期刊文献+

基于视觉注意力增强CBAM-U-Net模型的视网膜血管分割 被引量:2

Retinal vascular segmentation based on visual attention enhanced CBAM-U-Net model
下载PDF
导出
摘要 针对现有视网膜血管分割方法存在着分割不足、抗干扰能力弱等问题,提出了一种基于视觉注意力增强模型的视网膜血管分割方法。首先,通过白平衡和滤波对视网膜眼底图像进行增强预处理;然后,在医学图像分割网络中引入视觉注意力模型,提高视网膜血管特征的显著性,利用构建的深度分割网络进行模型训练;最后,利用训练得到的模型进行视网膜血管的分割预测。该方法在DRIVE和STARE两种数据集上进行实验,得到的Dice分割系数分别为0.867和0.898,Jaccard分割系数分别为0.765和0.816,平均准确率分别为97.53%和98.25%,平均灵敏度分别为80.16%和86.12%,平均特异性分别为99.87%和99.03%,实验结果表明该方法能够有效提高视网膜血管的分割精度。
出处 《计算机应用研究》 CSCD 北大核心 2020年第S02期321-323,共3页 Application Research of Computers
  • 相关文献

参考文献6

二级参考文献103

  • 1International Diabetes Federation. IDF Diabetes Atlas[R]. 6thed, Brussels, Belgium: International Diabetes Federation, 2014.
  • 2Abramoff M D, Garvin M K, Sonka M. Retinal imaging andimage analysis[J]. IEEE Reviews in Biomedical Engineering,2010, 3: 169-208.
  • 3Winder R J, Morrow P J, McRitchie I N, et al. Algorithms fordigital image processing in diabetic retinopathy[J]. ComputerMedical Imaging and Graphics, 2009, 33: 608-622.
  • 4Kirbas C, Quek F A review of vessel extraction techniques andalgorithms[J]. ACM Computing Surveys, 2004, 36(2): 81-121.
  • 5Hoover A, Kouznetsova V, Goldbaum M. Locating blood vesselsin retinal images by piecewise threshold probing of amatched filter response[J]. IEEE Transactions on Medical Imaging,2000, 19(3): 203-210.
  • 6Staal J J, Abramoff M D, Niemeijer M, et al. Ridge based vesselsegmentation in color images of the retina[J]. IEEE Transactionson Medical Imaging, 2004, 23(4): 501-509.
  • 7Niemeijer M, Staal J J, van Ginneken B, et al. Comparativestudy of retinal vessel segmentation methods on a new publiclyavailable database[C] //Proceedings of SPIE. Bellingham: Societyof Photo-Optical Instrumentation Engineers Press, 2004,5370: 648-656.
  • 8Odstrcilik J, Kolar R, Budai A, et al. Retinal vessel segmentationby improved matched filtering: evaluation on a newhigh-resolution fundus image database[J]. IET Image Processing,2013, 7(4): 373-383.
  • 9Farnell D J J, Hatfield F N, Knox P, et al. Enhancement ofblood vessels in digital fundus photographs via the applicationof multiscale line operators[J]. Journal of the Franklin Institute,2008, 345(7): 748-765.
  • 10Al-Diri B, Hunter A, Steel D, et al. REVIEW – a reference dataset for retinal vessel profiles[C] //Proceedings of 30th AnnualInternational Conference of the IEEE. Los Alamitos: Engineeringin Medicine and Biology Society Press, 2008: 2262–2265.

共引文献156

同被引文献3

引证文献2

二级引证文献3

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部