Impact of Reconstruction Algorithms on Diffuse Correlation Tomography Blood Flow Imaging

Abstract

Near-infrared diffuse correlation tomography (DCT) is an emerging technology for non-invasive imaging of the tissue blood flow. The flow imaging quality relies on the image reconstruction algorithm, which, however, is little studied thus far. In this study, we conducted the first investigation of reconstruction algorithm impact on DCT blood flow imaging. Two reconstruction algorithms, i.e., the finite element method (FEM) representing the imaging framework of partial differential equation, and the Nth-order linear (NL) approach, representing the imaging framework of integral equation that was recently proposed by us to incorporate the tissue morphological information, were compared. Both computer simulations and phantom experiment outcomes show that the NL approach performs much better in image accuracy and homogeneity over anomaly or background, when compared with the FEM at the same source-detector configuration and spatial resolution. This study demonstrates that the DCT blood flow imaging is substantially influenced by the reconstruction algorithm, thus it has great potential in future algorithm design and optimization.

Keywords

• Diffuse correlation tomography
• blood flow
• image reconstruction
• finite element method
• Nth-order linear approach