Related Experiment Video
Updated: Feb 16, 2026
Midface Hypoplasia and Cranial Base Morphology in Syndromic Craniosynostosis: A Comparative Analysis Study Using a Predictive Regression Model
Published on: November 4, 2025
Automatic Detection of Acromegaly From Facial Photographs Using Machine Learning Methods
Xiangyi Kong1, Shun Gong2, Lijuan Su3
1Department of Neurosurgery, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences, No. 1 Shuaifuyuan Hutong, Dongcheng District, Beijing 100730, China; Department of Breast Surgical Oncology, China National Cancer Center/Cancer Hospital, Chinese Academy of Medical Sciences, Peking Union Medical College, Chaoyangqu, Panjiayuan-Nanli 17, Beijing 100021, PR China.
Artificial intelligence (AI) can automatically detect acromegaly from facial photographs. This technology shows high accuracy, potentially improving early diagnosis and treatment outcomes for acromegaly.
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Endocrinology
Background:
- Early detection of acromegaly from facial photographs is a promising approach.
- This could reduce disease prevalence and improve cure rates.
Purpose of the Study:
- To develop and evaluate machine learning algorithms for automatic acromegaly detection from facial images.
- To assess the performance of AI in identifying acromegaly with high accuracy.
Main Methods:
- Utilized a dataset of 527 acromegaly patients and 596 controls.
- Employed facial landmark detection, frontalization, and machine learning algorithms (LM, KNN, SVM, RT, CNN, EM).
- Evaluated models on a separate dataset confirmed by growth hormone suppression tests.
Main Results:
- Achieved a positive predictive value (PPV) of 96%.
- Demonstrated a negative predictive value (NPV) of 95%.
- Reported a sensitivity and specificity of 96% for acromegaly detection.
Conclusions:
- Artificial intelligence effectively detects acromegaly from facial photographs.
- The developed AI models exhibit high sensitivity and specificity for early acromegaly diagnosis.
