Integrative multi-omics and clinical data with explainable AI: a deep learning framework for enhanced early detection of polycystic ovary syndrome(PCOS)

Abstract

Polycystic Ovary Syndrome (PCOS) is a complex endocrine disorder affecting 8-13% of reproductive-aged women, presenting varied symptoms including irregular cycles, hyperandrogenism, and potential infertility. Early detection remains challenging due to heterogeneous presentation and overlapping symptoms with normal puberty. In this study, we address enhanced early detection of PCOS through multi-omics and clinical data fusion with a custom domain adaptation strategy. Our approach is motivated by the challenges of integrating heterogeneous datasets ranging from transcriptomics to clinical into a unified predictive model.