Known as the second most common cancer worldwide, breast cancer is also by far the most frequently occurring tissue disease amongst women.
Malignant cells remaining in margins of the excised resection specimen often require a second surgery.
FLIm constitutes a label-free tissue imaging modality, which enables the analysis of various types of breast tissue, and that helps discriminate between cancer, fibrous, and adipose tissue components.
A compact scanning fiber-based system could enable intra-operative acquisition of FLIm images for real-time feedback for the surgeons. To that end, an overlay of the real image and the fluorescence life-time image within the surgeon’s fov is envisaged.
Currently, FLIm maps of specimens resected during surgery are acquired to identify positive margins. Reliable classification of FLIm results is achieved using machine learning algorithms with a Random Forest classifier and with accurate image registration procedures of the FLIm maps against histopathology images, the gold standard to grade tissue on a benign-malignancy scale.
Collaborators
Dr. Richard J. Bold, M.D. (UC Health)
Dr. Candice A. M. Sauder, M.D. (UC Health)
Dr. Morgan A. Darrow, M.D. (UC Health)
Funding
NIH (National Institutes of Health): R03 EB026819