Spatially Resolved Fibre-Optic Probe for Cervical Precancer Detection Using Fluorescence Spectroscopy and...

Area of Science:

• Biomedical Optics
• Medical Spectroscopy
• Cancer Diagnostics

Background:

• Early detection of cervical cancer relies on identifying subtle biochemical and morphological changes in the epithelium.
• Conventional methods for cervical cancer detection can be invasive and time-consuming.
• Fluorescence spectroscopy offers a non-invasive, real-time approach to detect these changes with high accuracy.

Purpose of the Study:

• To develop and validate a custom-designed, spatially resolved fiber-optic probe (SRFOP) for cervical cancer detection using fluorescence spectroscopy.
• To classify different grades of cervical precancer based on their fluorescence spectra.
• To evaluate the efficacy of a robust classification algorithm integrating principal component analysis (PCA) and artificial neural networks (ANN).

Main Methods:

• Utilized a custom SRFOP with 77 fibers arranged in concentric rings for spectral acquisition.
• Excited cervical tissue samples (n=28) with a 405 nm laser diode and collected fluorescence spectra (400-700 nm) using a USB 4000 spectrometer.
• Applied PCA for dimensionality reduction, selecting the top 10 principal components as input features for an ANN classifier.

Main Results:

• Proximal fibers (PFs) of the SRFOP demonstrated superior performance over distal fibers (DFs) in capturing discriminatory spectral features from the epithelium.
• The combined PCA-ANN classification approach achieved high diagnostic performance.
• Achieved an average sensitivity of 93.33%, specificity of 96.67%, and accuracy of 95.57% in discriminating cervical precancer grades and normal tissues.

Conclusions:

• Spatially resolved fluorescence spectroscopy, coupled with PCA-ANN, is a highly effective non-invasive technique for early cervical cancer detection.
• The SRFOP design and data analysis pipeline show significant promise for clinical application in cervical precancer diagnosis.
• This approach offers a sensitive and specific method for differentiating normal cervical tissue from various grades of precancer.