Investigation of speckle contrast optical tomography performance for imaging cerebral blood flow in a human head

High-density speckle contrast optical tomography (HD-SCOT) is a potentially attractive technique for bedside imaging of cerebral blood flow (CBF). To evaluate the performance of HD-SCOT, we built a pipeline with an anatomical head model for simulating measurements and reconstructed images. We observed that the cortical region is well represented by measurements with source-detector distance at least 29 mm. Including larger source-detector distances can reduce localization errors but with reduced signal-to-noise ratio (SNR). Imaging performance is highly dependent on the exposure time, with optimal exposure time dependent on the noise model. Future HD-SCOT systems will be designed using these results.