Full-Wave Image Reconstruction in Transcranial Photoacoustic Computed Tomography Using a Finite Element Method
IEEE Transactions on Medical Imaging
|September 9, 2024
Summary
This study introduces an efficient imaging model for transcranial photoacoustic computed tomography, significantly reducing computational demands for clearer human brain imaging through skull aberration correction. The new method enhances speed and accuracy for whole-brain simulations.
Area of Science:
- Biomedical Imaging
- Computational Physics
- Acoustics
Background:
- Transcranial photoacoustic computed tomography (TPCT) faces challenges due to skull-induced acoustic aberrations.
- Current full-wave reconstruction methods are computationally intensive, using a unified elastic wave equation for skull propagation.
Purpose of the Study:
- To develop an efficient discrete imaging model for TPCT.
- To reduce computational resources required for human brain imaging via TPCT.
Main Methods:
- Proposed a finite element discretization model applying the elastic wave equation only to the skull (solid) and the acoustic wave equation to soft tissues (liquid).
- Explicitly modeled solid-liquid interfaces with elastic-acoustic coupling.
- Utilized coarser, irregular meshes for geometric conformity, reducing system size by 20 times.
- Derived a matched forward-adjoint operator pair for optimization algorithms.
Main Results:
- Achieved a 20-fold reduction in linear system size, enabling faster and more accurate whole-brain simulations.
- Validated the reconstruction framework through numerical simulations and phantom experiments.
Conclusions:
- The proposed discrete imaging model significantly enhances the efficiency and speed of TPCT for human brain imaging.
- This approach offers a computationally feasible solution for overcoming skull-induced acoustic aberrations in TPCT.
