High-Speed Optical Microscopy Imaging of a Large Number of Spots of Tissue Samples and Detection of Cancer Using Artificial Intelligence (AI)

Cancer is the leading cause of death worldwide, with almost 10 million deaths in 2020 alone. Cancer does not usually show symptoms until it has spread and is generally diagnosed at a very later stage. This late-stage diagnosis significantly reduces the chances of survival. Early diagnosis helps?significantly?in?detecting?and?treating?cancer. However, the?traditional?method of manual imaging and identification of?cancer tissue characteristics?is?time-consuming?and tedious.?The?search?for?a better alternative for?robust and?accurate?cancer?detection?requires?high-volume imaging and statistical analysis of cancerous?tissue.?In this study, we report high-speed optical transmission/ reflectance?imaging of tissue samples using Tissue Microarrays?(TMA) cancer/control?tissue samples. These images are analyzed using the?artificial intelligence?(AI) algorithm. A conventional Olympus BX61 motorized research-grade microscope and custom automatic scanning patterns are used for imaging. The?image analysis is then performed using the AI-based Convolutional Neural Network (CNN) algorithm. The?results?show?robust and?accurate detection of control and cancer?stages?of tissues. The potential for pathological applications of the technique in cancer detection?is?discussed.