Three-dimensional Bayesian optical diffusion tomography with experimental data.

Area of Science:

• Biomedical Optics
• Diffuse Optical Imaging
• Computational Imaging

Background:

• Accurate reconstruction of absorbers within scattering media is crucial for biomedical applications.
• Traditional methods often struggle with limited detection planes and complex scattering environments.
• Developing robust algorithms for three-dimensional (3D) imaging is essential for advancing diffuse optical imaging (DOI).

Purpose of the Study:

• To develop and validate a novel algorithm for reconstructing 3D absorbers in scattering media using frequency domain measurements.
• To assess the algorithm's performance with single source-detector plane measurements.
• To investigate the impact of multiple modulation frequencies on reconstruction accuracy.

Main Methods:

• Utilized frequency domain measurements of transmitted light in a single source-detector plane.
• Employed Bayesian regularization and iterative coordinate descent optimization for image reconstruction.
• Incorporated estimation of detector noise, source-detector coupling, and background diffusion coefficients.
• Explored the use of multiple modulation frequencies.

Main Results:

• Successfully reconstructed a three-dimensional absorber embedded in a scattering medium.
• Demonstrated the algorithm's utility and the necessity of a 3D model for accurate imaging.
• Showcased that out-of-plane scattering enables the recovery of 3D features from single-plane measurements.
• Validated the algorithm's ability to estimate noise, coupling, and diffusion parameters.

Conclusions:

• The developed algorithm effectively reconstructs 3D absorbers in scattering media from single-plane measurements.
• The study underscores the critical importance of employing three-dimensional models in diffuse optical imaging.
• Out-of-plane scattering provides valuable information for recovering volumetric features, enhancing imaging capabilities.