Neural Network Based Image Compression Approach to Improve the Quality of Biomedical Image for Telemedicine

Aims: In Telemedicine, the use of digital utilization for medical diagnosis helps medical practitioners for better and fast treatment of patients, but at the same time it increase the storage resource requirement for archive the images as they are in high resolution and size. To minimize the size it must be compressed before transmission and stored. On the other hand, the compression will reduce the image affinity, particularly when the images are compressed at lower bit rates. The reconstructed images endure from overcrowding artifacts and the image quality will be severely besmirched under the circumstance of high compression ratios.
Methodology: To meet these defy, numerous amalgamated compression algorithms solely for medical imaging are developed in the recent years. But a need of accurate technique/s is highly essential to avoid any lethal results. Artificial intelligence (AI) techniques are highly accurate and hence preferred for automated image classification, segmentation and compression.
Conclusion: To accomplish the goal of performance enhancement with respect to compression ratios and deciphered picture quality an algorithm is developed using advance Artificial Neural Networks (ANN) for image compression (IC) and the results are compared with existing image compression techniques viz JPEG2000. This work is explored in the context of Magnetic Resonance (MR) image classification, segmentation and compression.