Multi-species blood identification is especially useful in animal quarantine, import and export, criminal cases, forensic examination, and wildlife conservation. Raman spectroscopy is a non-destructive, label-free, and highly specific method for providing chemical information on materials and serving as an analytical tool to characterize biological samples. Machine learning approaches in conjunction with Raman tweezers are proposed to extract the Raman spectral characteristics of single red blood cells (RBCs), then the cells based on the spectral characteristics are classified, and some classification prediction models to achieve single cell identification of different blood species are achieved finally.
|