Histopathological Analysis of Ovarian Cancer Using Deep Learning

Sulaiman Khan, Fathima Farha, Manahil Zulfiqar, Md. Rafiul Biswas, Zubair Shah

Abstract: Ovarian cancer is a leading gynecological malignancy with high mortality rates, necessitating accurate prognostic tools and personalized therapeutic strategies to improve patient outcomes. Deep learning techniques have shown great promise in revolutionizing the diagnosis and management of ovarian cancer, particularly in histopathology. This study synthesizes and analyzes the existing literature on the role of deep learning in histopathology analysis of ovarian cancer. This systematic research article searched online repositories such as PubMed, CINAHL, IEEE Xplore, Scopus, and Google Scholar to retrieve relevant articles. After screening and pre-processing, 22 studies fulfilled the inclusion criteria. The scoping review found that the use of deep learning techniques has rapidly increased from the year 2022 and has shown outstanding performance in prediction, diagnosis, classification, and management of ovarian cancer, particularly in the field of histopathology. Histopathological images were frequently used as the predominant clinical data type. It could happen because convolution neural network can automatically extract hierarchical features from raw data without the need for feature engineering, thus making them well-suited for tasks involving histopathological image analysis, which is prevalent in diagnosing ovarian cancer.

External IDs:dblp:conf/fllm/KhanFZBS24