Optimisation of Raman spectral processing for classification of radiotherapeutic toxicity

Background: Severe radiation toxicity can continue years after the completion of radiotherapy for prostate cancer patients. Currently, it is impossible to predict before treatment which patients will experience these long-term side effects. New approaches based on vibrational spectroscopy have advantages over lymphocyte and genomic assays in terms of minimal sample preparation, speed and cost. A high throughput method has been developed to measure Raman spectra from liquid plasma in a cover glass bottomed 96 well plate. However, the Raman spectra can show contributions from glass and water. The current study aims to optimise pre-processing steps to improve classification performance. Methods: Blood samples (n=32) were obtained from prostate cancer patients enrolled on the EU-funded REQUITE study (www.requite.eu) through a collaboration with the University of Leicester. Raman spectra were recorded from plasma samples using an in-house developed high throughput method. Extended multiplicative scattering correction (EMSC) was used for background correction. Results: Raman spectra of plasma were corrected individually and together for glass and water interference. It was observed that a good model efficiency was achieved for prostate cancer patients with no/minimal radiation toxicity (grade 0-1) and severe radiation toxicity (grade 2-3) after corrections with both glass and water rather than individually. Conclusions: It could be concluded from the study that to achieve good overall model efficiency, both glass and water corrections are required, when compared to no corrections, only glass corrections and only water corrections.