A Parallel Approach For Determining Region–of–Interest Area of Digital Medical Images
The development of technology is triggered many research areas. One of them is medical area that digital medical images have vital importance. This development has brought some problems like storing clinical data and medical images. As a solution to this problem, Picture Archiving and Communication Systems (PACS) are being used generally. However, there are several approaches as well. Importance of each part of the digital medical image can be changeable. Region-of-interest (ROI) is being used to determine the important segment of digital medical images. Chan-Vese active contour method is one of the developed algorithm to determination of ROI. In this paper, Chan-Vese active contour method is parallelized based on task parallelism on high performance computing (HPC) is proposed to improve the ROI determination. Finally, a series of experiments is tested to specify the effectiveness of the proposed approach.
Keywords— Region–Of–Interest,Image Processing Algorithms, ParallelComputing, Chan–Vese Segmentation Model, Preprocessing.