High-resolution three-dimensional blood flow tomography in the subdiffuse regime using laser speckle contrast imaging

Area of Science:

• Biomedical Optics
• Computational Imaging
• Neuroscience

Background:

• High-resolution visualization of cerebral hemodynamics is crucial for understanding brain diseases.
• Current optical imaging methods struggle with both spatial resolution and temporal speed over large fields of view.

Purpose of the Study:

• To develop a high spatial resolution computational optical imaging technique for reconstructing blood flow maps in complex tissues over a large field of view.
• To overcome limitations of existing methods in spatial and temporal resolution for hemodynamic imaging.

Main Methods:

• Utilized laser speckle contrast imaging (LSCI) principles combined with perturbation Monte Carlo simulations.
• Employed a mini-batch gradient descent with adaptive learning rate for iterative reconstruction of blood flow maps.
• Leveraged parallelization and vectorization for efficient forward and derivative calculations.

Main Results:

• Successfully reconstructed high-resolution blood flow maps in simulated phantoms and murine cerebral tissue.
• Achieved reconstruction errors below 2% for most vasculature, demonstrating robustness to noise.
• Showcased ability to visualize blood flow down to 500 μm depth with high spatial and temporal resolution.

Conclusions:

• Demonstrated a powerful computational imaging technique for large-field-of-view, high-resolution blood flow visualization.
• The method requires minimal LSCI images post-structural capture for computationally intensive reconstruction.
• Well-suited for dynamic monitoring of blood flow in applications like functional neural imaging.