Change detection of medical images using dictionary learning techniques and principal component analysis

Abstract

Abstract. Automatic change detection methods for identifying the changes of serial MR images taken at different times are of great interest to radiologists. The majority of existing change detection methods in medical imaging, and those of brain images in particular, include many preprocessing steps and rely mostly on statistical analysis of magnetic resonance imaging (MRI) scans. Although most methods utilize registration software, tissue classification remains a difficult and overwhelming task. Recently, dictionary learning techniques are being used in many areas of image processing, such as image surveillance, face recognition, remote sensing, and medical imaging. We present an improved version of the EigenBlockCD algorithm, named the EigenBlockCD-2. The EigenBlockCD-2 algorithm performs an initial global registration and identifies the changes between serial MR images of the brain. Blocks of pixels from a baseline scan are used to train local dictionaries to detect changes in the follow-up scan. We use PCA to reduce the dimensionality of the local dictionaries and the redundancy of data. Choosing the appropriate distance measure significantly affects the performance of our algorithm. We examine the differences between L 1 and L 2 norms as two possible similarity measures in the improved EigenBlockCD-2 algorithm. We show the advantages of the L 2 norm over the L 1 norm both theoretically and numerically. We also demonstrate the performance of the new EigenBlockCD-2 algorithm for detecting changes of MR images and compare our results with those provided in the recent literature. Experimental results with both simulated and real MRI scans show that our improved EigenBlockCD-2 algorithm outperforms the previous methods. It detects clinical changes while ignoring the changes due to the patient's position and other acquisition artifacts.

Change Detection of Medical Images for Three Dimensional Volumetric Data

“…We extend our 2D change detection algorithms [17][18][19][20] to 3D volumetric data. The 3D change detection algorithms are named as 3D EigenBlockCD-2 and 3D AEDL-2.…”

“…After the co-registration step, the tenth slice from the test volume and its corresponding tenth transformed slice from the reference volume are aligned as shown in Figure 4a and 4c, respectively. The algorithms used in 2D [17][18][19][20] can be easily extended for 3D volumes which we present below.…”

“…In previous publications, we extended primary work in the area of background modeling and dictionary learning techniques and developed three variant methods [17][18][19][20]. Our first algorithm, (Adaptive EigenBlock Dictionary Learning (AEDL)) [17] , performs image registration locally to capture the local spatial changes in the test image using a series of local sparse minimization processes and the knowledge of compressive sensing.…”

“…Both, the AEDL and the EigenBlockCD algorithms ignore insignificant changes due to patient's position, noise, and other acquisition related artifacts, when these unrelated disease changes are within a small neighbourhood. To account for large sizes of insignificant changes, i.e., large shifts and rotations, we designed second versions of the two algorithms, the AEDL-2 and EigenBlockCD-2 by adding a co-registration step at the beginning [19,20].…”

Assessment of Data Fusion Algorithms for Earth Observation Change Detection Processes

“…Determining changed areas in images of the same scene taken at different points in time is of considerable interest given that it offers a large number of applications in various disciplines [2], including: video-surveillance [3], medical diagnoses and treatment [4], vehicle driving support [5] and remote detection.…”