Optical microscopy is the standard for cytopathology diagnosis, but whole slide scanners are slow and expensive. This paper introduces a Multi-Camera Array Scanner (MCAS) for rapid 3D imaging of thick specimens at high resolution. MCAS uses 48 micro-cameras for parallelized imaging, achieving significantly faster scanning than conventional methods. The authors demonstrate MCAS's use in digitizing entire cytology samples and integrating machine learning for adenocarcinoma detection (0.73 recall) and slide-level classification (0.969 AUC) in lung specimens.

Kanghyun Kim, Amey Chaware, Clare B. Cook, Shiqi Xu, Monica Abdelmalak, Colin Cooke, Kevin C. Zhou, Mark Harfouche, Paul Reamey, Veton Saliu, Jed Doman, Clay Dugo, Gregor Horstmeyer, Richard Davis, Ian Taylor-Cho, Wen-Chi Foo, Lucas Kreiss, Xiaoyin Sara Jiang, Roarke Horstmeyer