Manual analysis of histopathological images is often time-consuming and painstaking, and it is also prone to errors arising from subjective evaluation criteria and human mistakes. To address these issues, we created a fully automated workflow (AW) to enumerate jejunal crypts in a microcolony survival assay to quantify gastrointestinal damage from radiation. After obtaining images of jejunal slices from irradiated mice, the AW performed cropping and normalizing of the individual slice images for resolution and color; using deep learning to detect crypts on each slice; using a tailored algorithm to enumerate the crypts; and tabulating and saving the results. A graphical user interface was developed to allow users to review and correct the automated results. Manual counting of crypts exhibited a mean absolute percent deviation of (34 ± 26)% between individuals vs. the group mean across counters, which was reduced to (11 ± 6)% across the 3 most-experienced counters. AW counts deviated from experts’ mean counts by (10 ± 8)%. With a single button press, the AW processed a sample image dataset from 60 mice in a few hours. The AW thereby allowed rapid, automated evaluation of the microcolony survival assay with accuracy comparable to that of trained experts and without subjective inter-observer variation.
Baikalov et al. (2026) studied this question.