrong Meng, Longyun Sun, Xinlin Jiao, Wenjing Zhang, Yanli Ban, Yugang Chi, Guowei Tao, Baoxia Cui
eClinicalMedicine.2024; 78: 102969. CrossRef
TRIM16 and PRC1 Are Involved in Pancreatic Cancer Progression and Targeted by Delphinidin
Donghua Wang, Long Lv, Jinghu Du, Kui Tian, Yu Chen, Manyu Chen
Chemical Biology & Drug Design.2024;[Epub] CrossRef
Identification of Ferroptosis‐Related Gene in Age‐Related Macular Degeneration Using Machine Learning
Meijiang Zhu, Jing Yu
Immunity, Inflammation and Disease.2024;[Epub] CrossRef
Integration of bioinformatics analysis reveals ZNF248 as a potential prognostic and immunotherapeutic biomarker for LIHC: machine learning and experimental evidence
Lifang Weng
American Journal of Cancer Research.2024; 14(11): 5230. CrossRef
Development and Validation of a Nomogram for Predicting Suicidal Ideation Among Rural Adolescents in China
Yunjiao Luo, Yuhao Wang, Yingxue Wang, Yihan Wang, Na Yan, Blen Shiferaw, Louisa Mackay, Ziyang Zhang, Caiyi Zhang, Wei Wang
Psychology Research and Behavior Management.2024; Volume 17: 4413. CrossRef
High serum mannose in colorectal cancer: a novel biomarker of lymph node metastasis and poor prognosis
Xueling Wang, Haoran Li, Xiaotian Chang, Zibin Tian
Frontiers in Oncology.2023;[Epub] CrossRef
Development of a nomogram for predicting nosocomial infections among patients after cardiac valve replacement surgery
Xue Yao, Na Li, Ranran Lu, Xujing Wang, Yujun Zhang, Shuhui Wang
Journal of Clinical Nursing.2023; 32(7-8): 1466. CrossRef
Machine learning-based warning model for chronic kidney disease in individuals over 40 years old in underprivileged areas, Shanxi Province
Wenzhu Song, Yanfeng Liu, Lixia Qiu, Jianbo Qing, Aizhong Li, Yan Zhao, Yafeng Li, Rongshan Li, Xiaoshuang Zhou
Frontiers in Medicine.2023;[Epub] CrossRef
Artificial intelligence in colorectal surgery: an AI-powered systematic review
A. Spinelli, F. M. Carrano, M. E. Laino, M. Andreozzi, G. Koleth, C. Hassan, A. Repici, M. Chand, V. Savevski, G. Pellino
Techniques in Coloproctology.2023; 27(8): 615. CrossRef
Prognostic significance of neutrophil count on in-hospital mortality in patients with acute type A aortic dissection
Weiqi Feng, Huili Li, Qiuji Wang, Chenxi Li, Jinlin Wu, Jue Yang, Ruixin Fan
Frontiers in Cardiovascular Medicine.2023;[Epub] CrossRef
Machine learning algorithms assisted identification of post-stroke depression associated biological features
Xintong Zhang, Xiangyu Wang, Shuwei Wang, Yingjie Zhang, Zeyu Wang, Qingyan Yang, Song Wang, Risheng Cao, Binbin Yu, Yu Zheng, Yini Dang
Frontiers in Neuroscience.2023;[Epub] CrossRef
Risk prediction of bronchopulmonary dysplasia in preterm infants by the nomogram model
Yang Gao, Dongyun Liu, Yingmeng Guo, Menghan Cao
Frontiers in Pediatrics.2023;[Epub] CrossRef
LncRNA model predicts liver cancer drug resistance and validate in vitro experiments
Qiushi Yin, Xiaolong Huang, Qiuxi Yang, Shibu Lin, Qifeng Song, Weiqiang Fan, Wang Li, Zhongyi Li, Lianghui Gao
Frontiers in Cell and Developmental Biology.2023;[Epub] CrossRef
Artificial intelligence–assisted treatment strategy for T1 colorectal cancer after endoscopic resection
Katsuro Ichimasa, Shin-ei Kudo, Jonathan Wei Jie Lee, Tetsuo Nemoto, Khay Guan Yeoh
Gastrointestinal Endoscopy.2023; 97(6): 1148. CrossRef
A risk prediction model for type 2 diabetes mellitus complicated with retinopathy based on machine learning and its application in health management
Hong Pan, Jijia Sun, Xin Luo, Heling Ai, Jing Zeng, Rong Shi, An Zhang
Frontiers in Medicine.2023;[Epub] CrossRef
LASSO-based machine learning algorithm to predict the incidence of diabetes in different stages
Qianying Ou, Wei Jin, Leweihua Lin, Danhong Lin, Kaining Chen, Huibiao Quan
The Aging Male.2023;[Epub] CrossRef
