Advanced compressed sensing photoacoustic imaging techniques using iteratively enhanced L1 regularization

Photoacoustic imaging (PAI) offers significant advantages but faces challenges in data processing and reconstruction. Sparse reconstruction techniques and compressed sensing theory have advanced its development. Regularization methods are commonly used to address the ill-posed problem of reconstructing PAI images from incomplete and noisy experimental data. While the L1-norm efficiently promotes sparsity in regularization methods, it often results in undesirable edge smoothing, which can degrade image quality. To address this limitation, this paper employs the Modified Iterative Reweighted L1 (MIRL1) regularization method, which aims to enhance sparsity while preserving critical image features such as edges. Experiments with simulated signals containing Gaussian noise show that MIRL1 outperforms traditional L1-norm-based SPGL1 and L0-norm-based SL0 algorithms in terms of image quality and noise robustness. MIRL1 enhances image clarity and maintains strong resistance to noise, providing superior image recovery results. This finding is significant for the development of sparse sampling PAI imaging algorithms.