Medical image compression has recently evolved as an area of research for progressive transmission. The distance based medical diagnosis, demands for high quality imaging at faster data transfer rate. As the information’s are highly informative, each pixel information defines a sample observation. Hence the coding in medical diagnosis need to be of higher accuracy than conventional image coding. In the approach of image coding multi spectral coding is developed as new coding approach to achieve the objective of higher visualization accuracy. With this observation in this paper a multi spectral coding using multi wavelet transformation is developed. The multi spectral coding is improved by a band selective approach using inter band correlation factor. The evaluation factors for such a coding technique are observed to be improved over conventional multi-spectral coding.