Artificial intelligence in gynecologic oncology: current status and development prospects

Gynecologic oncology is a significant challenge for modern healthcare, requiring innovative approaches to diagnosis, treatment, and outcome prediction. Artificial intelligence (AI) is revolutionizing oncology practice, opening up new opportunities to improve the care of patients with gynecologic malignancies. This review assesses the current state of AI use in gynecologic oncology based on a systematic literature review. The methodology included searching in PubMed, Web of Science, and Scopus databases for the period 2010–2024 using the key terms “artificial intelligence”, “machine learning”, and “deep learning” in combination with “cervical cancer”, “ovarian cancer”, and “endometrial cancer”. The results show that of the 71 studies, 34 focused on cervical cancer, 21 – on ovarian cancer, and 13 – on endometrial cancer, with 49 % of the studies using imaging data and 51 % using numerical data. AI demonstrates high accuracy in cervical cancer screening (up to 95 %), significantly improves outcome prediction in ovarian cancer (AUC 0.77–0.88), and shows promising results in endometrial cancer diagnosis. Radiomics-based machine learning models achieve a C-index of 0.85 for predicting disease-free survival and overall survival in cervical cancer after chemoradiation therapy. The discussion highlights the theoretical significance of AI for personalized medicine in gynecologic oncology and its practical importance for improving clinical outcomes, but also highlights the need to address validation issues, ethical considerations, and integration into clinical practice. Prospects include the development of multimodal models, the integration of more data, and the creation of explainable algorithms to enhance clinician confidence.