Pre-Screening for Acromegaly using Consumer 3D Structured-Light Facial Scans

Background : Acromegaly is a rare yet progressive endocrine disorder often diagnosed late despite accurate biochemical tests. Standard diagnostics are only initiated after clinical suspicion, which causes delays. Objective : To describe a privacy-preserving, on-device pre-screening system using 3D structured-light (SL) facial scans from consumer smartphones and to propose a high-specificity, prospective study in risk-enriched cohorts. Methods : We reviewed acromegaly epidemiology, the capabilities of Apple’s ‘’TrueDepth” near-infrared (NIR) SL sensor ( Apple Inc., Cupertino, CA, USA ), and the Apple Augmented Reality Developer Kit (ARKit) with an anonymized 3D face mesh. We detail an on-device machine learning (ML) pipeline and a two-stage prospective study. The primary endpoint is specificity ≥ 99.5% compared to insulin-like growth factor-I (IGF-I) > 1.3 × upper limit of normal (ULN). Sample size calculations estimate 8,000-10,000 negatives and approximately 140 positives. Voice analysis is included only as exploratory. Results and discussion : Prior studies show facial geometry supports acromegaly detection (sensitivity/specificity approximately 0.93). TrueDepth offers approximately 1 mm precision under controlled conditions. In sleep apnea clinics (prevalence ≥ 1.35/1,000), a specificity ≥ 99.5% yields workable positive predictive values (approximately 35%). Conclusion : Consumer SL 3D scans are technically and ethically suitable for pre-screening in high-risk groups. A prospective, multi-site validation study powered for high specificity is recommended.