Classifying breast cancer tissue by Raman spectroscopy with one-dimensional convolutional neural network

Area of Science:

• Biomedical Engineering
• Medical Diagnostics
• Computational Biology

Background:

• Breast cancer is the most common cancer in women, with current diagnostic methods facing limitations in sensitivity, speed, and reliance on expert interpretation.
• There is a critical need for innovative, automated diagnostic technologies to improve breast cancer detection and patient outcomes.
• Existing diagnostic tools often lack the required sensitivity for early-stage detection and can be time-consuming.

Purpose of the Study:

• To develop and validate a novel automated method for classifying healthy versus cancerous breast tissues.
• To investigate the efficacy of combining Raman spectroscopy with a one-dimensional convolutional neural network (1D-CNN) for breast tissue analysis.
• To compare the performance of the 1D-CNN approach against traditional classifiers like Fisher Discrimination Analysis (FDA) and Support Vector Machine (SVM).

Main Methods:

• Raman spectroscopy was employed to collect spectral data from breast tissue samples obtained from 20 patients.
• A one-dimensional convolutional neural network (1D-CNN) model was developed and trained using the collected spectral data for tissue classification.
• Comparative analysis was performed using Fisher Discrimination Analysis (FDA) and Support Vector Machine (SVM) classifiers on the same dataset.

Main Results:

• The 1D-CNN model demonstrated superior performance in classifying breast tissues.
• The 1D-CNN achieved an overall diagnostic accuracy of 92%, with a sensitivity of 98% and a specificity of 86%.
• Fisher Discrimination Analysis and Support Vector Machine classifiers were also evaluated for comparative performance.

Conclusions:

• The combination of Raman spectroscopy and 1D-CNN provides an effective and automated method for breast tissue classification.
• This approach shows significant potential for improving the accuracy and efficiency of breast cancer diagnosis.
• The study lays the groundwork for future advancements in automated cancer diagnostic systems.