A Stand‐Off Laser‐Induced Breakdown Spectroscopy (LIBS) System for Remote Bacteria Identification

Yong Cheng,Shuqing Wang,Fei Chen,Jiahui Liang,Yan Zhang,Lei Zhang,Wangbao Yin,Suotang Jia,Liantuan Xiao
DOI: https://doi.org/10.1002/jbio.202400332
IF: 3.3899
2024-09-22
Journal of Biophotonics
Abstract:This study utilizes laser‐induced breakdown spectroscopy (LIBS) and machine learning methods to achieve remote bacterial identification. The Random Forest (RF) model demonstrated excellent performance in classification accuracy, recall, precision, and F1‐score, providing more accurate identification results and showing broad application prospects. Bacteria are the primary cause of infectious diseases, making rapid and accurate identification crucial for timely pathogen diagnosis and disease control. However, traditional identification techniques such as polymerase chain reaction and loop‐mediated isothermal amplification are complex, time‐consuming, and pose infection risks. This study explores remote (~3 m) bacterial identification using laser‐induced breakdown spectroscopy (LIBS) with a Cassegrain reflective telescope. Principal component analysis (PCA) was employed to reduce the dimensionality of the LIBS spectral data, and the accuracy of support vector machine (SVM) and Random Forest (RF) algorithms was compared. Multiple repeated experiments showed that the RF model achieved a classification accuracy, recall, precision, and F1‐score of 99.81%, 99.80%, 99.79%, and 0.9979, respectively, outperforming the SVM model and providing more accurate remote bacterial identification. The method based on laser‐induced plasma spectroscopy and machine learning has broad application prospects, supporting noncontact disease diagnosis, improving public health, and advancing medical research and technological development.
optics,biochemical research methods,biophysics
What problem does this paper attempt to address?