Intelligent convolution neural network-assisted SERS to realize highly accurate identification of six pathogenic .
Journal:
Chemical communications (Cambridge, England)
Published Date:
May 9, 2023
Abstract
Based on label-free SERS technology, the relationship between the Raman signals of pathogenic microorganisms and purine metabolites was analyzed in detail. A deep learning CNN model was successfully developed, achieving a high accuracy rate of 99.7% in the identification of six typical pathogenic species within 15 minutes, providing a new method for pathogen identification.