Fully-automated segmentation of muscle and inter-/intra-muscular fat from magnetic resonance images of calves and thighs: an open-source workflow in Python
Bibliographic record
Abstract
Abstract Background INTER- and INTRAmuscular fat (IMF) is elevated in high metabolic states and can promote inflammation. While magnetic resonance imaging (MRI) excels in depicting IMF, the lack of reproducible tools prevents the ability to measure change and track intervention success. Methods We detail an open-source fully-automated iterative threshold-seeking algorithm (ITSA) for segmenting IMF from T1-weighted MRI of the calf and thigh within three cohorts (CaMos Hamilton (N = 54), AMBERS (N = 280), OAI (N = 105)) selecting adults 45–85 years of age. Within the CaMos Hamilton cohort, same-day and 1-year repeated images (N = 38) were used to evaluate short- and long-term precision error with root mean square coefficients of variation; and to validate against semi-automated segmentation methods using linear regression. The effect of algorithmic improvements to fat ascertainment using 3D connectivity and partial volume correction rules on analytical precision was investigated. Robustness and versatility of the algorithm was demonstrated by application to different MR sequences/magnetic strength and to calf versus thigh scans. Results Among 439 adults (319 female(89%), age: 71.6 ± 7.6 yrs, BMI: 28.06 ± 4.87 kg/m2, IMF%: 10.91 ± 4.57%), fully-automated ITSA performed well across MR sequences and anatomies from three cohorts. Applying both 3D connectivity and partial volume fat correction improved precision from 4.99% to 2.21% test–retest error. Validation against semi-automated methods showed R2 from 0.92 to 0.98 with fully-automated ITSA routinely yielding more conservative computations of IMF volumes. Quality control shows 7% of cases requiring manual correction, primarily due to IMF merging with subcutaneous fat. A full workflow described methods to export tags for manual correction. Conclusions The greatest challenge in segmenting IMF from MRI is in selecting a dynamic threshold that consistently performs across repeated imaging. Fully-automated ITSA achieved this, demonstrated low short- and long-term precision error, conducive of use within RCTs.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.003 | 0.006 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.001 |
| Open science | 0.003 | 0.003 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.015 | 0.009 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".