首页> 外文会议>International Conference on Information Technology and Electrical Engineering >Segmentation of optic disc and optic cup in colour fundus images based on morphological reconstruction
【24h】

Segmentation of optic disc and optic cup in colour fundus images based on morphological reconstruction

机译:基于形态重建的彩色眼底图像中视盘和视杯的分割

获取原文

摘要

Glaucoma disease is one of the retinal abnormalities caused by size changing of optic disc and optic cup. The disease is commonly diagnosed by quantifying vertical cup-to-disc ratio (CDR). Therefore, this study proposes two schemes for segmenting optic disc and optic cup based on morphological reconstruction. Retinal vessel removal is considered during the process. The proposed schemes are evaluated on 50 colour fundus images collected from Drishti-GS dataset. The final segmentation results are compared to the available ground truth image. Segmentation of optic disc achieves the level of accuracy, sensitivity and specificity at 96.3%, 83.12% and 99.37%, respectively. The evaluation for optic cup segmentation obtains the accuracy of 99.13%, sensitivity of 80.68% and specificity of 99.53%. These successful results in segmenting optic disc and optic cup indicate that the proposed schemes have a potential to be implemented as part of the development of computerised glaucoma diagnosis systems.
机译:青光眼疾病是由视盘和视杯的大小变化引起的视网膜异常之一。通常通过量化垂直杯碟比(CDR)来诊断该疾病。因此,本研究提出了基于形态学重建的两种分割视盘和视杯的方案。在该过程中考虑了视网膜血管的去除。在从Drishti-GS数据集中收集的50幅彩色眼底图像上对提出的方案进行了评估。将最终的分割结果与可用的地面真实图像进行比较。视盘的分割分别达到96.3 \%,83.12 \%和99.37 \%的准确性,敏感性和特异性水平。光学杯分割的评估获得了99.13%的准确度,80.68%的灵敏度和99.53%的特异性。这些成功的分割视盘和视杯的结果表明,提出的方案有潜力作为计算机化青光眼诊断系统开发的一部分来实施。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号