Related Experiment Video
Updated: Jan 9, 2026
13:35
Segmentation and Linear Measurement for Body Composition Analysis using Slice-O-Matic and Horos
Published on: March 21, 2021
11.6K
Associations Between Deep Learning-Derived Fat, Muscle, and Bone Measures From Abdominal Computed Tomography Scans
Jennifer L St Sauver1, Brandon R Grossardt2, Alexander D Weston3
1Division of Epidemiology, Mayo Clinic, Rochester, MN.
Mayo Clinic Proceedings. Digital Health
|December 1, 2025
Summary
Low muscle density detected via abdominal CT scans is linked to a higher fall risk in adults. This association is particularly strong in middle-aged individuals, suggesting early detection potential.
Area of Science:
- Radiology
- Gerontology
- Biomedical Engineering
Background:
- Fall risk is a significant concern for adults, impacting mobility and quality of life.
- Abdominal computed tomography (CT) provides detailed body composition data, but its association with fall risk is not fully understood.
- Assessing body composition through CT may offer novel insights into fall prevention strategies.
Purpose of the Study:
- To investigate the association between abdominal CT-derived body composition measures and the risk of falls in adults aged 20-89 years.
- To identify specific body composition biomarkers that predict fall risk.
- To explore age-related differences in the association between body composition and falls.
Main Methods:
- Utilized abdominal CT scans from 3972 individuals (aged 20-89) within the Rochester Epidemiology Project (2010-2020).
- Employed a deep learning algorithm to quantify subcutaneous adipose tissue, visceral adipose tissue, skeletal muscle area and density, and vertebral bone area and density.
- Assessed fall risk using Cox proportional hazards models, adjusting for body mass index and chronic conditions, with falls identified via ICD codes and chart review.
Main Results:
- No significant association was found between fat or bone measures and fall risk (all adjusted P > .05).
- Lower skeletal muscle density was significantly associated with an increased risk of falls (adjusted HR for lowest tertile vs. middle: 2.31).
- The association between low muscle density and falls was most pronounced in the 45-64 age group (adjusted HR: 4.98).
Conclusions:
- Abdominal CT-derived muscle density is a potential biomarker for fall risk.
- Reduced muscle density may indicate underlying physiological changes contributing to falls, even in middle age.
- These findings suggest that CT-based muscle density assessment could aid in early identification of individuals at higher risk of falls.
