Predicting clinical prognosis in gastric cancer using deep learning-based analysis of tissue pathomics images

Long Liu ¹, Yu Zhang ², Xiao-Hong Zhao ³

¹Department of Gastroenterology, Taizhou Hospital of Zhejiang Province affiliated to Wenzhou Medical University, Linhai, Zhejiang, China; Department of Hepatobiliary and Pancreatic Surgery, The Second Affiliated Hospital of Zhejiang University School of Medcine, Hangzhou, Zhejiang, China.
²Department of Radiation Oncology, The First Affiliated Hospital of Anhui Medical University, Hefei, Anhui, China.
³School of Pharmacy, Hangzhou Normal University, Hangzhou, Zhejiang, China.
⁴Science Research Department, Taizhou Hospital of Zhejiang province Affiliated to Wenzhou Medical University, Linhai, Zhejiang, China.
⁵Department of Thoracic Surgery, Taizhou Hospital of Zhejiang Province affiliated to Wenzhou Medical University, Linhai, 317000, China.
⁶Department of Gastroenterology, The Second Affiliated Hospital of Baotou Medical College, Inner Mongolia University of Science and Technology, Baotou, Inner Mongolia, China.
⁷Department of Gastroenterology, Taizhou Hospital of Zhejiang Province affiliated to Wenzhou Medical University, Linhai, Zhejiang, China.
⁸Department of Pathology, Taizhou Hospital of Zhejiang Province, Linhai, Zhejiang, China.
⁹Department of Gastroenterology, Taizhou Hospital of Zhejiang Province affiliated to Wenzhou Medical University, Linhai, Zhejiang, China; Key Laboratory of Minimally Invasive Techniques & Rapid Rehabilitation of Digestive System Tumor of Zhejiang Province, Taizhou Hospital of Zhejiang Province affiliated to Wenzhou Medical University, Linhai, Zhejiang Province, China; Institute of Digestive Disease, Taizhou Hospital of Zhejiang Province affiliated with Wenzhou Medical University, Linhai, Zhejiang, China.

⁰Department of Gastroenterology, Taizhou Hospital of Zhejiang Province affiliated to Wenzhou Medical University, Linhai, Zhejiang, China; Department of Hepatobiliary and Pancreatic Surgery, The Second Affiliated Hospital of Zhejiang University School of Medcine, Hangzhou, Zhejiang, China.

Computer Methods and Programs in Biomedicine +

|

June 13, 2025

View abstract on PubMed

Summary

A machine learning pathomics model accurately predicts overall survival (OS) in gastric cancer patients after surgery. Integrating this model with clinical data improves prognostic accuracy for personalized treatment decisions.

Area Of Science

Oncology
Medical Imaging
Artificial Intelligence

Background

Gastric cancer prognosis remains challenging, necessitating advanced predictive tools.
Accurate prediction of overall survival (OS) is crucial for guiding treatment strategies in gastric cancer patients.

Purpose Of The Study

To evaluate a machine learning-based pathomics model for predicting OS in gastric cancer patients post-surgery.
To assess the added value of integrating pathomics with clinical parameters for enhanced prognostic accuracy.

Main Methods

Retrospective analysis of 160 gastric cancer patients undergoing radical surgery.
Development of an optimal pathomics model using six machine learning methods and calculation of a pathomics score (Pathscore).
Construction of a nomogram integrating Pathscore with clinical factors (age, M stage, TNM stage) for OS prediction; validation using TCGA and GEO databases.

Main Results

The GBM-based pathomics model showed high predictive performance (1-, 3-, 5-year AUCs: 0.837, 0.970, 0.963).
The nomogram incorporating Pathscore and clinical data achieved superior OS prediction (1-, 3-, 5-year AUCs: 0.954, 0.939, 0.898).
Bioinformatics analysis indicated the model reflects tumor immune status and NRP1 expression.

Conclusions

Machine learning pathomics models are effective for predicting OS in gastric cancer patients.
Integrating pathomics with clinical parameters significantly enhances prognostic prediction accuracy.
This approach offers a reliable foundation for personalized treatment decisions in gastric cancer management.

Keywords:

Gastric cancer Immune level Machine learning Pathomics Survival

Related Concept Videos

Cancer Survival Analysis

Cancer survival analysis focuses on quantifying and interpreting the time from a key starting point, such as diagnosis or the initiation of treatment, to a specific endpoint, such as remission or death. This analysis provides critical insights into treatment effectiveness and factors that influence patient outcomes, helping to shape clinical decisions and guide prognostic evaluations. A cornerstone of oncology research, survival analysis tackles the challenges of skewed, non-normally...