abstract
- © Springer International Publishing Switzerland 2015.In this paper, three segmentation quality evaluation methods are tested and analyzed. The purpose of this study is to test the reliability of the three evaluation methods (Target Overlap, F-measure and Jaccard Index) when used to assess the segmentation quality for the purpose of image segmentation in the Magnetic Resonance guided High Intensity Focused Ultrasound (MRgHIFU) therapy treatment planning. After the analysis of the evaluation methods, F-measure is considered as the best option because it takes into account relevant parameters for segmentation quality evaluation such as recall, precision, pixels detected correctly and incorrectly and the weighting factor for recall and precision. For the evaluation of a single region, Jaccard Index should be used in order to confirm that the region was correctly segmented, so the accuracy of the segmentation process will be more reliable.