Clustering-based compression connected to cloud databases in telemedicine and long-term care applications

Abstract

A novel clustering-based compression work connected to cloud databases is proposed for the applications of telemedicine and long-term care in this study, where the goal is to enhance information transfer rate and storage capacity to further improve communication between medical staffs and patients in long-term care and telemedicine. The proposed system mainly involves three-dimensional histogram competitive Hopfield neural network (CHNN) clustering, regionalization, and modified block truncation coding (BTC). Three-dimensional histogram CHNN clustering and regionalization are proposed to achieve better clustering accuracy within three-dimensional spaces and simultaneously overcome the problems of fluctuating initial values of clustering. Modified BTC is also proposed to analyze clustering regions with different compression rates according to their importance in order to greatly preserve important image feature information under the condition of smaller image sizes. The experimental results indicate that the proposed system is adaptive and performs better than several previous methods. It is also suggested being suitable for the applications of telemedicine and long-term care connected to cloud databases.