Application of state of the art raman chemical imaging and chemometrics to accelerate and improve patient prostate biopsy assessment for cancer.

A pathologists' task in assessing a prostate biopsy for cancerous tissue is of prime importance since failure to correctly identify tissue can lead to a failure to accurately diagnose the patient and predict their likely prognosis. Histological staining assists in the assessment of prostate biopsies by making important features of the biopsy more visible. However staining has some key limitations. Firstly, a pathologist still needs to examine the biopsy at high magnification which makes rigorous tissue assessment unrealistically laborious and tiring. Computer based image processing tools can be implemented to help shoulder this burden but are dependent on the quality of the biopsy stain. Secondly, staining is inconsistent and the intensity and clarity of the stain can vary greatly from hospital to hospital. This can mislead the pathologist and will confound any computerised pathology tool. The third problem is that the staining process itself is time consuming, laborious and expensive.

In order to overcome these limitations we propose a novel data driven approach (Figure 1). Raman Chemical Imaging will be used to generate biochemical maps of prostate tissue independent of the staining process. The Raman data generated will be optimally combined with that arising from existing methodologies (staining and digital imaging) in order to allow automatic identification of important features of prostate biopsy. Data modelling (chemometric) tools will be developed to merge imaging data with existing patient records to predict patient outcome. This approach will provide additional information on the patient to the prostate cancer pathologist and the medical team. Hence the decision to recommend a patient for treatment or not will be better informed.

Award Date
21 October 2016
Award Value
Principal Investigator
Professor Aoife Gowen
Host Institution
University College Dublin
Health Research Awards