Prediction model based on radiomics and clinical features for preoperative lymphovascular invasion in gastric cancer patients
Ping Wang, Kaige Chen, Ying Han, Min Zhao, Nanding Abiyasi, Haiyong Peng, Shaolei Yan, Jiming Shang, Naijian Shang, Wei Meng
Future Oncology.2023; 19(23): 1613. CrossRef
Liang-Ge-San: a classic traditional Chinese medicine formula, attenuates acute inflammation via targeting GSK3β
Liling Yang, Lijun Yan, Weifu Tan, Xiangjun Zhou, Guangli Yang, Jingtao Yu, Zibin Lu, Yong Liu, Liyi Zou, Wei Li, Linzhong Yu
Frontiers in Pharmacology.2023;[Epub] CrossRef
Synergistic inhibition of NUDT21 by secretory S100A11 and exosomal miR‐487a‐5p promotes melanoma oligo‐ to poly‐metastatic progression
Bin Zeng, Yuting Chen, Hao Chen, Qiting Zhao, Zhiwei Sun, Doudou Liu, Xiaoshuang Li, Yuhan Zhang, Jianyu Wang, H. Rosie Xing
Molecular Oncology.2023; 17(12): 2743. CrossRef
Identification of immune-related lncRNA in sepsis by construction of ceRNA network and integrating bioinformatic analysis
Tianfeng Wang, Si Xu, Lei Zhang, Tianjun Yang, Xiaoqin Fan, Chunyan Zhu, Yinzhong Wang, Fei Tong, Qing Mei, Aijun Pan
BMC Genomics.2023;[Epub] CrossRef
Applying machine learning algorithms to develop a survival prediction model for lung adenocarcinoma based on genes related to fatty acid metabolism
Dan Cong, Yanan Zhao, Wenlong Zhang, Jun Li, Yuansong Bai
Frontiers in Pharmacology.2023;[Epub] CrossRef
Identification of raffinose family oligosaccharides in processed Rehmannia glutinosa Libosch using matrix‐assisted laser desorption/ionization mass spectrometry image combined with machine learning
Huizhi Li, Shishan Zhang, Yanfang Zhao, Jixiang He, Xiangfeng Chen
Rapid Communications in Mass Spectrometry.2023;[Epub] CrossRef
Development and validation of a LASSO-based prediction model for immunosuppressive medication nonadherence in kidney transplant recipients
Lei Dong, Xiao Zhu, Hongyu Zhao, Qin Zhao, Shan Liu, Jia Liu, Lina Gong
Renal Failure.2023;[Epub] CrossRef
An artificial intelligence prediction model outperforms conventional guidelines in predicting lymph node metastasis of T1 colorectal cancer
Zheng Hua Piao, Rong Ge, Lu Lu
Frontiers in Oncology.2023;[Epub] CrossRef
Silicon versus Superbug: Assessing Machine Learning’s Role in the Fight against Antimicrobial Resistance
Tallon Coxe, Rajeev K. Azad
Antibiotics.2023; 12(11): 1604. CrossRef
Predictive model for early death risk in pediatric hemophagocytic lymphohistiocytosis patients based on machine learning
Li Xiao, Yang Zhang, Ximing Xu, Ying Dou, Xianmin Guan, Yuxia Guo, Xianhao Wen, Yan Meng, Meiling Liao, Qinshi Hu, Jie Yu
Heliyon.2023; 9(11): e22202. CrossRef
Multimodal and multi-omics-based deep learning model for screening of optic neuropathy
Ye-ting Lin, Qiong Zhou, Jian Tan, Yulin Tao
Heliyon.2023; 9(12): e22244. CrossRef
Exploration and validation of key genes associated with early lymph node metastasis in thyroid carcinoma using weighted gene co-expression network analysis and machine learning
Yanyan Liu, Zhenglang Yin, Yao Wang, Haohao Chen
Frontiers in Endocrinology.2023;[Epub] CrossRef
Current problems and perspectives of pathological risk factors for lymph node metastasis in T1 colorectal cancer: Systematic review
Katsuro Ichimasa, Shin‐ei Kudo, Hideyuki Miyachi, Yuta Kouyama, Kenichi Mochizuki, Yuki Takashina, Yasuharu Maeda, Yuichi Mori, Toyoki Kudo, Yuki Miyata, Yoshika Akimoto, Yuki Kataoka, Takafumi Kubota, Tetsuo Nemoto, Fumio Ishida, Masashi Misawa
Digestive Endoscopy.2022; 34(5): 901. CrossRef
Tumor Location as a Prognostic Factor in T1 Colorectal Cancer
Katsuro Ichimasa, Shin-ei Kudo, Yuta Kouyama, Kenichi Mochizuki, Yuki Takashina, Masashi Misawa, Yuichi Mori, Takemasa Hayashi, Kunihiko Wakamura, Hideyuki Miyachi
