Application of artificial intelligence in the diagnosis and prognostic prediction of ovarian cancer

• Published in: Computers in biology and medicine Vol. 146; p. 105608
• Main Authors: Zhou, Jingyang, Cao, Weiwei, Wang, Lan, Pan, Zezheng, Fu, Ying
• Format: Journal Article
• Language: English
• Published: United States Elsevier Ltd 01.07.2022; Elsevier Limited
• Subjects: Accuracy; Algorithms; Artificial intelligence; Biomarkers; Cancer; Data collection; Datasets; Diagnosis; Internal Medicine; Iterative methods; Machine learning; Medical diagnosis; Medical prognosis; Methods; Other; Ovarian cancer; Quality assessment; Reliability analysis; Reliability aspects; Reviews; Tumors; Ultrasonic imaging; Artificial intelligence; Ovarian cancer
• ISSN: 0010-4825; 1879-0534; 1879-0534
• DOI: 10.1016/j.compbiomed.2022.105608

In recent years, the wide application of artificial intelligence (AI) has dramatically improved the work efficiency of clinicians and reduced their workload. This review provides a glance at the latest advances in AI-assisted diagnosis and prognostic prediction of ovarian cancer (OC). We performed an advanced search in PubMed and IEEE/IET Electronic Library, and included 39 articles in this review. A comprehensive and objective criterion was built to assess the reliability and quality of all studies from four aspects: the size of datasets for model development, research design, the division of training sets and test sets, and the type of quantitative performance indicators. This review analyzed the construction of AI models, including data pre-processing methods, feature selection techniques, AI classifiers, or algorithms. Additionally, we compared the performance of these models built on different datasets, which may support researchers for further iteration and development of AI. Finally, we discussed the challenges and future directions for AI application in medicine. •Advances of artificial intelligence in ovarian cancer over two decades are reviewed.•A customized scoring system is created to assess the quality of researches.•Literature search is performed in both PubMed and IEEE/IET Electronic Library.