首页> 外文期刊>International Journal of Chemistry and Chemical Engineering >Computer Aided Diagnosis (CAD) of Bright Lesion Detection in Fundus Images
【24h】

Computer Aided Diagnosis (CAD) of Bright Lesion Detection in Fundus Images

机译:基底图像中明亮病变检测的计算机辅助诊断(CAD)

获取原文
获取原文并翻译 | 示例
           

摘要

Diabetes mellitus will lead to vision loss that is the retina gets damaged. Often diabetic retinopathy has no early warning signs. For early detection and treatment the screening program will help a lot. Early symptoms of this disease are exudates, so early diagnosis and treatment at right time is very important to prevent blindness. In this paper our idea lies in the use of Block Variation of Local Correlation Coefficients (BVLC) and Block Difference of Inverse Probabilities (BDIP) texture features to characterize detected lesions using Active Contour technique (ACT). Then, Support Vector Machine (SVM) classifier is utilized to classify the detected lesions and the accuracy obtained is about 96.6%, Mathew correlation coefficient is about 0.972 and fisher score is about 0.9625.From these techniques we can reduce false positives for the detection of bright lesions in Fundus images.
机译:糖尿病将导致视力丧失,即视网膜受损。 糖尿病视网膜病常无期没有预警标志。 对于早期检测和治疗,筛查计划将有所帮助。 这种疾病的早期症状是渗出物,因此在正确的时间早期诊断和治疗非常重要,无法防止失明。 在本文中,我们的想法在于使用局部相关系数(BVLC)的块变化和横向概率(BDIP)纹理特征的块差异,以使用主动轮廓技术(ACT)表征检测到的病变。 然后,使用支持向量机(SVM)分类器来分类检测到的病变,获得的精度约为96.6%,Mathew相关系数约为0.972,Fisher分数约为0.9625.从这些技术中可以减少检测的误报 眼底图像中的明亮病变。

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