Study of AI Algorithms to Classify Breast Cancer Types from Breast Tissue Images

What this Trial Is About

Breast cancer is a common and potentially deadly disease that highlights the importance of early and accurate diagnosis to improve patient outcomes. Traditional diagnosis relies on manual examination of tissue samples under a microscope, which can be subjective and slow. This research aims to improve breast cancer diagnosis by using advanced computer techniques called deep learning to classify types of breast cancer from tissue images. The types include noninvasive cancers, confined within ducts or lobules, and invasive cancers that spread into surrounding tissue and make up about 70% of cases with generally worse outcomes. The study involves analyzing breast tissue samples collected through biopsy or mastectomy using histopathology, the current gold standard diagnostic method. Researchers will use deep learning models to classify the tissues as normal, benign, noninvasive in situ, or invasive cancer. This AI approach hopes to improve the accuracy and speed of diagnosis compared to manual methods. The model is designed to be adaptable to different healthcare settings, aiming to increase access to reliable breast cancer screening, especially in underserved or low-resource areas. Participants are women who have undergone biopsy due to suspected abnormal cell growth in the breast. Their tissue samples will be analyzed with the AI model within one week of biopsy or surgery. Researchers will assess the accuracy of the AI in classifying breast tissue types. The study focuses on improving cancer detection to support early treatment decisions and better outcomes. Participants may undergo biopsy and mastectomy procedures as part of their care, with histopathological analysis used for diagnosis. The study does not include follow-up screenings and excludes pregnant women and those with severe medical conditions that might affect results or safety.

AI Model for Classifying Breast Cancer From Histopathology Images