Deep Learning-Based Mitosis Detection in Breast Cancer Histopathology Images: A Mapping Study

Mitosis detection refers to the process of identifying and locating mitotic cells within histological or microscopic images. Mitosis, a crucial stage in the cell cycle, involves cell division, and detecting mitotic figures is particularly important in various fields, especially in cancer diagnosis and research. Pathologists takes more time to detect the mitosis in traditional way, also there may be huge variants between experts. As a result, there has been significant interest in developing automated methods for mitosis detection in histopathology images. Several automatic mitosis detection techniques have been proposed in recent years. Five electronic research databases were searched in order to complete the mapping investigation. We took into account research articles released between January 2013 and May 2023. There were 227 studies in the first batch of results. A total of 79 studies were chosen from this set. We conducted a direct search of research organizations’ and individuals’ publications as well as snowballing techniques to find the 73 researchers who completed these investigations. 79 papers addressing Deep learning-based mitotic detection in breast cancer were found through the mapping project, and these have been examined to gather pertinent data for a number of research concerns. This mapping study will help the future researchers to easily identify the relevance within each study and will lead to continuous advancements in deep learning architectures and algorithms will likely lead to improved accuracy and efficiency in mitosis detection. Refinement of models, such as convolutional neural networks (CNNs) and more sophisticated architectures, may enhance the ability to detect mitotic figures with higher precision.