Abstract:Objective To study the automatic segmentation method of liver magnetic resonance imaging (MRI) sequence image
based on Chan Vese (CV) model. Methods Based on the traditional CV model, the energy functional of CV model was improved,
and the Dirac function was replaced by a new edge indicator function to optimize the parameters of CV model, so as to improved
the segmentation accuracy and speed of CV model. The results of image segmentation were quantitatively evaluated by Jaccard
coefficient and Dice coefficient, and the data were analyzed in this paper. Results The improved CV model had better anti noise
performance, which was suitable for image segmentation of disturbed image and complex scene image segmentation. Moreover, the
improved CV model segmentation time and iteration times, Jaccard coefficient and Dice coefficient were lower than the traditional
CV model, the difference between the two models was statistically significant (P<0.001). Conclusion The improved CV model
algorithm has a significant effect on liver MRI image segmentation. The automatic segmentation method of liver MRI sequence
image based on CV model can extract the target quickly and accurately, which is ideal in medical image segmentation.
高倩倩,孙世春. 基于CV模型的肝脏核磁共振序列图像自动分割方法[J]. 中国医疗设备, 2020, 35(8): 64-66.
GAO Qianqian, SUN Shichun. Automatic Segmentation of Liver MRI Sequence Image Based on CV Model. China Medical Devices, 2020, 35(8): 64-66.