Deep-learning analysis suggests a facial image signal associated with low muscle mass in a clinical cohort enriched for chronic liver disease: a proof-of-concept study
- 1Department of Gastroenterology and Hepatology, Division of Internal Medicine, Medical University of Graz, Graz, Austria
- 2CBmed GmbH - Center for Biomarker Research in Medicine, Graz, Austria
- 3Department of Ethics and Social Studies, Faculty of Catholic Theology, University of Graz, Graz, Austria
Abstract
Background: Sarcopenia is a clinically relevant complication of chronic liver disease. Because changes in muscle mass, adiposity, and nutritional status may affect facial morphology, this exploratory study examined whether standardized frontal facial photographs contain an algorithmically detectable signal associated with low measured muscle mass.
Methods: Prospectively collected clinical data and facial photographs from 255 participants were retrospectively analyzed using the 2010 European Working Group on Sarcopenia in Older People (EWGSOP1) framework. The binary target was low versus normal measured muscle mass. Stage 1 developed ImageNet-pretrained VGG16, MobileNetV2, and DenseNet121 models in 204 participants and evaluated them in a 51-participant internal hold-out set. Stage 2 independently redeveloped the task using four region-specific EfficientNet-B0 and DenseNet121 components integrated into a single decision pipeline and evaluated it separately within the same source cohort. A clinical-variable-only comparator, human image assessment, Grad-CAM, and controlled perturbation analyses were also performed.
Results: Overall, 173 participants had low and 82 had normal muscle mass. Stage-1 balanced accuracy was 0.71 for VGG16, 0.70 for DenseNet121, and 0.63 for MobileNetV2 The integrated stage-2 system showed high sensitivity (94.3%) but limited specificity (50.0%), resulting in a balanced accuracy of 0.72. Human image assessment showed low interrater agreement. Grad-CAM and perturbation analyses indicated contributions from facial regions but also substantial dependence on peripheral and non-face image content. The clinical-variable comparator outperformed the stage-1 facial models.
Conclusion: Standardized facial photographs contained a image-derived information associated with low measured muscle mass across two independently implemented modelling stages. The integrated stage-2 pipeline achieved high sensitivity, supporting further investigation of facial image analysis as a low-burden pre-screening approach. However, limited specificity, potential clinical and diagnostic-pathway confounding, dependence on non-face image content, and the absence of external validation indicate that these findings remain exploratory. Prospective external validation and assessment of incremental value beyond routine clinical variables are required before clinical implementation.
Related
- Bayesian re-analysis suggests participation rather than referral underlies the benefits of social prescribing for chronic diseases
- MOG Antibody Status Is Associated With Distinct Clinical Profiles and Treatment Responses in Chronic Relapsing Inflammatory Optic Neuropathy
- Reinforcement Learning for Chronic Care Pathway Optimization: A Unified Framework across Three Clinical Goal Types
- Multi-omics profiling suggests deep brain stimulation-associated expansion of CLIC3+ NK cells attenuates neuroinflammation in Parkinson's disease
- Clinical factors associated with early multidomain symptomatic recovery in hospitalized patients with chronic lung disease and post-COVID conditions: a secondary analysis of a prospective controlled cohort
BibTeX
@article{fuchs2026deeplearning,
title = {{Deep-learning analysis suggests a facial image signal associated with low muscle mass in a clinical cohort enriched for chronic liver disease: a proof-of-concept study}},
author = {Thomas Fuchs and Akhila Abdulnazar and Lavra Rebol and Peter Fickert and Thomas Gremsl and Vanessa Stadlbauer},
journal = {Frontiers in Cell and Developmental Biology},
year = {2026},
month = oct,
doi = {10.3389/fcell.2026.1964880},
url = {https://doi.org/10.3389/fcell.2026.1964880},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Deep-learning analysis suggests a facial image signal associated with low muscle mass in a clinical cohort enriched for chronic liver disease: a proof-of-concept study
AU - Thomas Fuchs
AU - Akhila Abdulnazar
AU - Lavra Rebol
AU - Peter Fickert
AU - Thomas Gremsl
AU - Vanessa Stadlbauer
PY - 2026
DA - 2026/10/12
JO - Frontiers in Cell and Developmental Biology
DO - 10.3389/fcell.2026.1964880
UR - https://doi.org/10.3389/fcell.2026.1964880
AB - Background: Sarcopenia is a clinically relevant complication of chronic liver disease. Because changes in muscle mass, adiposity, and nutritional status may affect facial morphology, this exploratory study examined whether standardized frontal facial photographs contain an algorithmically detectable signal associated with low measured muscle mass. Methods: Prospectively collected clinical data and facial photographs from 255 participants were retrospectively analyzed using the 2010 European Working Group on Sarcopenia in Older People (EWGSOP1) framework. The binary target was low versus normal measured muscle mass. Stage 1 developed ImageNet-pretrained VGG16, MobileNetV2, and DenseNet121 models in 204 participants and evaluated them in a 51-participant internal hold-out set. Stage 2 independently redeveloped the task using four region-specific EfficientNet-B0 and DenseNet121 components integrated into a single decision pipeline and evaluated it separately within the same source cohort. A clinical-variable-only comparator, human image assessment, Grad-CAM, and controlled perturbation analyses were also performed. Results: Overall, 173 participants had low and 82 had normal muscle mass. Stage-1 balanced accuracy was 0.71 for VGG16, 0.70 for DenseNet121, and 0.63 for MobileNetV2 The integrated stage-2 system showed high sensitivity (94.3%) but limited specificity (50.0%), resulting in a balanced accuracy of 0.72. Human image assessment showed low interrater agreement. Grad-CAM and perturbation analyses indicated contributions from facial regions but also substantial dependence on peripheral and non-face image content. The clinical-variable comparator outperformed the stage-1 facial models. Conclusion: Standardized facial photographs contained a image-derived information associated with low measured muscle mass across two independently implemented modelling stages. The integrated stage-2 pipeline achieved high sensitivity, supporting further investigation of facial image analysis as a low-burden pre-screening approach. However, limited specificity, potential clinical and diagnostic-pathway confounding, dependence on non-face image content, and the absence of external validation indicate that these findings remain exploratory. Prospective external validation and assessment of incremental value beyond routine clinical variables are required before clinical implementation.
SN - 2296-634X
ER -