In this article, we analyze the performance of artificial neural network, in classification of medical images using wavelets as feature extractor. This work classifies the mammographic image, MRI images, CT images, and ultrasound images as either normal or abnormal. We have tested the proposed approach using 50 mammogram images (13 normal and 37 abnormal), 24 MRI brain images (9 normal and 15 abnormal), 33 CT images (11 normal and 22 abnormal), and 20 ultrasound images (6 normal and 14 abnormal). Four kind of neural network models such as BPN (Back Propagation Network), Hopfield, RBF (Radial Basis Function), and PNN (Probabilistic neural network) were chosen for study. To improve diagnostic accuracy, the feature extracted using wavelets such as Harr, Daubechies (db2, db4, and db8), Biorthogonal and Coiflet wavelets are given as input to the neural network models. Good classification percentage of 96% was achieved using the RBF when Daubechies (db4) wavelet based feature extraction was used. We observed that the classification rate is almost high under the RBF neural network for all the dataset considered.
This article presents the detailed analysis of the local pixel grouping-principle component analysis (LPG-PCA) algorithm in denoising and deblurring of medical images. Inefficient diagnosis of the medical images containing lot of information is often affected by the noise and artifacts. In order to remove these noises and artifacts, a statistical decorrelation technique, LPG-PCA is used which is found to be one of the efficient methods, which could be used in improving the performance of medical images. For better preservation of local structures of the image, a pixel and its nearest neighbors are modeled as a vector variable, which leads to the selection of similar intensity characteristics. Denoising method used in this article is done in two stages for improving the denoising performance. The smoothening caused by the denoising process is removed by using LPG-PCA along with adaptive sparse domain representations in the deblurring process. This involves clustering of data and finding the subdictionary of each cluster using LPG-PCA. Experimental results show that an average improvement of 2.9 and 5.1 dB is found in the computed tomography and magnetic resonance imaging images using denoising and deblurring process.
This paper presents the performance analysis of the LPG-PCA algorithm in deblurring of medical images. Medical images containing lot of information which are often affected by noise and artifacts, which leads to the inefficient diagnosis. LPG-PCA which is a statistical decorrelation technique is found to be one of the efficient methods which could be used in improving the performance of medical images. For better preservation of fine structures in an image, a pixel and its nearest neighbors are modeled as a vector variable whose training samples are selected using a moving window in the image. Such a local vector variable preservation leads to the selection of the similar intensity characteristics. This property of LPG-PCA technique is applied in image deblurring process using adaptive sparse domain regularization technique. This method involves clustering of data and finding the Sub dictionary of each cluster using LPG-PCA. Then the dictionary for input patch is selected using SVD technique and deblurring is done using regularization. Performance analysis of this technique is found using various image quality measures and results are found to be efficient than other conventional methods.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.