FREQUENCY-DOMAIN NEAR-INFRARED PHOTO DIFFUSION IMAGING - INITIAL EVALUATION IN MULTITARGET TISSUELIKE PHANTOMS

In this paper, an initial evaluation of our finite element based frequ ency-domain image reconstruction algorithm is performed for experiment s where multiple millimeter-sized heterogeneities are embedded within a tissue-equivalent (optically) background medium having multicentimet er dimensions. The cases considered consist of several interesting geo metry and optical property contrast combinations including (i) two dif ferent-sized targets with the same contrast at three different separat ion distances; (ii) two different-sized targets with different contras ts at two different separation distances; and (iii) three targets with the same and different sizes and contrasts, respectively. The reconst ruction algorithm that has been used is an enhanced version of our ori ginally developed regularized least squares approach that now includes total variation minimization, dual meshing, and spatial low-pass filt ering. Quantitative measures of image quality including the size, loca tion, and shape of the embedded heterogeneities along with errors in t heir recovered optical property values are presented. The results show that multiple targets can be clearly detected for all combinations na tions of locations, sizes, and contrast levels considered, but the qua ntitative nature of this detection is influenced by these parameters. (C) 1998 American Association of Physicists in Medicine.