Journal of ovarian research · 2026
Artificial intelligence in ovarian pathophysiology and management: a systematic review and meta-analysis.
СтатьяСистематический обзорМетаанализНазвание и аннотация: Перевод пока не подготовлен
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Yu Hang, Peng Lifan
Аннотация
BACKGROUND: The integration of artificial intelligence (AI) into reproductive medicine and gynecologic oncology has driven transformative advances in the diagnosis, prognosis, and management of ovarian conditions, including malignancies, endocrine disorders, and functional impairments.
OBJECTIVE: This PRISMA 2020-compliant systematic review and meta-analysis synthesizes evidence from 81 peer-reviewed studies to evaluate the efficacy, clinical applications, and limitations of AI and machine learning (ML) models across the full spectrum of ovarian health and disease.
METHODS: A systematic literature search was conducted across PubMed/MEDLINE, Embase, Web of Science, Cochrane Library, and Scopus for studies published between January 2015 and January 2026. Eligible studies included the development, validation, or clinical application of AI/ML models for ovarian pathophysiology or management. Study quality was appraised using the QUADAS-2 tool for diagnostic accuracy studies and the PROBAST tool for predictive modeling studies. Narrative synthesis and random-effects meta-analysis were performed, with pre-specified subgroup analyses, meta-regression, and sensitivity analyses to explore heterogeneity. Applications were categorized into ovarian cancer (diagnosis, staging, prognosis, treatment guidance), reproductive endocrinology (ovarian reserve, Polycystic Ovary Syndrome (PCOS), ovarian stimulation), and fundamental ovarian biology.
RESULTS: AI models demonstrated high diagnostic accuracy for ovarian cancer using multimodal data, including ultrasound radiomics (pooled sensitivity 89-94%, specificity 85-91%, AUC 0.92) and integrated serum biomarkers (AUC 0.94). Explainable AI (XAI) platforms effectively predicted complete surgical cytoreduction in advanced ovarian cancer (pooled AUC 0.87). In reproductive medicine, AI algorithms optimized ovarian stimulation protocols and predicted follicular growth, with reliable performance for forecasting ovarian response in IVF (pooled AUC 0.81). Substantial heterogeneity (I² 68-85% across primary analyses) was identified, driven predominantly by retrospective study design, variable AI architectures, and lack of standardized validation. Only 22% of included studies reported prospective, multicenter external validation.
CONCLUSION: AI holds significant promise for personalizing and improving care for ovarian conditions. Translation from research to routine clinical practice requires rigorous prospective multicenter validation, standardized methodological and reporting frameworks, seamless clinical workflow integration, and robust ethical and regulatory governance to ensure responsible, equitable implementation.
TRIAL REGISTRATION: PROSPERO Registration: CRD42026348721 https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42026348721.
OBJECTIVE: This PRISMA 2020-compliant systematic review and meta-analysis synthesizes evidence from 81 peer-reviewed studies to evaluate the efficacy, clinical applications, and limitations of AI and machine learning (ML) models across the full spectrum of ovarian health and disease.
METHODS: A systematic literature search was conducted across PubMed/MEDLINE, Embase, Web of Science, Cochrane Library, and Scopus for studies published between January 2015 and January 2026. Eligible studies included the development, validation, or clinical application of AI/ML models for ovarian pathophysiology or management. Study quality was appraised using the QUADAS-2 tool for diagnostic accuracy studies and the PROBAST tool for predictive modeling studies. Narrative synthesis and random-effects meta-analysis were performed, with pre-specified subgroup analyses, meta-regression, and sensitivity analyses to explore heterogeneity. Applications were categorized into ovarian cancer (diagnosis, staging, prognosis, treatment guidance), reproductive endocrinology (ovarian reserve, Polycystic Ovary Syndrome (PCOS), ovarian stimulation), and fundamental ovarian biology.
RESULTS: AI models demonstrated high diagnostic accuracy for ovarian cancer using multimodal data, including ultrasound radiomics (pooled sensitivity 89-94%, specificity 85-91%, AUC 0.92) and integrated serum biomarkers (AUC 0.94). Explainable AI (XAI) platforms effectively predicted complete surgical cytoreduction in advanced ovarian cancer (pooled AUC 0.87). In reproductive medicine, AI algorithms optimized ovarian stimulation protocols and predicted follicular growth, with reliable performance for forecasting ovarian response in IVF (pooled AUC 0.81). Substantial heterogeneity (I² 68-85% across primary analyses) was identified, driven predominantly by retrospective study design, variable AI architectures, and lack of standardized validation. Only 22% of included studies reported prospective, multicenter external validation.
CONCLUSION: AI holds significant promise for personalizing and improving care for ovarian conditions. Translation from research to routine clinical practice requires rigorous prospective multicenter validation, standardized methodological and reporting frameworks, seamless clinical workflow integration, and robust ethical and regulatory governance to ensure responsible, equitable implementation.
TRIAL REGISTRATION: PROSPERO Registration: CRD42026348721 https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42026348721.