AI-Driven Automated Tool for Abdominal CT Body Composition Analysis in Gastrointestinal Cancer Management

2025-03-10

AI Summary

Manual CT-based quantification of abdominal muscle and adipose tissue (subcutaneous/visceral) in gastrointestinal cancer patients is time-consuming, costly, and poorly scalable, hindering clinical prognostic assessment. To address this, we propose an AI-driven automated quantitative analysis tool featuring a novel multi-view localization module integrated with an interactive correction mechanism and a 2D nnUNet-based segmentation network, all accessible via an intuitive graphical user interface enabling real-time clinician intervention. The method preserves clinical interpretability and control while substantially improving efficiency and accuracy—achieving 90% localization accuracy and Dice scores of 0.967 for muscle and adipose tissue segmentation. This tool delivers a standardized, reproducible protocol for abdominal body composition quantification, overcoming key limitations of manual analysis and providing robust, imaging-derived biomarkers to support precision management of gastrointestinal cancers.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Calibration & Uncertainty Quantification
• Humans and AI: Brain-Sensing and Analysis

Application Category

• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

The incidence of gastrointestinal cancers remains significantly high, particularly in China, emphasizing the importance of accurate prognostic assessments and effective treatment strategies. Research shows a strong correlation between abdominal muscle and fat tissue composition and patient outcomes. However, existing manual methods for analyzing abdominal tissue composition are time-consuming and costly, limiting clinical research scalability. To address these challenges, we developed an AI-driven tool for automated analysis of abdominal CT scans to effectively identify and segment muscle, subcutaneous fat, and visceral fat. Our tool integrates a multi-view localization model and a high-precision 2D nnUNet-based segmentation model, demonstrating a localization accuracy of 90% and a Dice Score Coefficient of 0.967 for segmentation. Furthermore, it features an interactive interface that allows clinicians to refine the segmentation results, ensuring high-quality outcomes effectively. Our tool offers a standardized method for effectively extracting critical abdominal tissues, potentially enhancing the management and treatment for gastrointestinal cancers. The code is available at https://github.com/NanXinyu/AI-Tool4Abdominal-Seg.git}{https://github.com/NanXinyu/AI-Tool4Abdominal-Seg.git.

Problem

• Automates abdominal CT scan analysis for cancer management.
• Reduces time and cost of manual tissue composition analysis.
• Improves accuracy in identifying muscle and fat tissues.

Innovation

• AI-driven automated abdominal CT analysis tool
• Multi-view localization and nnUNet-based segmentation
• Interactive interface for clinician refinement