A Pilot Study Using Biospeckle Photography for Optical Breast Cancer Screening

Biospeckle is a powerful, affordable, non-destructive, and growing optical medical imaging field. The information-rich biospeckle pattern that appears difficult to be interpreted opens new artificial intelligence opportunities. In this preliminary study, we employed biospeckle photography associated with local contrast parameters as a new optical imaging approach for breast cancer screening, a major cause of mortality among women worldwide. The proposed method initially generated biospeckle images for normal and cancerous breast tissues. Then, the analysis approach was performed by converting each biospeckle image into a set of local contrast matrices from which the statistical parameters of their intensity histograms were computed. The multivariate analysis of the biospeckle statistical data using the principal component analysis (PCA) succeeded in finding a clear classification of the biospeckle images revealing the possibility of discrimination between normal and cancerous breast tissues. The results indicated the potential of biospeckle photography associated with local contrast measurements and histogram’s statistical parameters for nondestructive and automatic breast cancer detection.