Comparison of two- and three-dimensional reconstruction methods in optical tomography

Area of Science:

• Biomedical optics
• Image reconstruction
• Optical imaging

Background:

• Optical near-infrared imaging is crucial for non-invasively probing biological tissues.
• Reconstructing images from scattering media presents significant computational challenges.
• Developing efficient and accurate 3D reconstruction algorithms is essential for clinical applications.

Purpose of the Study:

• To present a novel three-dimensional (3D) image reconstruction scheme for optical near-infrared imaging.
• To evaluate the performance of this scheme in highly scattering materials.
• To investigate the impact of using simplified two-dimensional (2D) models for 3D data reconstruction.

Main Methods:

• Developed a 3D image reconstruction algorithm for optical parameters from transillumination measurements.
• Tested the algorithm on a cylindrical phantom with embedded absorbing perturbations.
• Analyzed the effect of mismatched 2D reconstruction models on 3D tomographic data.

Main Results:

• The 3D reconstruction scheme accurately maps optical parameters within scattering volumes.
• Using 2D models for 3D data introduces artifacts, varying with measurement type.
• Mean time-of-flight data showed better agreement between 2D and 3D models than dc intensity data.
• Artifacts from model mismatch were substantially reduced by utilizing mean time data.

Conclusions:

• The proposed 3D reconstruction method offers a viable approach for optical near-infrared imaging.
• Employing mean time-of-flight data is critical for mitigating artifacts when using simplified 2D reconstruction models.
• This finding enhances the practical applicability of faster, less memory-intensive 2D models in 3D imaging scenarios.