Image reconstruction using interval simulated annealing in electrical impedance tomography.

Area of Science:

• Medical Imaging
• Computational Electromagnetics

Background:

• Electrical impedance tomography (EIT) is an imaging modality that reconstructs internal electrical impedance distributions.
• EIT reconstruction is a nonlinear, nonconvex optimization problem, often ill-posed and challenging for derivative-based methods.
• Standard simulated annealing (SA) can solve EIT reconstruction but incurs high computational costs due to full objective function evaluations.

Purpose of the Study:

• To propose a computationally efficient variation of simulated annealing for EIT reconstruction.
• To reduce the computational burden of EIT imaging while maintaining solution quality.

Main Methods:

• A modified simulated annealing (SA) algorithm is introduced for EIT reconstruction.
• The proposed method involves partial evaluation of the objective function at each iteration.
• Guarantees on the algorithm's behavior are maintained despite partial evaluations.

Main Results:

• The modified SA algorithm significantly reduces the computational cost of EIT reconstruction.
• Partial objective function evaluation leads to faster convergence without compromising accuracy.
• The approach addresses the ill-posed nature of EIT reconstruction problems.

Conclusions:

• The proposed partial evaluation SA method offers a computationally efficient alternative for EIT reconstruction.
• This advancement can accelerate the application of EIT in various fields.
• Further research can explore adaptive strategies for partial evaluation in EIT.