![]() | Only 14 pages are availabe for public view |
Abstract Retrieving and managing Content-Based Medical Images Retrieval (CBMIR) are considered more important now, especially with increasing in medical imaging and expanding the medical image database. Also, these systems allow for the benefit of medical images in having a better grasp on and deeper insights into the causes and treatment of different diseases, not only for diagnostic purposes. CBMIR, therefore, played an important role in the field of image processing and the extraction of low-level features such as color histograms, edges, texture, and shape, as well as similarity measures for comparison and retrieval of medical images. The majority of the methods already used in CBMIR enhance the retrieval of a medical image and disease diagnosis by reducing the issue of the semantic gap between low visual and high semantic levels. To overcome these problems, there is a critical need for an efficient and accurate content-based medical image retrieval method. This thesis proposes an efficient method of Retrieval based on Query Expansion (RbQE) for the retrieval of Computed Tomography (CT), Magnetic Resonance Imaging (MRI), and histopathological images. RbQE is based on expanding the features of querying and exploiting the pre-trained learning models AlexNet, VGGNets, and ResNets to extract compact, deep, and high-level features from medical images. There are two searching procedures in RbQE: a rapid search and a final search. In the rapid search, the original query is expanded by retrieving the top-ranked images from each class and is used to reformulate the query by calculating the mean values for deep i ii features of the top-ranked images, resulting in a new query for each class. In the final search, the new query that is most similar to the original query will be used for retrieval from the database. The performance of the proposed method has been compared to state-of-the-art methods on five publicly available standard datasets, namely, TCIA-CT, EXACT09-CT, NEMA-CT, OASIS-MRI, and KIMIA Path960. Experimental results show that the proposed method exceeds the compared methods by 0.84%, 4.86%, 1.24%, 14.34%, and 1.96% in average retrieval precision (ARP) on the retrieval of the top ten, for TCIA-CT, EXACT09-CT, NEMA-CT, OASIS-MRI, and KIMIA Path960, respectively. But for KIMIA Path960, the ARPs for the top five, fifteen, and twenty exceed the compared methods by 2.14%, 6.74%, and 9.42%, respectively. |