Dictionary Learning Method Based on K‐Sparse Approximation and Orthogonal Procrustes Analysis for Reconstruction in Bioluminescence Tomography
Linzhi Su,Limin Chen,Wenlong Tang,Huimin Gao,Yi Chen,Chengyi Gao,Huangjian Yi,Xin Cao
DOI: https://doi.org/10.1002/jbio.202400308
IF: 3.3899
2024-10-09
Journal of Biophotonics
Abstract:This study takes the system matrix A and the surface photon flux Φ as inputs, proposing a dictionary learning method based on K‐sparse approximation and Orthogonal Procrustes analysis (KSAOPA) for sparse reconstruction in bioluminescence tomography (BLT). The proposed method has been shown to excel in terms of localization accuracy, morphological recovery, and in vivo applicability. Bioluminescence tomography (BLT) is one kind of noninvasive optical molecular imaging technology, widely used to study molecular activities and disease progression inside live animals. By combining the optical propagation model and inversion algorithm, BLT enables three‐dimensional imaging and quantitative analysis of light sources within organisms. However, challenges like light scattering and absorption in tissues, and the complexity of biological structures, significantly impact the accuracy of BLT reconstructions. Here, we propose a dictionary learning method based on K‐sparse approximation and Orthogonal Procrustes analysis (KSAOPA). KSAOPA uses an iterative alternating optimization strategy, enhancing solution sparsity with k‐coefficients Lipschitzian mappings for sparsity(K‐LIMAPS) in the sparse coding stage, and reducing errors with Orthogonal Procrustes analysis in the dictionary update stage, leading to stable and precise reconstructions. We assessed the method performance through simulations and in vivo experiments, which showed that KSAOPA excels in localization accuracy, morphological recovery, and in vivo applicability compared to other methods.
optics,biochemical research methods,biophysics