Subtype-Specific Detection in Stage Ia Breast Cancer: Integrating Raman Spectroscopy, Machine Learning, and Liquid...

Area of Science:

• Biomedical engineering
• Medical diagnostics
• Spectroscopy

Background:

• Early breast cancer detection and subtyping are crucial for effective treatment.
• Current diagnostic methods can be invasive or lack specificity for early stages.
• Vibrational spectroscopy offers a label-free approach to analyze biological samples.

Purpose of the Study:

• To integrate Raman spectroscopy (RS) with machine learning for early breast cancer detection.
• To classify the four major breast cancer subtypes using blood plasma.
• To establish a non-invasive, personalized diagnostic tool for early-stage breast cancer.

Main Methods:

• Collected blood plasma samples from breast cancer patients and healthy controls.
• Performed Raman spectroscopy to acquire detailed spectral data.
• Utilized machine learning algorithms for spectral pattern analysis and classification.

Main Results:

• Identified significant spectral patterns in plasma associated with breast cancer biomarkers.
• Achieved high classification performance for the four major breast cancer subtypes at stage Ia.
• Reported average sensitivity of 90%, specificity of 95%, and cross-validated AUC of 0.98.

Conclusions:

• Raman spectroscopy combined with machine learning shows significant potential for early breast cancer detection and subtyping.
• This integrated approach offers a non-invasive, personalized method for cancer diagnostics and monitoring.
• This study represents a novel application of RS and machine learning for classifying early-stage breast cancer subtypes.