Methods for quickly gauging the antibacterial efficacy of new pharmaceuticals can assist healthcare providers in the timely administration of effective treatment. Rapid antibacterial susceptibility testing (RAST) techniques developed for this task are time consuming and require bulky, expensive equipment, which hinders their use in resource-limited areas. An inexpensive testing solution from Pennsylvania State University applies machine learning to the analysis of time-resolved dynamic laser speckle imaging (DLSI) results.

DLSI technology captures the change in bacterial motion in response to antibiotic treatments, and its use with machine learning enables quick determination of the minimum inhibitory concentration (MIC) of specific drugs. The system reduces detection time by eliminating the need for advanced optical setup and by correlating bacterial micromotion with the inhibitory effects of antibiotics.

The technique was demonstrated to accurately define the MIC of ampicillin and gentamicin for a model strain of Escherichia coli in 60 minutes, compared to six hours using the currently U.S. Food and Drug Administration-approved phenotype-based RAST technique. The machine learning algorithm was trained and validated using the overnight results of a gold standard antibacterial susceptibility testing method enabling prediction of MIC with a similarly high accuracy yet substantially faster.

Additional research on this proof of concept is needed to evaluate the method’s performance on a broader panel of microorganisms and antibiotics.

Monitoring bacterial motion enables rapid and accurate identification of antibiotic resistance and the minimum drug dosage needed to inhibit bacterial growth. Source: Keren Zhou, Pennsylvania State University