The tumor immune microenvironment(TIME)of colorectal cancer contains indicators of unique therapeutic outcomes for each cancer patient. Deep learning-based imaging cytometry(DL-IC), which can obtain objective and reproducible cell- related information in tissue sections, has attracted attention as an analytical method for clarifying this indicator. This study demonstrates the validation process of Cu-Cyto, one of DL-IC, regarding cell identification accuracy. Acquisition of"spatial structure"information in TIME is useful for biomarker retrieval and contributes to precision oncology.