Related Experiment Video
Updated: Mar 15, 2026
Detection and Isolation of Cancer in Prostate Biopsies Using Stimulated Raman Histology and Artificial Intelligence
Published on: June 10, 2025
Neural networks improve brain cancer detection with Raman spectroscopy in the presence of operating room light
Michael Jermyn1, Joannie Desroches2, Jeanne Mercier2
1McGill University, Montreal Neurological Institute and Hospital, Department of Neurology and Neurosurgery, 3801 University Street, Montreal, Quebec H3A 2B4, CanadabPolytechnique Montreal, Department of Engineering Physics, CP 6079, Succ. Centre-Ville, Montreal, Quebec H3C 3A7, Canada.
Abstract:
Invasive brain cancer cells cannot be visualized during surgery and so they are often not removed. These residual cancer cells give rise to recurrences.
Insights
Artificial neural networks (ANNs) coupled with in vivo Raman spectroscopy can detect invasive brain cancer cells during surgery. This approach overcomes operating room light artifacts, potentially reducing residual cancer and improving patient survival.
Area of Science:
- Neuro-oncology
- Medical Spectroscopy
- Artificial Intelligence in Medicine
Background:
- Invasive brain cancer cells are often missed during surgery, leading to tumor recurrence.
- Current methods lack the sensitivity to detect residual microscopic cancer cells.
- Raman spectroscopy shows promise for intraoperative cancer detection but is susceptible to artifacts.
Purpose of the Study:
- To evaluate the efficacy of in vivo Raman spectroscopy combined with artificial neural networks (ANNs) for detecting invasive brain cancer cells.
- To assess the ability of ANNs to overcome spectral artifacts in the operating room environment.
- To determine if this integrated approach can simplify intraoperative use and improve surgical outcomes.
Main Methods:
- Utilized in vivo Raman spectroscopy to acquire spectral data from patients with grade 2 to 4 gliomas.
- Developed and applied artificial neural networks (ANNs) with nonparametric and adaptive models to analyze spectral data.
- Tested the system's robustness against spectral artifacts generated by operating room lights.
Main Results:
- ANNs successfully detected complex nonlinear spectral characteristics indicative of invasive cancer cells.
- The coupled ANN-Raman spectroscopy approach demonstrated robustness against spectral artifacts.
- The method showed potential for simplifying intraoperative Raman spectroscopy use within standard surgical workflows.
Conclusions:
- Artificial neural networks can enhance the reliability of in vivo Raman spectroscopy for intraoperative brain cancer detection.
- This technology has the potential to significantly reduce residual cancer after surgery.
- Improved detection of invasive cancer cells may lead to enhanced patient survival rates for glioma patients.