Journal of the Anus, Rectum and Colon.2022; 6(1): 9. CrossRef
The Importance of Being “That” Colorectal pT1: A Combined Clinico-Pathological Predictive Score to Improve Nodal Risk Stratification
Alessandro Gambella, Enrico Costantino Falco, Giacomo Benazzo, Simona Osella-Abate, Rebecca Senetta, Isabella Castellano, Luca Bertero, Paola Cassoni
Frontiers in Medicine.2022;[Epub] CrossRef
A Predictive Model for Qualitative Evaluation of PG-SGA in Tumor Patients Through Machine Learning
Xiangliang Liu, Yuguang Li, Wei Ji, Kaiwen Zheng, Jin Lu, Yixin Zhao, Wenxin Zhang, Mingyang Liu, Jiuwei Cui, Wei Li
Cancer Management and Research.2022; Volume 14: 1431. CrossRef
Deep Submucosal Invasion Is Not an Independent Risk Factor for Lymph Node Metastasis in T1 Colorectal Cancer: A Meta-Analysis
Liselotte W. Zwager, Barbara A.J. Bastiaansen, Nahid S.M. Montazeri, Roel Hompes, Valeria Barresi, Katsuro Ichimasa, Hiroshi Kawachi, Isidro Machado, Tadahiko Masaki, Weiqi Sheng, Shinji Tanaka, Kazutomo Togashi, Chihiro Yasue, Paul Fockens, Leon M.G. Moo
Gastroenterology.2022; 163(1): 174. CrossRef
LASSO ‐derived nomogram predicting new‐onset diabetes mellitus in patients with kidney disease receiving immunosuppressive drugs
Lina Shao, Chuxuan Luo, Chaoyun Yuan, Xiaolan Ye, Yuqun Zeng, Yan Ren, Binxian Ye, Yiwen Li, Juan Jin, Qiang He, Xiaogang Shen
Journal of Clinical Pharmacy and Therapeutics.2022; 47(10): 1627. CrossRef
Using random forest algorithm for glomerular and tubular injury diagnosis
Wenzhu Song, Xiaoshuang Zhou, Qi Duan, Qian Wang, Yaheng Li, Aizhong Li, Wenjing Zhou, Lin Sun, Lixia Qiu, Rongshan Li, Yafeng Li
Frontiers in Medicine.2022;[Epub] CrossRef
Identification of diagnostic signatures associated with immune infiltration in Alzheimer’s disease by integrating bioinformatic analysis and machine-learning strategies
Yu Tian, Yaoheng Lu, Yuze Cao, Chun Dang, Na Wang, Kuo Tian, Qiqi Luo, Erliang Guo, Shanshun Luo, Lihua Wang, Qian Li
Frontiers in Aging Neuroscience.2022;[Epub] CrossRef
Development and validation of a predictive model combining clinical, radiomics, and deep transfer learning features for lymph node metastasis in early gastric cancer
Qingwen Zeng, Hong Li, Yanyan Zhu, Zongfeng Feng, Xufeng Shu, Ahao Wu, Lianghua Luo, Yi Cao, Yi Tu, Jianbo Xiong, Fuqing Zhou, Zhengrong Li
Frontiers in Medicine.2022;[Epub] CrossRef
Identification of Drug-Induced Liver Injury Biomarkers from Multiple Microarrays Based on Machine Learning and Bioinformatics Analysis
Kaiyue Wang, Lin Zhang, Lixia Li, Yi Wang, Xinqin Zhong, Chunyu Hou, Yuqi Zhang, Congying Sun, Qian Zhou, Xiaoying Wang
International Journal of Molecular Sciences.2022; 23(19): 11945. CrossRef
Development and validation of a machine learning model to predict the risk of lymph node metastasis in renal carcinoma
Xiaowei Feng, Tao Hong, Wencai Liu, Chan Xu, Wanying Li, Bing Yang, Yang Song, Ting Li, Wenle Li, Hui Zhou, Chengliang Yin
Frontiers in Endocrinology.2022;[Epub] CrossRef
Preoperative prediction of lymph node status in patients with colorectal cancer. Developing a predictive model using machine learning
Morten Hartwig, Karoline Bendix Bräuner, Rasmus Vogelsang, Ismail Gögenur
International Journal of Colorectal Disease.2022; 37(12): 2517. CrossRef
A retrospective analysis based on multiple machine learning models to predict lymph node metastasis in early gastric cancer
Tao Yang, Javier Martinez-Useros, JingWen Liu, Isaias Alarcón, Chao Li, WeiYao Li, Yuanxun Xiao, Xiang Ji, YanDong Zhao, Lei Wang, Salvador Morales-Conde, Zuli Yang
Frontiers in Oncology.2022;[Epub] CrossRef
Challenges and opportunities in the application of artificial intelligence in gastroenterology and hepatology
Chrysanthos D Christou, Georgios Tsoulfas
World Journal of Gastroenterology.2021; 27(37): 6191. CrossRef

