Dimensionality Reduction Based Optimization Algorithm for Sparse 3-D Image Reconstruction in Diffuse Optical Tomography

• DOI: 10.1038/srep22242
• Publication Date: 2016-03-04T10:20:22Z
• Authors: Tanmoy Bhowmik; Hanli Liu; Zhou Ye; Soontorn Oraintara
• Keywords: Diffuse optical imaging; Regularization

Diffuse optical tomography (DOT) is a relatively low cost and portable imaging modality for reconstruction of properties in highly scattering medium, such as human tissue. The inverse problem DOT ill-posed, making high-quality image critical challenge. Because the nature sparsity DOT, regularization has been utilized to achieve reconstruction. However, conventional approaches using sparse optimization are computationally expensive have no selection criteria optimize parameter. In this paper, novel algorithm, Dimensionality Reduction based Optimization (DRO-DOT), proposed. It reduces dimensionality by reducing number unknowns two steps thereby makes overall process fast. First, it constructs resolution voxel basis on sensing-matrix find an support. Second, reconstructs inside To compensate reduced sensitivity with increasing depth, depth compensation incorporated DRO-DOT. An efficient method optimally select parameter proposed obtaining image. DRO-DOT also able reconstruct high-resolution images even limited optodes spatially set-up.