Geometric-sensitivity-difference Algorithm for Fluorescence Diffuse Optical Tomography Reconstruction

Abstract

Fluorescence Diffuse Optical Tomography is an emerging non-invasive imaging technique. Here we have introduced an alternative approach for Fluorescence Diffuse Optical Tomography which optimizes the data model fit based on pairing of the source and detector such that one source is shared by two detectors or one detector is shared by two sources, which in comparison to the conventional data model fit which does not use any kind of pairing. This new reconstruction algorithm is called Geometric Sensitivity Difference (GSD) method which effectively reduces the variation of reconstruction sensitivity with respect to imaging depth. Here the GSD method for source sharing detector pairs is demonstrated using simulated continuous-wave measurements in an outer circular-array imaging geometry, of which the native sensitivity varies strongly with respect to the depth. The results of the GSD method are compared to that of two methods: one is the conventional baseline method which utilizes the native sensitivity and the second method is depth compensation algorithm (DCA) which employs active depth adaptive scheme which counteracts the dependence of reconstruction sensitivity with respect to imaging depth.