Related Experiment Video
Updated: Jan 7, 2026
07:54
Surface-enhanced Resonance Raman Scattering Nanoprobe Ratiometry for Detecting Microscopic Ovarian Cancer via Folate Receptor Targeting
Published on: March 25, 2019
8.6K
Minimally Invasive, Label-Free, Point-of-Care Histopathological Diagnostic Platform of Malignant Tumors of the Female
Liangliang Jiang1, Siqi Gong2, Zibo Gao2
1Harbin Medical University Cancer Hospital; No. 150, Haping Road, Nangang District, Harbin, Heilongjiang 150081, China.
The Journal of Physical Chemistry Letters
|January 5, 2026
Summary
This study introduces an ultra-fast, label-free method using Raman spectroscopy and machine learning for gynecological cancer diagnosis. The technology accurately identifies tissue types in under a minute, aiding surgical decisions.
Area of Science:
- Biomedical Optics
- Computational Pathology
- Gynecologic Oncology
Background:
- Intraoperative histopathology is crucial for gynecologic cancer surgery.
- Current methods like frozen-section pathology are time-consuming and require significant resources.
- There is a need for rapid, accurate diagnostic tools at the point of care.
Purpose of the Study:
- To develop and validate an ultrarapid, label-free histopathology platform using Raman spectroscopy and machine learning.
- To assess the diagnostic accuracy of this platform for gynecologic tissue specimens.
- To enable real-time intraoperative decision-making for gynecologic cancer surgery.
Main Methods:
- Acquisition of 4750 Raman spectra from 85 human gynecologic tissue specimens across 19 histopathological classes.
- Preprocessing of spectral data and classification using five machine-learning algorithms (Support-vector machines, Random Forest, k-Nearest Neighbor).
- Performance evaluation using stratified 70%:30% train-test splits.
Main Results:
- Support-vector machines achieved 100% accuracy (AUC = 1.00) across all classes, outperforming other algorithms.
- Single-spectra acquisition took 30 seconds, with automated prediction under 8 seconds, enabling decisions within 1 minute.
- Raman spectroscopy identified biochemical alterations (nucleic acids, amino acids, collagen) invisible to routine microscopy.
Conclusions:
- Raman spectroscopy coupled with machine learning provides an ultrarapid, label-free platform for accurate discrimination of malignant, benign, and premalignant gynecologic lesions.
- This technology has the potential to reduce operative time, minimize repeat surgeries, and improve accessibility of high-quality histopathology.
- The platform enables real-time diagnostic decisions at the point of care, enhancing surgical precision in gynecologic oncology.
