标题：Corrigendum to “Hierarchical retinal blood vessel segmentation based on feature and ensemble learning” (Neurocomputing (2015) 149(Part PB) (708–717)(S0925231214010169)(10.1016/j.neucom.2014.07.059))
作者：Wang S.; Yin Y.; Cao G.; Wei B.; Zheng Y.; Yang G.
作者机构：[Wang, S] School of Computer Science and Technology, Shandong University, Jinan, 250101, China;[ Yin, Y] School of Computer Science and Technology, Sh 更多
通讯作者地址：[Yin, Y] School of Computer Science and Technology, Shandong UniversityChina;
摘要：The authors regret “a programming error exists in their experiments which invalidates the conclusions of their paper. After modification, the performance of the proposed method is the third best on the DRIVE database in terms of accuracy, and remains the best on the STARE database. The detail of the programming error is as follows: When computing Accuracy, they tried to assess whether their result OR and the ground truth GT were the same by this equation: count=sum(sum((OR-GT)==0)), they failed to notice that OR and GT that they used were unsigned integers, the negative values of (OR-GT) was automatically reassigned to 0, so all the accuracies in their paper are inaccurate, below are the list of changes (Tables 3, 4, 6,7): They also declare a typo error of SP of Drive (trained on STARE) in Table 10, the value is changed to 0.9574 from 0.9774. The main text and Tables 5, 8, 9, 11, in which the accuracy values referred to the above mentioned Tables must also be updated: Section 1, paragraph 3: Experimental results show that our approach is competitive with state-of-the-art by achieving sensitivity/specificity/accuracy/AUC values of 0.8173/0.9733/0.9533/0.9475 for the DRIVE database and of 0.8104/0.9791/0.9621/0.9751 for the STARE database. Section 5.4, paragraph 2: The best case accuracy, sensitivity, specificity, AUC for the DRIVE database are 0.976359, 0.903219, 0.986776, 0.960497, respectively, and the worst case measures are 0.921903, 0.667402, 0.955739 and 0.938519, respectively. The best case vessel segmentation result for the STARE database has an accuracy of 0.989061; sensitivity, specificity, AUC are 0.94471, 0.99432, and 0.988388, respectively. The worst case accuracy is 0.906358; sensitivity, specificity, and AUC are 0.548304,0.948144, and 0.962269, respectively”. The authors would like to apologise for any inconvenience caused. Yilong Yin is the Director of MLA Lab and a Professor of the Shandong University. He received his pH.D. degree in 2000 from the Jilin University. From 2000 to 2002, he worked as a post-doctoral fellow in the Department of Electronic Science and Engineering, Nanjing University. His research interests include machine learning, data mining, and biometrics. © 2016 Elsevier B.V.