PubReader
ePub Link

Cite

CITE: export Copy Download Format; Close

Download Citation

Download a citation file in RIS format that can be imported by all major citation management software, including EndNote, ProCite, RefWorks, and Reference Manager.

Format:

RIS — For EndNote, ProCite, RefWorks, and most other reference management software
BibTeX — For JabRef, BibDesk, and other BibTeX-specific software

Include:

Citation for the content below
Citation and abstract for the content below

LASSO-Based Machine Learning Algorithm for Prediction of Lymph Node Metastasis in T1 Colorectal Cancer

Cancer Res Treat. 2021;53(3):773-783. Published online December 29, 2020

DOI: https://doi.org/10.4143/crt.2020.974

XML Download

Abstract
Introduction
Materials and Methods
Results
Discussion
Electronic Supplementary Material
NOTES
REFERENCES

LASSO-Based Machine Learning Algorithm for Prediction of Lymph Node Metastasis in T1 Colorectal Cancer

Fig. 1 Selection of significant parameters in clinicopathologic variables in the training set and definition of linear predictor. (A) Ten time cross-validation for tuning parameter selection in the LASSO model. (B) LASSO coefficient profiles. The LASSO was used for regression of high dimensional predictors. The method uses an L1 penalty to shrink some regression coefficients to exactly zero. The binomial deviance curve was plotted versus log (λ), where λ is the tuning parameter (A). LASSO coefficient profiles of clinicopathologic variables (B). LASSO, least absolute shrinkage and selection operator.

Fig. 2 Comparison of AUROC between LASSO model in the training and validation sets and Japanese criteria in the validation set. AUC, area under the curve; AUROC, area under the receiver operating characteristic; CI, confidence interval; LASSO, least absolute shrinkage and selection operator.

Fig. 3 Decision curve analysis of Japanese criteria and LASSO model in the training (A) and validation (B) set. The y-axis measures the net benefit. The green line represents the LASSO model. The red line represents the Japanese criteria. The gray line represents the assumption that all patients underwent surgeries. The black line represents the assumption that patients underwent no surgeries. The net benefit was calculated by subtracting the proportion of all patients who are false positive from the proportion who are true positive, weighting by the relative harm of forgoing treatment compared with the negative consequences of an unnecessary treatment. The decision curve showed that if the threshold probability of a patient or doctor is >10%, using the LASSO model in the current study to predict LNM adds more benefit than the treat-all-patients scheme or the treat-none scheme. For example, if the personal threshold probability of a patient is 20% (i.e., the patient would opt for surgery if his/her probability of LNM was > 20%), then the net benefit is 0.35 when using the LASSO model to make the decision of whether to undergo surgery, with added benefit than the treat-all scheme or the treat-none scheme. This decision curve analysis showed that the net benefit was comparable on the basis of the Japanese criteria and the treat-all or treat-none strategies. LASSO, least absolute shrinkage and selection operator; LNM, lymph node metastasis.

