Abstract PO-062: Fully automated artificial intelligence based breast cancer scanning through transilluminated image acquisition for mass screening appliances
P. Ponram · Clinical Cancer Research · 2021
Abstract Early detection of breast cancer is an effective tool in better management and treatment of the same. Mass screening is one of the available methods to detect early breast masses. Current technologies such as ultrasonography and mammography are golden standards of breast cancer detection. But these imaging modalities have several complexities when used for mass screening such as labor-intensive and cost. Hence a low-cost automated breast screening device is proposed in this project. A transillumination-based non-ionizing and non-radiating near-infrared optical wave are used for the transillumination of breast masses and their corresponding images will be acquired. The acquired images will be given to a deep learning artificial intelligence network for automated detection and classification of the tumor. The AI network will be pre-trained with double validated data with clinician expertise in the domain. The AI outcome will be displayed to the user and also taken as backpropagation feedback to the AI network to train the system further. This device shall be used as a system for mass screening applications from which suspected cases can e referred for mammography for final confirmation of tumor. The automated system shall be used in rural places with less skilled human intervention for mass public screening programs. This will reduce the disease burden on the huge women population vulnerable to the disease. Citation Format: P. Ponram. Fully automated artificial intelligence based breast cancer scanning through transilluminated image acquisition for mass screening appliances [abstract]. In: Proceedings of the AACR Virtual Special Conference on Artificial Intelligence, Diagnosis, and Imaging; 2021 Jan 13-14. Philadelphia (PA): AACR; Clin Cancer Res 2021;27(5_Suppl):Abstract nr PO-062.