Search

The aim of this paper is to present an image compression method using feedforward backpropagation neural networks. Medical imaging is an efficient source for better diagnosis of the disease and also helps in assessing the severity of the disease. But due to the increasing size of the medical images, transferring and storage of images require huge bandwidth and storage space. Therefore, it is essential to derive an effective compression algorithm, which have minimal loss, time complexity, and increased reduction in size. With the concept of Neural Network, data compression can be achieved by producing an internal data representation. The training algorithm and development architecture gives less distortion and considerable compression ratio and also keeps up the capability of hypothesizing and is becoming important. The performance metrics of three algorithms, Levenberg Marquardt algorithm, Resilient backpropagation algorithm, and Gradient Decent algorithm have been computed on Magnetic Resonance Imaging (MRI) images and it is observed that Levenberg Marquardt algorithm is more accurate when compared to the other two algorithms. [ABSTRACT FROM AUTHOR]

In this paper, the construction of new lifting based wavelets by a new method of calculating lifting coefficients is presented. First of all, new basis functions are utilized to ease new orthogonal traditional wavelets. Then by using the decomposing poly-phase matrix the lifting steps are calculated using a simplified method. The interesting feature of lifting scheme is that the construction of wavelet is derived in spatial domain only; hence the difficulty in the design of traditional wavelets is avoided. Lifting scheme was used to generate second generation wavelets which are not necessarily translation and dilation of one particular function. Short and sharp basis functions are chosen so as to obtain the non-uniform nature of usual image classes. Implemented wavelets are applied on a number of medical images. It was found that the compression ratio (CR) and Peak Signal to Noise Ratio (PSNR) are far ahead of that are obtained with the popular traditional wavelets as well as the successful 5/3 and 9/7 lifting based wavelets. Set Partitioning in Hierarchical Trees (SPIHT) is used to incorporate compression. [ABSTRACT FROM AUTHOR]

For perfect diagnosis of brain tumour, it is necessary to identify tumour affected regions in the brain in Magnetic Resonance Imaging (MRI) images effectively and compression of these images for transmission over a communication channel at high speed with better visual quality to the experts. An attempt has been made in this paper for identifying tumour regions with optimal thresholds which are optimized with the proposed Hybrid Bacteria Foraging Optimization Algorithm (BFOA) and Particle Swarm Optimization (PSO) named (HBFOA-PSO) by maximizing the Renyi's entropy and Kapur's entropy. BFOA may be trapped into local optimal problem and delay in execution time (convergence time) because of random chemotaxis steps in the procedure of algorithm and to get global solution, a theory of swarming is commenced in the structure of HBFOA-PSO. Effectiveness of this HBFOA-PSO is evaluated on six different MRI images of brain with tumours and proved to be better in Peak Signal to Noise Ratio (PSNR), Mean Square Error (MSE) and Fitness Function. [ABSTRACT FROM AUTHOR]