Related Experiment Video
Updated: Sep 24, 2025
08:05
Detection and Isolation of Cancer in Prostate Biopsies Using Stimulated Raman Histology and Artificial Intelligence
Published on: June 10, 2025
691
Cancer detection from stained biopsies using high-speed spectral imaging.
Eugene Brozgol1,2, Pramod Kumar1,2, Daniela Necula3
1Physics Department and Nanotechnology Institute, Bar Ilan University, Ramat Gan, Israel.
Biomedical Optics Express
|May 6, 2022
Summary
Spectral imaging offers a powerful new method for automated cancer diagnosis in pathological biopsies. This approach significantly improves accuracy over traditional color imaging, aiding faster patient treatment.
Area of Science:
- Biomedical Engineering
- Medical Imaging
- Computational Pathology
Background:
- Current pathological biopsy analysis is slow and often manual.
- Existing imaging methods lack sufficient data for reliable automated diagnostics.
- Deep learning for pathology faces challenges with data validity and performance.
Purpose of the Study:
- To develop and validate a rapid spectral imaging system for pathological biopsies.
- To enhance the accuracy of cancer cell identification using spectral data.
- To leverage artificial intelligence for improved machine-aided diagnostics.
Main Methods:
- Rapid acquisition of spectral images from pathological biopsies.
- Development of spectral classification algorithms for data analysis.
- Application of artificial intelligence (AI) algorithms on spectral image data.
Main Results:
- Spectral information proved significantly more informative than color data.
- High accuracy achieved in identifying cancer cells using spectral imaging.
- AI algorithms demonstrated high performance with a small training set, highlighting rich spectral data.
- Key spectral differences identified in the nucleus, linked to aneuploidy.
Conclusions:
- Rapid spectral imaging provides a robust method for automated biopsy analysis.
- This technology can significantly improve diagnostic accuracy and speed.
- Spectral imaging has the potential to bridge the gap in machine-aided pathology, enhancing patient care.
