Towards rapid and early cancer diagnostics at the point of care using machine learning assisted vibrational spectroscopy

"Cancer continues to be one of the most prevalent diseases worldwide. Cancer is associated with a very high mortality rate (50% survival at 10-years) as most cancers are diagnosed at a late stage. Recent advances have been made in early detection, though the assays employed are still experimental, highly expensive and can suffer from poor sensitivity and specificity.

On the other hand, vibrational spectroscopy (Mid-IR and Raman) has shown to be robust in detecting cancer-specific analytes within the blood and other bodily fluids.

This Ph.D. project focuses on advancing cancer diagnostics through vibrational spectroscopy, specifically targeting novel biomarkers such as circulating tumor DNA (ct-DNA), epigenetic markers, and microRNA (miRNA). Utilizing infrared and Raman spectroscopy, the project aims to unveil unique vibrational profiles associated with these biomolecules in biological samples.

Innovative on-chip photonic device platform will be developed to enhance the precision of biomarker detection. Integrating machine learning, the research aims to develop computational models for the identification, classification, and prediction of cancer based on these specific biomarkers. Anticipated outcomes include a robust set of vibrational spectroscopy-based biomarkers detection, offering a significant stride toward personalized cancer diagnostics through non-invasive and real-time assessment of ct-DNA, epigenetic markers, and miRNA expression.

This research holds promise for transforming cancer diagnosis with early detection and precise characterisation of cancer subtypes. "