Practical fully three-dimensional reconstruction algorithms for diffuse optical tomography.

Area of Science:

• Biomedical Optics
• Medical Imaging
• Computational Science

Background:

• Diffuse optical tomography (DOT) is a powerful imaging technique but faces challenges due to its ill-posed nature.
• Accurate 3D reconstruction in DOT is computationally intensive and prone to artifacts.

Purpose of the Study:

• To develop an efficient and accurate fully three-dimensional (3D) reconstruction algorithm for diffuse optical tomography (DOT).
• To overcome the ill-posed nature of 3D DOT and reduce computational costs associated with iterative reconstruction.

Main Methods:

• Implemented a pseudodynamic (PD) approach integrating an ordinary differential equation to bypass explicit matrix inversion.
• Utilized the adjoint-Broyden update formula for low-rank Jacobian updates, avoiding costly reevaluations.
• Integrated calculations along the quadratic path from the perturbation equation with the Hessian for further efficiency.

Main Results:

• PD reconstructions demonstrated superior absorption coefficient maps with reduced artifacts and improved quantitative accuracy and contrast recovery.
• The Jacobian update formula significantly reduced the scaling of computation time with measurement set dimension.
• Increased data dimension improved reconstruction but required careful distribution of measurement points for optimal contrast recovery.

Conclusions:

• The developed PD algorithm offers an efficient and artifact-free solution for 3D DOT reconstruction.
• Computational efficiency is enhanced through Jacobian updates and Hessian-based integration.
• Optimal contrast recovery necessitates not only sufficient data but also strategic placement of measurement points.