Fast and Noninvasive Diagnosis of Cervical Cancer by Coherent Anti-Stokes Raman Scattering

Area of Science:

• Biomedical Optics
• Cancer Diagnostics
• Cellular Imaging

Background:

• Cervical cancer is a leading cause of death in women globally.
• Current detection methods (cytopathology, HPV testing, histopathology) have limitations like subjectivity and time constraints.
• There is a critical need for noninvasive, accurate, and rapid early detection techniques.

Purpose of the Study:

• To develop and validate a novel noninvasive, label-free imaging approach for early detection of cervical precancerous lesions.
• To assess the efficacy of CARS, SHG/TPF, and Raman imaging combined with deep convolutional neural networks (DCNNs) for classifying cervical cells.

Main Methods:

• Utilized Coherent Anti-Stokes Raman Scattering (CARS) and Second Harmonic Generation/Two-Photon Fluorescence (SHG/TPF) imaging on liquid-based cytology Pap smears.
• Performed Raman spectral imaging for biochemical analysis of normal, low-grade squamous intraepithelial lesion (LSIL), and high-grade squamous intraepithelial lesion (HSIL) cells.
• Trained DCNNs using morphotextural and spectral data from CARS, SHG/TPF, and Raman images for cell classification.

Main Results:

• CARS, SHG/TPF, and Raman imaging-based DCNNs achieved 100% accuracy in discriminating between normal and cancerous Pap smears.
• The imaging techniques effectively utilized both morphotextural and spectral information for accurate classification.
• Normal, LSIL, and HSIL cells were distinguishable based on nucleus-to-cytoplasm ratio and cell morphology.

Conclusions:

• CARS/SHG/TPF microscopy represents a promising label-free imaging technique for rapid screening of cytopathological samples.
• This noninvasive approach has significant potential for the early detection of cervical cancer.
• The integration of advanced imaging and DCNNs offers a powerful tool for cytopathology.