A Study on the Reconstruction of Bayesian Diffuses Optical Tomography |

ABSTRACT:

Following the assembly of a triple-modality SPECT-CT-OTsmall animal imaging system provided intrinsically co-registered projectiondata of all three submodalities and under the assumption and investigation ofdual-labeled probes consisting of both Huorophores and radionuclides, a novelmulti-modal reconstruction strategy is presented in this paper aimed at improvingfluorescence mediated tomography (FMT). The following reconstruction procedureis proposed: Firstly, standard X-ray CT image reconstruction Is performedemploying the FDK algorithm. Secondly, standard SPECT image reconstruction isperformed using OSEM. Thirdly, from the reconstructed CT volume data thesurface boundary of the imaged object is extracted for finite elementdefinition. Finally, the reconstructed SPECT data is used as a prioriinformation within a Bayesian reconstruction framework for optical (FMT)reconstruction. We provide results of this multi-modal approach using phantomexperimental data and illustrate that this strategy does suppress artifacts andfacilitates quantitative analysis for optical imaging studies.  We present acombined classification and reconstruction algorithm for diffuse opticaltomography (DOT). DOT is a nonlinear ill-posed inverse problem. Therefore, someregularization is needed. We present a mixture of Gaussians prior, whichregularizes the DOT reconstruction step. During each iteration, the parametersof a mixture model are estimated. These associate each reconstructed pixel withone of several classes based on the current estimate of the optical parameters.This classification is exploited to form a new prior distribution to regularizethe reconstruction step and update the optical parameters.