Related Experiment Video
Updated: Sep 10, 2025
Author Spotlight: Advancing Alzheimer's Research – Exploring Early Detection and Multi-Omics Approaches
Published on: December 15, 2023
BioCompNet: A Deep Learning Workflow Enabling Automated Body Composition Analysis toward Precision Management of
Jianyong Wei1,2, Hongli Chen1, Lijun Yao1
1Department of Endocrinology and Metabolism, Shanghai Sixth People's Hospital Affiliated to Shanghai Jiao Tong University School of Medicine, Shanghai Diabetes Institute, Shanghai Key Laboratory of Diabetes Mellitus, Shanghai Key Clinical Center for Metabolic Disease, Shanghai 200233, China.
BioCompNet, a deep learning tool, automates body composition analysis from MRI scans. This enhances cardiometabolic risk assessment by providing fast, accurate quantification of bone, muscle, and adipose tissue.
Area of Science:
- Biomedical imaging
- Artificial intelligence in medicine
- Cardiovascular research
Background:
- Body composition (BC) is crucial for cardiometabolic risk stratification.
- Current medical image analysis for BC is time-consuming and limited.
- Automated quantification of BC components is needed for clinical applications.
Purpose of the Study:
- To develop BioCompNet, a deep learning workflow for automated body composition analysis.
- To quantify 15 biomechanically critical body composition components from dual-parametric MRI.
- To establish a scalable framework for precision cardiometabolic risk assessment.
Main Methods:
- Developed an end-to-end deep learning workflow (BioCompNet) using a hierarchical U-Net architecture.
- Integrated dual-parametric (water/fat) MRI sequences for quantification.
- Trained and validated the model on large datasets from community and tertiary hospital cohorts.
Main Results:
- BioCompNet achieved high accuracy (Dice coefficients > 0.93) in quantifying abdominal and thigh BC components.
- Demonstrated excellent interreader reliability (ICC ≥ 0.881) for all quantified features.
- Significantly reduced processing time from ~129 minutes to ~0.12 minutes per case.
Conclusions:
- BioCompNet enables rapid, accurate, and comprehensive volumetric analysis of body composition.
- The workflow offers a scalable solution for precision cardiometabolic risk assessment.
- BioCompNet supports clinical decision-making through automated body composition quantification.
