Local Connectedness Constraint and Contrast Normalization Based Microaneurysm Detection
Diabetic retinopathy is an eye disease that damages the retina and leads to vision loss.As the first sign of diabetic retinopathy,microaneurysm(MA)usually appears as a round small spot,which seems similar to incontinuous vessels and background noise dots.It is hard for the individual algorithms published so far to distinguish them.In this paper,we model the MA detection problem as finding the interest spots from an image.Based on the characteristics of the target,connectedness rules are set up to constrain the MA inside a window.Hemorrhages and noise points are removed that way.Local contrast normalization is also used to better classify MAs from dots within vessels.Through comprehensive experiments on DIARETDB1 dataset and ROC dataset,we show that dots within vessels and noise points in the background can be well removed.Our method outperforms others with high sensitivity and specificity.
Microaneurysm detection Local connectedness constraint Region growing Local contrast normalization
Mengxue Liu Qi Yu Jie Yang Yu Qiao Xun Xu
Institute of Image Processing and Pattern Recognition,Shanghai Jiao Tong University,Shanghai,China Shanghai General Hospital,Shanghai Jiao Tong University,Shanghai,China
国际会议
第七届全国模式识别学术会议(The 7th Chinese Conference on Pattern Recognition,CCPR2016)
成都
英文
392-403
2016-11-03（万方平台首次上网日期，不代表论文的发表时间）