Spatially-Adaptive Temporal Smoothing for Reconstruction of
Dynamic and Gated Image Sequences

 In this project we propose a method for spatio-temporal reconstruction of dynamic or gated image sequences. In a method we proposed previously, temporal smoothing in a Karhunen-Loève (KL) transform domain was used prior to reconstruction to reduce the effect of noise. In this work we improve on one of our group’s previous techniques by making the temporal smoothing adapt spatially to local characteristics in the projection data. This improves the noise performance of the temporal smoothing, while significantly lessening the possibility of signal distortion.

 In the proposed method, spatial regions of the projection-data sequence having similar time characteristics are identified by an unsupervised k-means clustering algorithm. A different KL transformation is designed for each spatial region in projection space, adapting the smoothing to the local temporal behavior. Finally, images are reconstructed from the smoothed projections by using the expectation-maximization (EM) algorithm.

 Experimental computer simulation results are shown that demonstrate potential improvements in image quality obtained by this technique for dynamic and gated imaging applications in brain, lesion, and cardiac imaging.

For more details please refer to:  Spatially-Adaptive Temporal Smoothing for Reconstruction of Dynamic and Gated Image Sequences
IEEE MIC'00 October 15-20, 2000 Lyon, France

For more details send a request to: jovan [at] brankov [dot] com

Cardiac experiment

 Reconstructed images in cardiac experiment. The proposed KL/Clustering method sacrifices some noise performance compared with the KL method, in favor of improved accuracy of the motion.

Dynamic Brain Imaging

 Reconstructed images in brain experiment. Images are scale to compensates for decreasing radio tracer concentration in the blood. Proposed KL/Clustering method sacrifices some noise performance compared with the KL method, in favor of improved accuracy of the time behavior of brain parts. See the occipital cortex region.

Dynamic Lung Lesion Imaging

 Similar to brain experiment images are scale to compensates for decreasing radio tracer concentration in the blood. Ones more the KL/clustering method improved accuracy of the time behavior of the small lesion in the lung.