An adaptive parameter selection strategy based on maximizing the probability of data for robust fluorescence...

Area of Science:

• Biomedical imaging
• Computational modeling
• Medical physics

Background:

• Fluorescent molecular tomography (FMT) faces ill-posed inverse problems.
• Existing regularization methods (L2, L1 norms) performance depends on parameter selection.
• Current parameter selection strategies lack universality and incur high computational costs.

Purpose of the Study:

• To develop a universally applicable adaptive parameter selection method for FMT.
• To improve the performance of FMT reconstruction algorithms.
• To address the limitations of classical parameter selection strategies.

Main Methods:

• Proposed a maximizing the probability of data (MPD) strategy for adaptive parameter selection.
• Utilized maximum a posteriori (MAP) and maximum likelihood (ML) estimations to model regularization parameters.
• Employed multiple iterative estimates to determine stable optimal regularization parameters.

Main Results:

• The MPD strategy successfully determined stable regularization parameters for both L2 and L1 norm-based algorithms.
• Achieved good reconstruction performance in numerical simulations and in vivo experiments.
• Demonstrated the universal applicability of the MPD strategy.

Conclusions:

• The MPD strategy offers a robust and universally applicable solution for regularization parameter selection in FMT.
• This method enhances the accuracy and reliability of FMT reconstruction.
• The findings have significant implications for advancing FMT applications in biomedical research.