Fig. 1

Fig. 2

Fig. 3

LASSO-Based Machine Learning Algorithm for Prediction of Lymph Node Metastasis in T1 Colorectal Cancer

Table 1

Comparison of clinicopathological characteristics between the training and validation set

	Training set (n=221)	Validation set (n=95)	p-value
Sex
Male	123 (55.7)	59 (62.1)	0.347
Female	98 (44.3)	36 (37.9)
Age (yr)
< 70	177 (80.1)	77 (81.1)	0.966
≥ 70	44 (19.9)	18 (18.9)
Preoperative CEA (ng/mL)
< 5.0	208 (94.1)	73 (76.8)	< 0.001
≥ 5.0	13 (5.9)	22 (23.2)
Tumor location
Colon	129 (58.4)	60 (63.2)	0.502
Rectum	92 (41.6)	35 (36.8)
Retrieved lymph node numbers	12 (7–17)	14 (12–21)	< 0.001
Depth of invasion (μm)
< 1,000	20 (9.0)	6 (6.3)	0.557
≥ 1,000	201 (91.0)	89 (93.7)
Histologic grade^a)
G1	93 (42.1)	28 (29.5)	0.047
G2, G3, etc.	128 (57.9)	67 (70.5)
LVI
Present	27 (12.2)	14 (14.7)	0.668
Absent	194 (87.8)	81 (85.3)
Gross morphology
Pedunculated	17 (7.7)	16 (16.8)	0.025
Flat and sessile	204 (92.3)	79 (83.2)
Tumor budding^b)
Low grade	88 (39.8)	38 (40.0)	> 0.99
High grade	133 (60.2)	57 (60.0)
Background adenoma
Present	196 (88.7)	84 (88.4)	> 0.99
Absent	25 (11.3)	11 (11.6)
LNM
Negative	192 (86.9)	83 (87.4)	> 0.99
Positive	29 (13.1)	12 (12.6)
MMR
MMR-deficient	14 (6.3)	8 (8.4)	0.669
MMR-proficient	207 (93.7)	87 (91.6)

Values are presented as number (%) or median (IQR). CEA, carcinoembryonic antigen; IQR, interquartile range; LNM, lymph node metastasis; LVI, lymphovascular invasion; MMR: mismatch repair.

^a)Histologic grade: G1, well differentiated; G2, moderately differentiated; G3, poorly differentiated,

^b)Tumor budding grade was as follows: grade 1, 0–4/grade 2, 5–9/grade 3, 10 or more. Grade 1 was defined as low grade and grade 2/3 as high grade.

Table 2

Univariable analysis for predicting lymph node metastasis in the training set (n=221)

