Intraoperative tissue identification for gastrointestinal disorder by Raman spectroscopy and machine learning

In this study, we evaluated surgical specimens obtained from patients for detecting resection merging in Hirschsprung's disease. Conventional multivariate analyses successfully characterized Raman spectral data. Furthermore, the Raman spectroscopic approach combined with machine learning methods successfully predicted whether the target specimen was healthy or diseased by the decision algorithm. Toward practical use, we developed a portable Raman spectroscopic system and a fiber-optic Raman probe for laparoscopic surgery. we performed in vivo Raman measurement of abdominal organs using a live porcine during laparoscopy.