UC Irvine

Perfusion is an important biomarker that can be used as a diagnostic metric and a predictor of tissue viability, or, when altered, as the target of treatment. Laser speckle imaging (LSI) is one of the optical methods that have received wide recognition in skin viability assessment applications. These applications require more than a single perfusion value. They also need perfusion distribution, a high refresh rate, and patient comfort during the procedure, all of which LSI can provide as a non-invasive technique for full-field real-time imaging. Despite being widely adopted, LSI operates under assumptions that, while making it relatively simple to use, may limit its clinical applicability. To address these limitations, a deeper look into LSI validation methods and adjacent techniques is needed.First, to enhance LSI measurement accuracy and reliability, the extent to which day-to-day instrument variability degrades the reproducibility of LSI data was assessed, and a static-phantom-imaging-based β-correction approach was developed and validated. Correcting for β empirically brought variability within a single system down from ~13% to ~2%. Next, the β-correction approach was tested in the burn wound assessment study. Applied to a clinically relevant heterogeneous burn wound model, LSI separated SPT from DPT burns with up to 94% accuracy at 72 h post-burn. Regressing the perfusion values against histology produced prediction maps of burn depth and remaining healthy dermis, with errors falling within standard surgical excision thickness. Lastly, to battle the static scattering effect on the LSI measurement, which is critical in burn research, a Dynamic Light Scattering Imaging (DLSI) technique was studied. Being adjacent to LSI, DLSI uses the same founding principle while accounting for static scattering and for the type of dynamics present in the sample, including mixtures of more than one dynamic regime. The barriers keeping DLSI from clinical use include processing time and model complexity, both of which are addressed in this work. Collectively, this work advances LSI toward clinical applicability and addresses several of its inherent limitations through a modification of the adjacent technique, DLSI. The methods developed here and the results they produced show how speckle-based techniques can become valuable tools in tissue viability assessment, with the modified DLSI technique in particular offering, once tested in skin, a path to earlier and more accurate diagnosis and better burn healing outcomes.