Signal in the Scatter: Advancing Speckle-Based Optical Sensing for Deep Cerebral Blood Flow Monitoring

This thesis presents the development and application of novel, non-invasive optical techniques for monitoring cerebral blood flow (CBF) and cerebral blood volume (CBV), addressing the critical need for cost-effective and scalable solutions in cerebrovascular health assessment. The research introduces advancements in interferometric speckle visibility spectroscopy (iSVS) and speckle contrast optical spectroscopy (SCOS) to overcome the challenges of signal attenuation and noise when measuring blood flow through the scalp and skull. The depth sensitivity of these optical methods was first experimentally determined. Using iSVS on phantoms, rabbits, and human subjects, a two-layer decay model was observed as the source-to-detector (S-D) distance was varied, allowing for the quantification of the transition point from superficial to cerebral signal detection. Complementing this, a multi-channel SCOS system was used with temporary occlusion of the superficial temporal artery to experimentally isolate and quantify the influence of scalp blood flow, providing direct evidence of brain-to-scalp signal sensitivity and establishing optimal S-D configurations. Another primary focus of this work was enhancing the signal-to-noise ratio (SNR) of deep-tissue measurements. A comprehensive theoretical framework for iSVS was developed to evaluate its SNR in the presence of detector noise, confirming its superiority in photon-limited regimes and revealing relaxed constraints on the reference beam. In parallel, a compact, fiber-free SCOS device was engineered, demonstrating a 70-fold increase in signal collection over traditional fiber-based systems with enhanced stability. The SNR for SCOS was further improved through an optimization-based, adaptive noise calibration framework that mitigates artifacts from cerebral blood volume fluctuations, significantly lowering the signal detection threshold for reliable CBF measurement. Building upon these foundational advancements, the research progressed to clinical applications. The technology's modularity was demonstrated by engineering a portable, six-channel SCOS system for simultaneous, real-time measurements at multiple brain locations. This system was validated in a preliminary study on a patient with traumatic brain injury, demonstrating its potential for characterizing regional cerebrovascular dysfunction by comparing blood flow dynamics against structural MRI data. Furthermore, the compact SCOS device was used to assess stroke risk in a 50-person cohort by monitoring cerebrovascular reactivity during a breath-holding task; this revealed significant discrepancies between CBF and CBV responses that correlated with risk scores.