Lara Riem · bioRxiv (Cold Spring Harbor Laboratory) 2025 · cross-sectional MRI imaging study · n=48

Big Bones Mean Big Muscles: AI Quantifies 71 Individual Muscles Across the Body, Revealing Widespread Links Between Muscle, Bone, and Size

Cited 1 times in the scientific literature.

Level 4 - case-series / case-control

Cross-sectional imaging analysis and algorithm validation study (graded by design analogy)

OpenAlex W4409235283 · doi:10.1101/2025.04.01.646599 · record verified 2026-08-31

What was done

Researchers developed and validated an AI algorithm to generate 3D segmentations of 71 individual muscles and 13 bones from whole-body MRI scans. The dataset included 48 healthy adults (24 males, 24 females, aged 18–49 years). The authors evaluated muscle volume, asymmetry, length, and fat fraction, comparing volumetric relationships across whole-body, regional, and individual muscle-bone levels as well as across sexes.

What was found

Total muscle volume was the primary predictor of individual muscle volume, followed by bone volume, which significantly correlated with muscle size across whole-body, regional, and individual levels. Muscle-to-body size relationships (mass, height, BMI) differed between sexes, whereas bone-to-body size relationships did not. Muscle asymmetry and fat fraction varied between distinct muscles but showed no statistically significant differences between sexes. Specific numerical correlation coefficients and effect sizes were not provided in the abstract.

Why it matters

This study provides an automated, comprehensive mapping of whole-body muscular and skeletal volumes, demonstrating that skeletal frame size is an anatomical determinant of individual muscle size variation.

Limits

The sample size was small (n = 48) and restricted to healthy adults aged 18–49, limiting generalizability to older adults, clinical populations, or athletic extremes. The cross-sectional design prevents causal conclusions regarding musculoskeletal development or adaptation, and the abstract omitted specific quantitative metrics (correlation coefficients, p-values).

Cited by