Deep learning-based computed tomography (CT) derived body composition classifier for colorectal cancer patients

2026-08-16

AI Summary

This study addresses the clinical bottleneck of time-consuming and expert-dependent body composition analysis in colorectal cancer CT imaging by proposing a deep learning-based automated assessment framework. Through systematic optimization of architectures such as GoogLeNet, the method achieves precise regression prediction of skeletal muscle and adipose tissue indices, complemented by a web-based tool for seamless clinical integration. Experimental results demonstrate that the model predicts skeletal muscle area with a mean error of only 4.96% and achieves 80% classification accuracy on an independent test set. By significantly reducing manual segmentation costs and lowering the barrier to specialized expertise, this work validates the feasibility and efficiency of AI-assisted body composition analysis for routine clinical application in oncology.

Abstract

Background: Accurate body composition analysis using Computed Tomography (CT) scans is essential for assessing skeletal muscle area (SMA) and skeletal muscle density (SMD), key markers of nutritional status in cancer patients. Conventional manual methods are labour-intensive and require specialist expertise, limiting their routine clinical use. Therefore, this study serves as a feasibility and pilot investigation to explore the potential of deep learning-based automated regression for body composition analysis within a clinical workflow. Methods: Four deep learning architectures (AlexNet, UNet, GoogLeNet, and ResNet34) were trained to predict SMA, SMD, subcutaneous fat area (SFA), and visceral fat area (VFA) from CT scans of colorectal cancer patients. Systematic hyperparameter optimization identified the most accurate models, which were subsequently implemented in a web application for clinical use. Results: GoogLeNet achieved the best performance, with a mean percentage error (PE) of 4.96% for SMA prediction, while AlexNet reached 8.12% for SMD. Independent testing demonstrated robust accuracy, correctly classifying body composition metrics in 80% of cases. The web application delivered rapid and consistent outputs, supporting integration into clinical workflows. Conclusion: Optimized deep learning models, particularly GoogLeNet and AlexNet, can automate CT-derived body composition analysis with a Mean Percentage Error (PE) of 4.96% for SMA and 8.12% for SMD. These tools have the potential to streamline clinical practice by reducing the time and expertise required for manual segmentation. Further validation in larger, more diverse datasets is warranted.

Problem

• Body Composition Analysis
• Computed Tomography
• Colorectal Cancer
• Manual Segmentation
• Clinical Workflow

Innovation

• Deep Learning
• Body Composition Analysis
• Computed Tomography
• Automated Regression
• Clinical Workflow Integration