An innovative medical image retrieval is proposed to extract low-level image texture features. These low level texture features were extracted directly using Modified Discrete Cosine Transform (MDCT). MDCT coefficients represent dominant directions and gray level variations of the image. Our proposed method uses a hierarchical similarity measure for efficient medical image retrieval and also reduces the search space in a large image database. In an experiment using a database of 200 images, our method shows a higher performance in the retrieval. The retrieval image is the relevance between a query image and database image, the relevance similarity is ranked according to the closest similar measures computed by the Euclidean distance. The experimental results show that, using our MDCT approach, it is easy to identify main objects and reduce the influence of background in the image, and thus improve the performance of medical image retrieval. © 2010 Published by Elsevier Ltd.