	LN positive, n (%)	Univariable analysis
	LN positive, n (%)	OR (95% CI)	p-value
Sex
Female	14/98 (14.3)	Reference
Male	15/123 (12.2)	0.833 (0.37–1.83)	0.647
Age (yr)
< 70	24/177 (13.6)	Reference
≥ 70	5/33 (11.4)	0.817 (0.26–2.12)	0.699
Preoperative CEA (ng/mL)
< 5.0	27/208 (13.0)	Reference
≥ 5.0	2/13 (15.4)	1.218 (0.18–4.86)	0.803
Tumor location
Colon	14/129 (10.9)	Reference
Rectum	15/92 (16.3)	1.6 (0.72–3.53)	0.239
Depth of invasion (μm)
< 1,000	3/20 (15.0)	Reference
≥ 1,000	26/201 (12.9)	0.841 (0.25–3.78)	0.794
Histologic grade^a)
G1	7/93 (7.5)	Reference
G2, G3, etc.	22/128 (17.2)	2.549 (1.08–6.70)	0.040
LVI
Absent	20/194 (10.3)	Reference
Present	9/27 (33.3)	4.35 (1.67–10.84)	0.001
Gross morphology
Pedunculated	2/17 (11.8)	Reference
Flat and sessile	27/204 (13.2)	1.144 (0.29–7.51)	0.863
Tumor budding^b)
Low grade	10/133 (7.5)	Reference
High grade	19/88 (21.6)	3.387 (1.52–7.97)	0.003
Background adenoma
Present	22/196 (11.2)	Reference
Absent	7/25 (20.0)	3.075 (1.09–7.96)	0.024
MMR
MMR-deficient	1/14 (7.1)	Reference
MMR-proficient	28/207 (13.5)	2.033 (0.38–37.65)	0.502
CD3_IM
Low (< 11.3)	10/55 (18.2)	Reference
High (≥ 11.3)	19/166 (11.4)	0.581 (0.25–1.38)	0.203
CD3_TC
Low (< 9.2)	2/39 (5.1)	Reference
High (≥ 9.2)	27/182 (14.8)	3.222 (0.90–20.53)	0.121
CD8_IM
Low (< 28.2)	29/199 (14.6)	Reference
High (≥ 28.2)	0/27 (0)	0 (NA-1.68e+26)	0.990
CD8_TC
Low (< 18.5)	24/133 (18.0)	Reference
High (≥ 18.5)	5/88 (5.7)	0.273 (0.08–0.69)	0.011
FOXP3_IM
Low (< 15.1)	16/99 (16.2)	Reference
High (≥ 15.1)	13/122 (10.7)	0.618 (0.27–1.35)	0.230
FOXP3_TC
Low (< 20.3)	23/157 (14.6)	Reference
High (≥ 20.3)	6/64 (9.4)	0.602 (0.21–1.47)	0.290

CEA, carcinoembryonic antigen; CI, confidence interval; IM, invasive margin; LN, lymph node; LVI, lymphovascular invasion; MMR, mismatch repair; NA, not available; OR, odds ratio; TC, tumor center.

^a)Histologic grade: G1, well differentiated; G2, moderately differentiated; G3, poorly differentiated,

^b)Tumor budding grade was as follows: grade 1, 0–4/grade 2, 5–9/grade 3, 10 or more. Grade 1 was defined as low grade and grade 2/3 as high grade.

Table 3

NRI and IDI in the training set and the validation set

	Training set (n=221)		Validation set (n=95)

	Japanese criteria vs. LASSO model	p-value	Japanese criteria vs. LASSO model	p-value
NRI (95% CI)^a)	0.722 (0.402–1.041)	< 0.001	0.447 (0.041–0.854)	0.039

IDI (95% CI)	0.187 (0.100–0.274)	< 0.001	0.121 (0.008–0.234)	0.034

CI, confidence interval; LASSO, least absolute shrinkage and selection operator; NRI, net reclassification improvement.

^a)Cutoff of “0, 0.1, 0.2, 1” was used in this analysis.

Table 1 Comparison of clinicopathological characteristics between the training and validation set

Values are presented as number (%) or median (IQR). CEA, carcinoembryonic antigen; IQR, interquartile range; LNM, lymph node metastasis; LVI, lymphovascular invasion; MMR: mismatch repair.

Histologic grade: G1, well differentiated; G2, moderately differentiated; G3, poorly differentiated,

Tumor budding grade was as follows: grade 1, 0–4/grade 2, 5–9/grade 3, 10 or more. Grade 1 was defined as low grade and grade 2/3 as high grade.

Table 2 Univariable analysis for predicting lymph node metastasis in the training set (n=221)

CEA, carcinoembryonic antigen; CI, confidence interval; IM, invasive margin; LN, lymph node; LVI, lymphovascular invasion; MMR, mismatch repair; NA, not available; OR, odds ratio; TC, tumor center.

Histologic grade: G1, well differentiated; G2, moderately differentiated; G3, poorly differentiated,

Tumor budding grade was as follows: grade 1, 0–4/grade 2, 5–9/grade 3, 10 or more. Grade 1 was defined as low grade and grade 2/3 as high grade.

Table 3 NRI and IDI in the training set and the validation set

CI, confidence interval; LASSO, least absolute shrinkage and selection operator; NRI, net reclassification improvement.

Cutoff of “0, 0.1, 0.2, 1” was used in this analysis.