Validation of the Linearity in Image Reconstruction Methods for Speckle Contrast Optical Tomography

Area of Science:

• Biomedical Optics
• Neuroimaging
• Optical Spectroscopy

Background:

• Speckle contrast optical spectroscopy (SCOS) non-invasively measures cerebral blood flow and brain function.
• Speckle contrast optical tomography (SCOT) extends SCOS for blood flow variation mapping.
• Linearity assumptions in SCOT image reconstruction may not always hold true.

Purpose of the Study:

• To develop and validate a forward model for SCOT using the Rytov approximation.
• To compare the Rytov approximation with the first Born approximation and Monte Carlo simulations.
• To demonstrate the utility of the Rytov approximation for accurate SCOT image reconstruction of human brain activity.

Main Methods:

• Constructed a SCOT forward model using the Rytov approximation to solve the correlation diffusion equation.
• Compared Rytov approximation results with first Born approximation and Monte Carlo simulations.
• Developed a blood flow index (BFi) and utilized a sensitivity matrix approach for linear forward and inverse problem solving.

Main Results:

• Rytov approximation results closely matched Monte Carlo simulations.
• First Born approximation showed significant deviation for blood flow variations exceeding 30%.
• Demonstrated successful experimental image reconstruction of human brain activations using high-density SCOS.

Conclusions:

• The Rytov approximation is recommended over the first Born approximation for SCOT when blood flow variations exceed 30%.
• The developed BFi is linearly related to local cerebral blood flow variations, enabling linear reconstruction methods.
• The validated Rytov-based forward model guides SCOT system design and data analysis for enhanced neuroimaging.