Fast and efficient image reconstruction for high density diffuse optical imaging of the human brain

Area of Science:

• Medical Imaging
• Computational Neuroscience
• Biophysics

Background:

• Real-time human brain imaging is crucial for advanced neuroimaging.
• Current methods for sensitivity map generation can be computationally intensive.

Purpose of the Study:

• To develop a fast and efficient sensitivity map generation technique for real-time neuroimaging.
• To improve the computational efficiency of brain imaging using Finite Element Models (FEM).

Main Methods:

• Developed a novel sensitivity map generation method using Finite Element Models (FEM).
• Utilized a reduced sensitivity matrix, leveraging sparsity and parallelization.
• Evaluated time and memory efficiency against conventional methods across various mesh densities (50,000 to 320,000 nodes).

Main Results:

• Achieved over a tenfold reduction in required memory.
• Demonstrated a fourfold decrease in computational time.
• Enabled near real-time image recovery.

Conclusions:

• The proposed FEM-based approach offers significant computational advantages for brain imaging.
• This method facilitates faster and more efficient neuroimaging, paving the way for real-time applications.
• The technique is scalable and effective across a range of mesh densities.