AI Models and Genomic Biomarkers in Rare Gynecological Pathologies: From Mayer-Rokitansky-Küster-Hauser (MRKH) and Primary Ovarian Insufficiency (POI) to Sex Cord-Stromal Tumors with Meigs’ and OHVIRA Syndromes

Abstract

The rare gynecological diseases, including MRKH syndrome, POI, OHVIRA syndrome, Meigs' syndrome and SCSTs, pose great difficulties for diagnosis due to genetic complexity and clinical heterogeneity. This review highlights the latest human-based scientific evidence on the use of genomic biomarkers and AI in the diagnosis, prognosis, and treatment of the mentioned syndromes and disorders. The human studies performed through WGS, WES, next-generation sequencing, and multiomics revealed critical biomarkers associated with the development of reproductive tract disorders, ovarian function, and tumor progression, which promotes precision medicine and personalization. The review also covers the increasing usage of machine learning, deep learning, explainable AI, digital pathology, radiogenomics and AI-assisted diagnostics in analyzing genomic, imaging and histopathological data for accurate classification and diagnostics of diseases. Despite the high potential of these technologies in clinic, there are still some limitations related to the lack of large multicentric human datasets, genetic heterogeneity and the need for AI validation. Further studies should be directed at the combination of advanced genomic technologies, multiomics and explainable AI for the improvement of precision gynecology.