Colorimetric sensor array for the rapid distinction and detection of various antibiotic-resistant psychrophilic bacteria in raw milk based-on machine learning.

In this study, a rapid, inexpensive, and accurate colorimetric sensor for detecting psychrophilic bacteria was designed, comprising gold (Au) nanoparticles (NPs) modified by d-amino acid (D-AA) as color-metric probes. Based on the aggregation of Au NPs induced by psychrophilic bacteria, a noticeable color shift occurred within 6 h. Depending on the various metabolic behaviors of bacteria to different D-AA, four primary psychrophilic bacteria in raw milk were successfully distinguished by learning the response patterns. Furthermore, the quantification of single bacteria and the practical application in milk samples could be realized. Notably, a rapid colorimetric method was constructed by combining Au/D-AA with antibiotics for the minimum inhibitory concentration of psychrophilic bacteria, which relied on differences in bacteria metabolic activity in response to diverse antibiotic treatments. Therefore, the method enables the rapid detection and susceptibility evaluation of psychrophilic bacteria, promoting clinical practicability and antibiotic management.