December 2025—The Journal of Clinical Microbiology has published an article describing ARUP Laboratories’ validation of a deep convolutional neural network to detect parasites in concentrated wet mounts of stool (Mathison BA, et al. J Clin Microbiol. 2025;63[11]:e0106225).
The convolutional neural network was trained with 4,049 unique parasite-positive specimens sourced from around the world. The clinical validation included 25 classes of parasites. The study found that artificial intelligence detected an additional 169 organisms that were not originally identified by a technologist. After discrepancy analysis, the positive agreement between AI and manual review was 98.6 percent.
ARUP Laboratories, 801-583-2787