Acromegaly facial changes analysis using last generation artificial intelligence methodology: the AcroFace system


Por: Rashwan, HA, Marqués-Pamies, M, Ruiz, S, Gil, J, Asensio-Wandosell, D, Martinez-Momblán, MA, Vázquez, F, Salinas, I, Ciriza, R, Jordà, M, Chanson, P, Valassi, E, Abdelnasser, M, Puig, D and Puig-Domingo, M

Publicada: 1 jun 2025
Resumen:
PurposeTo describe the development of the AcroFace system, an AI-based system for early detection of acromegaly, based on facial photographs analysis.MethodsTwo types of features were explored: (1) the visual/texture of a set of 2D facial images, and (2) geometric information obtained from a reconstructed 3D model from a single image. We optimized acromegaly detection by integrating SVM for geometric features and CNNs for visual features, each chosen for their strength in processing distinct data types effectively. This combination enhances overall accuracy by leveraging SVM's capability to manage structured, quantitative data and CNNs' proficiency in interpreting complex image textures, thus providing a comprehensive analysis of both geometric alignment and textural anomalies. ResNet-50, VGG-16, MobileNet, Inception V3, DensNet121 and Xception models were trained with an expert endocrinologist-based score as a ground truth.ResultsResNet-50 model as a feature extractor and Support Vector Regression (SVR) with a linear kernel showed the best performance (accuracy delta 1 of 75% and delta 3 of 89%), followed by the VGG-16 as a feature extractor and SVR with a linear kernel. Geometric features yield less accurate results than visual ones. The validation cohort showed the following performance: precision 0.90, accuracy 0.93, F1-Score 0.92, sensitivity 0.93 and specificity 0.93.ConclusionAcroFace system shows a good performance to discriminate acromegaly and non-acromegaly facial traits that may serve for the detection of acromegaly at an early stage as a screening procedure at a population level.

Filiaciones:
Rashwan, HA:
 Univ Rovira & Virgili, Dept Comp Engn & Math, Tarragona, Spain

Marqués-Pamies, M:
 Autonomous Univ Barcelona, Germans Trias Res Inst & Hosp, Serv Endocrinol & Nutr, Barcelona, Spain

 Hosp Granollers, Endocrinol Unit, Barcelona, Spain

Ruiz, S:
 Autonomous Univ Barcelona, Germans Trias Res Inst & Hosp, Serv Endocrinol & Nutr, Barcelona, Spain

Gil, J:
 Autonomous Univ Barcelona, Germans Trias Res Inst & Hosp, Serv Endocrinol & Nutr, Barcelona, Spain

 Inst Salud Carlos III, Dept Med, Autonomous Univ CIBERER Grp 747, Madrid, Spain

Asensio-Wandosell, D:
 Autonomous Univ Barcelona, Germans Trias Res Inst & Hosp, Serv Endocrinol & Nutr, Barcelona, Spain

 Inst Salud Carlos III, Dept Med, Autonomous Univ CIBERER Grp 747, Madrid, Spain

Martinez-Momblán, MA:
 Autonomous Univ Barcelona, Germans Trias Res Inst & Hosp, Serv Endocrinol & Nutr, Barcelona, Spain

 Univ Barcelona, Med & Hlth Sci Fac, Nursing Sch, Fundamental & Med Surg Nursing Dept, Barcelona, Spain

:
 Autonomous Univ Barcelona, Germans Trias Res Inst & Hosp, Serv Endocrinol & Nutr, Barcelona, Spain

:
 Autonomous Univ Barcelona, Germans Trias Res Inst & Hosp, Serv Endocrinol & Nutr, Barcelona, Spain

Ciriza, R:
 Spanish Assoc People Acromegaly, Huesca, Spain

:
 Autonomous Univ Barcelona, Germans Trias Res Inst & Hosp, Serv Endocrinol & Nutr, Barcelona, Spain

Chanson, P:
 Univ Paris Saclay, Hop Bicetre, AP HP,Physiol & Physiopathol Endocriniennes, Ctr Reference Malad Rares Hypophyse,Inserm,Serv En, F-94275 Le Kremlin Bicetre, France

:
 Autonomous Univ Barcelona, Germans Trias Res Inst & Hosp, Serv Endocrinol & Nutr, Barcelona, Spain

 Inst Salud Carlos III, Dept Med, Autonomous Univ CIBERER Grp 747, Madrid, Spain

Abdelnasser, M:
 Univ Rovira & Virgili, Dept Comp Engn & Math, Tarragona, Spain

Puig, D:
 Univ Rovira & Virgili, Dept Comp Engn & Math, Tarragona, Spain

:
 Autonomous Univ Barcelona, Germans Trias Res Inst & Hosp, Serv Endocrinol & Nutr, Barcelona, Spain

 Inst Salud Carlos III, Dept Med, Autonomous Univ CIBERER Grp 747, Madrid, Spain
ISSN: 15737403





Pituitary
Editorial
Kluwer Academic Publishers, ONE NEW YORK PLAZA, SUITE 4600, NEW YORK, NY, UNITED STATES, Países Bajos
Tipo de documento: Article
Volumen: 28 Número: 3
Páginas:
WOS Id: 001472125600001
ID de PubMed: 40257631
imagen Green Submitted, hybrid

MÉTRICAS