Abstract PR-01: Real-time, point-of-care pathology diagnosis via embedded deep learning

There is an urgent need for widespread cancer diagnosis in low resource settings, especially in contrast to areas with developed healthcare systems. According to a study in The Lancet, in the U.S. there is one pathologist for every 20,000 individuals, while in Sub-Saharan Africa, there is only one for every million. In addition, current telepathology systems for cancer diagnosis mostly rely on pathologists performing remotely, which is low-throughput and requires more time and resources. With the growth of telepathology, remote diagnosis becomes a viable solution to address the lack of skilled pathologists in developing regions. Here, we present a cost-efficient device that incorporates embedded deep learning to achieve real time, point-of-care diagnosis of whole pathology slides. We achieve this with a low-cost, 3D-printable microscope that uses the Raspberry Pi and camera module to capture high-resolution images of slides. Then, using a weakly-supervised deep-learning model run on the NVIDIA Jetson Nano, the device is able to accurately classify the whole slide without any pixel-level annotations. Furthermore, the model’s attention-based approach to diagnosis allows us to generate human-interpretable heatmaps displaying the regions most influential to the model’s diagnosis. Our device also incorporates a touch screen and batteries to increase accessibility as an easy-to-use and low maintenance device while still maintaining an efficient runtime given the available resources. Overall, we demonstrate that the device is capable of achieving accurate, high-throughput, and interpretable cancer diagnoses in low resource settings. Citation Format: Bowen Chen, Max Lu, Jana Lipkova, Faisal Mahmood. Real-time, point-of-care pathology diagnosis via embedded deep learning [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 PR-01.