A Cell Image Classification Method for Quality Control of Chimeric Mice with Humanized Livers

In this study, we developed a discriminator that determines the grade of human hepatocytes for quality control of chimeric mice with humanized livers. Currently, the grading of human hepatocytes is divided into four categories, Grade S, Grade A, Grade B, and Grade C, which are qualitatively judged visually by experts, and the grade judgment differs depending on the person who judges. Hence, in this study, we quantitatively determined the grade of human hepatocytes by developing a discriminator using deep learning. We will discuss the results of discriminators learning and test by the 3-class classification that combines Grade S and Grade A into one. As a result, the 3-class classification gave a discrimination accuracy of 93.2%.