Microbiologist at the Beth Israel Deaconess Medical Center (BIDMC) in Boston have developed an artificial intelligence system capable of detecting blood infections to assist in diagnosis. Findings are published in the Journal of Clinical Microbiology.
The automated AI-enhanced system was developed by combining a specialized microscope for the collection of high-resolutions image data and a trained convolutional neural network (CNN). Led by senior author James Kirby, MD, the director of the Clinical Microbiology Laboratory at BIDMC and an associate professor at Harvard Medical School, researchers hoped the system could help clinical microbiologists accurately diagnose blood infections to improve patient survival.
“This marks the first demonstration of machine learning in the diagnostic area,” said Kirby. “With further development, we believe this technology could form the basis of a future diagnostic platform that augments the capabilities of clinical laboratories, ultimately speeding the delivery of patient care.”
To achieve high accuracy rates, researchers trained the CNN AI system to sort bacteria by shape and distribution. The system was trained with 25,000 images of routine blood samples and 100,000 training images to categorize the bacteria into rod-shaped, round clusters and round chains or pairs. Results found the system had almost 95 percent accuracy in sorting by type and 93 percent accuracy when sorting 189 new images without human intervention.