On the optimal choice of the illumination function in photoacoustic tomography

Phuoc-Truong Huynh,Barbara Kaltenbacher

2024-11-11

Abstract:This work studies the inverse problem of photoacoustic tomography (more precisely, the acoustic subproblem) as the identification of a space-dependent source parameter. The model consists of a wave equation involving a time-fractional damping term to account for power law frequency dependence of the attenuation, as relevant in ultrasonics. We solve the inverse problem in a Bayesian framework using a Maximum A Posteriori (MAP) estimate, and for this purpose derive an explicit expression for the adjoint operator. On top of this, we consider optimization of the choice of the laser excitation function, which is the time-dependent part of the source in this model, to enhance the reconstruction result. The method employs the $A$-optimality criterion for Bayesian optimal experimental design with Gaussian prior and Gaussian noise. To efficiently approximate the cost functional, we introduce an approximation scheme based on projection onto finite-dimensional subspaces. Finally, we present numerical results that illustrate the theory.

Numerical Analysis,Optimization and Control

What problem does this paper attempt to address?

The main problem that this paper attempts to solve is the identification problem of the absorption density function \(a = a(x)\) in photoacoustic tomography (PAT), especially in the case of considering medium loss. Specifically, the author studied the wave equation model with a fractional - order damping term in the time domain, which can describe the frequency - dependent attenuation phenomenon during the acoustic wave propagation process. In order to improve the quality of the reconstruction results, the author also explored the optimization problem of the laser excitation function selection.

### Main contributions of the paper:
1. **Research on photoacoustic tomography problems in lossy media**:
- The author incorporated the absorption density \(a\) as a space - dependent factor into the source term of the partial differential equation, expanding the existing model.
- A Bayesian framework was adopted to solve the inverse problem and calculate the maximum a posteriori (MAP) estimate.
- By introducing an appropriate adjoint problem, the explicit expression of the adjoint state was derived, and different prior selections were compared in the discretization setting.

2. **Optimize experimental design**:
- In the Bayesian setting, considering the case where both the parameter and the observation space are infinite - dimensional, the A - optimality criterion was used to optimize the laser intensity function.
- A projection scheme was introduced to project the data - mismatch Hessian matrix onto a finite - dimensional subspace to efficiently approximate the cost function.
- The convergence properties of the projection scheme were analyzed, and it was proved that the solution of the approximate optimal experimental design (OED) problem converges to the solution of the original infinite - dimensional problem under certain conditions, demonstrating the stability and consistency of the method.

### Main technical details:
- **Fractional - order derivative**: The Caputo - Djrbashian - type time - fractional - order derivative was used to describe the memory effect and frequency - dependent attenuation in acoustic wave propagation.
- **Forward problem and adjoint problem**: The concept of weak solution was defined, and the well - posedness of the forward problem was proved. The weak form of the adjoint problem was derived through the integration - by - parts formula.
- **Bayesian inverse problem**: Assuming that the parameter \(a\) and the noise \(\eta\) both follow a Gaussian distribution, the expressions of the posterior covariance operator and the MAP estimate were derived.
- **Optimize the laser excitation function**: The A - optimality criterion was used to minimize the trace of the posterior covariance operator, and the optimization problem was approximately solved by the projection method.

### Numerical results:
- The author verified the validity of the theoretical analysis through numerical experiments and showed how the optimized laser excitation function improves the quality of the reconstruction results.

Overall, this paper introduced new mathematical models and optimization methods in the study of the inverse problem of photoacoustic tomography, providing a theoretical basis and technical support for improving image reconstruction quality.