Fenotipado multimodal basado en inteligencia artificial para el síndrome de Ehlers-Danlos hipermóvil: manifestaciones gastroenterológicas, mucocutáneas, neuro-ORL y emergenciológicas

Vollständiger Abstract

Worum geht es in dieser Arbeit?

Introduction: Hypermobile Ehlers-Danlos syndrome (hEDS) is a heterogeneous connective tissue disorder characterized by joint hypermobility, instability, chronic pain, and multisystem manifestations, lacking a definitive genetic biomarker, which complicates early diagnosis and phenotypic stratification (Hakim, 2024; Malfait et al., 2017). Objective: To synthesize the available evidence on the application of multimodal AI-based phenotyping to identify phenotypic subgroups of hEDS with gastroenterological, mucocutaneous, neuro-ORL and emergency manifestations. Methodology: Systematic review following PRISMA 2020 guidelines (Page et al., 2021). A systematic search was conducted in PubMed, LILACS, SciELO and Cochrane for studies published between 2015 and 2025 on hEDS phenotyping using AI, clustering techniques and multimodal analysis. Observational and technological development studies were included. Quality was assessed using the adapted Newcastle-Ottawa scale (Wells et al., 2000). Results: Seven relevant studies applying computational and AI models for hEDS phenotyping were identified. Machine learning models, including logistic regression and decision trees, identified stable phenotypic subgroups (7 clusters) with variable organ system involvement, including musculoskeletal, gastrointestinal and neurological domains (Pearson et al., 2025). AI-based facial phenotyping demonstrated ability to distinguish hEDS from other subtypes (AUC ≥ 0.97) (Murdock et al., 2025). Studies reported high prevalence of gastrointestinal manifestations (up to 49% SIBO), dysautonomia, neuropathy, and emergency complications (Cureus, 2025; Zhao, 2026). Conclusion: Multimodal AI-based phenotyping represents a promising tool for early identification and phenotypic stratification of hEDS. Integration of clinical, molecular, and wearable device data enables characterization of subgroups with specific manifestations and guides multidisciplinary management (Pearson et al., 2025; Zhao, 2026).