Facial Analysis in Acromegaly Using Machine Learning: Toward Earlier Diagnosis
Authors' conclusion
Our findings demonstrate that acromegaly-related facial features can be detected years before clinical diagnosis using machine learning. By combining high-performance deep models with interpretable approaches, this study supports the potential for artificial intelligence-based facial screening tools to aid in early detection of acromegaly.
Study context
- Population
- Facial photographs of patients with acromegaly (prediagnosis, postdiagnosis and combined images) and controls92 patients · 88 controls
- Comparison
- VGG-Face deep features + support vector machine (SVM) vs interpretable landmark-based facial-ratio model for identifying acromegaly
- Outcome
- Classification of acromegaly vs control facial images
- Follow-up
- Prediagnosis images (mean 7.47 years before diagnosis)
Methodological detailsmachine learning, 2-stage pipeline, 489 photographs from 92 patients with acromegaly and 254 images from 88 controls analysed
- Machine learning
- 2-stage pipeline
- 489 photographs from 92 patients with acromegaly and 254 images from 88 controls analysed
Performance
As reported in the abstract
| Test or model | AUC | Accuracy |
|---|---|---|
| Prediagnosis data set | 0.982 | 91.5% |
Prediagnosis data set
AUC0.982
Accuracy91.5%
- Interpretable model using facial ratios: moderate accuracy (AUC)
- 0.776
Original publication
Facial Analysis in Acromegaly Using Machine Learning: Toward Earlier Diagnosis. J Clin Endocrinol Metab. Published 21 August 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →