Abstract: Multi-dimensional image processing techniques are more helpful for medical image diagnosis process. In this research work brain MRI or CT scan images are collecting for input and attain the preprocessed image. Depending on scalability and integration on medical images we get multi model diagnosed image. The 2 scale multi dimensional sparse representation is applied on selected image for acquiring the original and de-noise image. This method improves the performance metrics such as accuracy 98.45%, recall 98.98% and throughput 97.58% have been improved.