Retrieval of 3D medical images via their texture features

While content-based image retrieval has been researched for more than two decades, retrieving 3D datasets has been progressing considerably slower, especially in the application to the medical domain. This is in part due to the limitation of processing speed while trying to retrieve high-resolution datasets in real-time. Another barrier is that most existing methods have been developed based on 2D images instead of 3D, leaving a gap to be filled. At present, a significant number of exploitations are focusing on the extraction of 3D shapes. As it happens, it appears that, to a large extent, the remaining information tends to be equally important in the task of clinical decision making. With this in mind, in this paper, a texture-based online system, MIRAGE, has been developed to facilitate CBIR for 3D images. Specifically, four texture-based approaches stemming from 2D forms are studied extensively through the application to 3D images using a collection of MR brain images and are implemented, which include 3D Local Binary Pattern (LBP), 3D Grey Level Co-occurrence Matrices (GLCM), 3D Wavelet Transforms (WT) and 3D Gabor Transforms (GT). Based on the nature of the content, each approach has its own advantages and disadvantages. For example, in terms of retrieval precision of tumours and processing speed, LBP not only achieves precision rate of up to 78% but also can perform retrieval in real time with sub-second processing speeds, outperforming the others