US researchers use AI for quick detection of bacteria

19 December 2017 (Last Updated December 19th, 2017 11:38)

US-based microbiologists at Beth Israel Deaconess Medical Center (BIDMC) have used artificial intelligence (AI) powered microscopes to diagnose blood infections and improve patient outcomes.

US researchers use AI for quick detection of bacteria
E coli Bacteria. Credit: NIAID/Flickr.

US-based microbiologists at Beth Israel Deaconess Medical Center (BIDMC) have used artificial intelligence (AI) powered microscopes to diagnose blood infections and improve patient outcomes.

Researchers demonstrated that an automated AI-enhanced microscope system can be used for rapid and accurate identification of bacterial images.

BIDMC Clinical Microbiology Laboratory director James Kirby said: “This marks the first demonstration of machine learning in the diagnostic area.

“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.”

Led by Kirby, the team used an automated microscope and trained a convolutional neural network (CNN) to differentiate bacteria depending on shape and distribution such as rod-shaped E-coli, round Staphylococcus clusters and pairs or chains of Streptococcus.

“It can provide an unprecedented level of detail as a research tool.”

The unschooled neural network was fed more than 25,000 blood sample images prepared during routine clinical workups.

From more than 100,000 training images generated by the researchers, the machine intelligence learned to categorise bacteria with an accuracy of approximately 95%.

These training images contained bacteria that were already detected by human clinical microbiologists.

When used to sort new images from 189 slides that had no human intervention, the algorithm is reported to have demonstrated 93% accuracy.

Researchers expect that further development and training of the AI-enhanced platform would provide a fully automated classification system, addressing the shortage of clinical microbiologists.

Kirby added: “The tool becomes a living data repository as we use it. And could be used to train new staff and ensure competency. It can provide an unprecedented level of detail as a research tool.”