Non-invasive optical blood flow monitoring systems for disease diagnosis and healthcare monitoring have been studied. Diffuse Speckle Contrast Analysis (DSCA) system can measure deep-tissue blood flow with a relatively simple system configuration, high speed, and high sensitivity. However, the relative blood flow index (BFI) is acquired with the system, and it changes with every acquisition. In this study, we adopt machine learning to overcome this limitation. DSCA system was established with a micro-size camera, and the correlation between conventional BFI and ML-based BFI was analyzed. This work will be the first step toward a quantitative Diffuse Speckle Contrast Velocimetry (DSCV).
|