Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

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

Xinyu Nan, Meng He, Zifan Chen, Bin Dong, Lei Tang, Li Zhang

TL;DR

GI cancers require accurate abdominal tissue prognostic metrics, but manual analysis is slow and costly. The authors present an AI-driven tool that combines a multi-view localization module with a high-precision 2D nnUNet segmentation pipeline, plus an interactive refinement interface, to automate abdominal tissue analysis on CT scans. Localization achieves around 90% accuracy and segmentation achieves a Dice score around 0.967, enabling robust measurement of muscle, subcutaneous fat, and visceral fat. This approach standardizes tissue extraction, reduces clinician workload, and supports scalable GI cancer management and treatment planning.

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.

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

TL;DR

GI cancers require accurate abdominal tissue prognostic metrics, but manual analysis is slow and costly. The authors present an AI-driven tool that combines a multi-view localization module with a high-precision 2D nnUNet segmentation pipeline, plus an interactive refinement interface, to automate abdominal tissue analysis on CT scans. Localization achieves around 90% accuracy and segmentation achieves a Dice score around 0.967, enabling robust measurement of muscle, subcutaneous fat, and visceral fat. This approach standardizes tissue extraction, reduces clinician workload, and supports scalable GI cancer management and treatment planning.

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.

Paper Structure

This paper contains 13 sections, 2 equations, 2 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Methodology
  3. Overview of the Proposed Automated Tool
  4. Step I: Abdominal Region Localization
  5. Stage II: Abdominal Tissue Segmentation and Interactive Refinement
  6. Experiments
  7. Data Acquisition and pre-processing
  8. Implementation Details
  9. Evaluation Metrics
  10. Localization and Segmentation Results Analysis
  11. Conclusion
  12. Compliance with ethical standards
  13. Acknowledgments

Figures (2)

  • Figure 1: An overview of our automatic annotation tool for abdominal components. (a) Stage I: A multi-view localization model consisting of a 3D ResNet18, a lightweight downsampling network and a multi-view fusion module to localize the start and end slice indexes of abdominal slices. (b) Stage II: A two-step process involving (i) a high-precision segmentation model based on 2D nnUNet architecture to segment abdominal muscle, SFA, and VFA of each slice, producing high-quality initial results, and (ii) an optional interactive interface for clinicians to further refine these segmentation results if needed.
  • Figure 2: Workflow of the interactive automatic annotation tool.