A novel approach to improve execution time performance of Medical image processing

This paper presents a novel approach to enhance the performance of Medical image processing, primarily using code optimization techniques. The goal is to work on medical imaging which is both time and life critical. The problem of performance improvement is tackled from multiple directions - architecture, algorithm, compiler and the code level, the approach being unique in th is work. Image acceleration techniques are useful for faster diagnosis and treatment of medical issues, which are experimented in this work. Performance of image processing algorithms is measured with varying parameters such as different operating systems, different processors, different image sizes and various compiler optimization techniques like loop fusion, constant folding etc. It is found that code optimization and certain compiler optimizations improve the performance.