ABSTRACT
This paper presents a new class of image noise smoothing algorithms utilizing the membership information of the neighboring pixels. The basic idea of this method is to compute the smoothed output using neighboring pixels from the same cluster to avoid image blurring. A fuzzy c-means algorithm is first applied to the image to separate the image pixels into a certain number of clusters. A membership function is defined as the probability that a pixel belongs to a cluster. The proposed method uses this membership function as a weight to calculate the weighted sum of the pixel values from its neighboring pixels. The results of the application of this algorithm to various images show that it can smooth images with edge enhancement. The smoothness of the resultant images can be controlled by the cluster number and window size.
Subject(s)
Algorithms , Artifacts , Fuzzy Logic , Image Enhancement/methods , Diagnostic Imaging , Humans , Information Storage and Retrieval/methods , Pattern Recognition, Automated/methods , Phantoms, Imaging , Signal Processing, Computer-AssistedABSTRACT
The maximum likelihood expectation maximization (MLEM) algorithm has several advantages over the conventional filtered back-projection (FBP) for image reconstruction. However, the slow convergence and the high computational cost for its practical implementation have limited its clinical applications. This study proposes the incorporation of a thresholding technique in both the MLEM and ordered subsets EM (OSEM) algorithm to accelerate convergence. The threshold is set to c*m, where m is the mean pixel value of the whole image. The reconstruction time is proportional to the total number of pixels, so a thresholding technique that nullifies the value of a pixel if it falls below a threshold, can effectively remove the non-active pixels and substantially accelerate reconstruction. Preliminary tests on simulated PET data reveal that the thresholding technique accelerates the convergence rate and reduce error in the reconstructed image. The reconstruction performance improves with the increase of the threshold level and the MSE reaches minimum for c value equals to about 1.