2026-04-17T11:57:28Zhttps://recercat.cat/oai/request

oai:recercat.cat:11351/130742025-10-24T10:59:24Zcom_2072_451665com_2072_378040col_2072_451666

Acromegaly facial changes analysis using last generation artificial intelligence methodology: the AcroFace system Rashwan, Hatem A. Asensio-Wandosell, Diego Martínez Momblan, Mª Antonia Marques Pamies, Montserrat Ruiz-Janer, Sabina Gil, Joan Acromegàlia Intel·ligència artificial Imatges - Anàlisi DISEASES::Musculoskeletal Diseases::Bone Diseases::Bone Diseases, Endocrine::Acromegaly PHENOMENA AND PROCESSES::Mathematical Concepts::Algorithms::Artificial Intelligence PSYCHIATRY AND PSYCHOLOGY::Psychological Phenomena::Mental Processes::Perception::Pattern Recognition, Physiological::Pattern Recognition, Visual::Facial Recognition ENFERMEDADES::enfermedades musculoesqueléticas::enfermedades óseas::enfermedades óseas endocrinas::acromegalia FENÓMENOS Y PROCESOS::conceptos matemáticos::algoritmos::inteligencia artificial PSIQUIATRÍA Y PSICOLOGÍA::fenómenos psicológicos::procesos mentales::percepción::reconocimiento fisiológico de modelos::reconocimiento visual de modelos::reconocimiento facial Acromegalia; Inteligencia artificial; Análisis facial Acromegàlia; Intel·ligència artificial; Anàlisi facial Acromegaly; Artificial intelligence; Facial analysis To describe the development of the AcroFace system, an AI-based system for early detection of acromegaly, based on facial photographs analysis. Two 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. ResNet-50 model as a feature extractor and Support Vector Regression (SVR) with a linear kernel showed the best performance (accuracy δ1 of 75% and δ3 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. AcroFace 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. Open Access Funding provided by Universitat Autonoma de Barcelona. This study was partially supported by a grant of the Aspire program by Pfizer international and by a grant from the Instituto de Salud Carlos III PMP22/00021 funded by the European Union-Next Generation EU to Manel Puig-Domingo. 2025-10-24T10:59:24Z 2025-10-24T10:59:24Z 2025-10-24T10:59:24Z 2025-05-12T10:42:02Z 2025-05-12T10:42:02Z 2025-04-21 info:eu-repo/semantics/article info:eu-repo/semantics/publishedVersion http://hdl.handle.net/11351/13074 Pituitary;28(3) https://www.doi.org/10.1007/s11102-025-01515-2 Attribution-NonCommercial 4.0 International http://creativecommons.org/licenses/by-nc/3.0/ info:eu-repo/semantics/openAccess Springer Scientia