Automated segmentation and feature discovery of age-related macular degeneration and Stargardt disease via self-attended neural networks.
Automated segmentation and feature discovery of age-related macular degeneration and Stargardt disease via self-attended neural networks.
复制标题
DOI：
10.1038/s41598-022-18785-6
复制
发表时间：
2022-08-26
影响因子：
4.6
通讯作者：
Hu, Zhihong Jewel
中科院分区：
文献类型：
--
作者：
Wang, Ziyuan;Sadda, Srinivas Reddy;Lee, Aaron;Hu, Zhihong Jewel
视网膜相关性黄斑变性（AMD）和Stargardt病分别是老年人和年轻人失明的主要原因。AMD的地图状萎缩（GA）和Stargardt萎缩是其终末期结局。有效的分割和量化这些萎缩性病变的方法是至关重要的临床研究。在这项研究中，我们开发了一种深度卷积神经网络（CNN），具有可训练的自我参与机制，用于准确的GA和Stargardt萎缩分割。与只能可视化CNN特征的传统事后注意机制相比，我们的自关注机制嵌入在完全卷积网络中，并直接参与训练CNN主动关注关键特征，以增强算法性能。我们应用自参与CNN对眼底自发荧光（FAF）图像上的AMD和Stargardt萎缩性病变进行分割。与现有的常规全卷积网络（U-Net）相比，我们的自参与CNN在AMD GA分割中实现了10.6%的Dice系数和17%的IoU（交集），在Stargardt萎缩分割中实现了22%的Dice系数和32%的IoU。随着纵向图像数据具有较长的时间，所开发的自我关注机制也可以应用于早期AMD和Stargardt特征的视觉发现。
Age-related macular degeneration (AMD) and Stargardt disease are the leading causes of blindness for the elderly and young adults respectively. Geographic atrophy (GA) of AMD and Stargardt atrophy are their end-stage outcomes. Efficient methods for segmentation and quantification of these atrophic lesions are critical for clinical research. In this study, we developed a deep convolutional neural network (CNN) with a trainable self-attended mechanism for accurate GA and Stargardt atrophy segmentation. Compared with traditional post-hoc attention mechanisms which can only visualize CNN features, our self-attended mechanism is embedded in a fully convolutional network and directly involved in training the CNN to actively attend key features for enhanced algorithm performance. We applied the self-attended CNN on the segmentation of AMD and Stargardt atrophic lesions on fundus autofluorescence (FAF) images. Compared with a preexisting regular fully convolutional network (the U-Net), our self-attended CNN achieved 10.6% higher Dice coefficient and 17% higher IoU (intersection over union) for AMD GA segmentation, and a 22% higher Dice coefficient and a 32% higher IoU for Stargardt atrophy segmentation. With longitudinal image data having over a longer time, the developed self-attended mechanism can also be applied on the visual discovery of early AMD and Stargardt features.
登录
查看更多内容
超越视网膜层：SD-OCT 图像中自动地理萎缩分割的深度投票模型
影响因子：
3
作者：
Ji Z;Chen Q;Niu S;Leng T;Rubin DL
通讯作者：
Rubin DL
影响因子：
3.4
作者：
Chu, Zhongdi;Wang, Liang;Wang, Ruikang K.
通讯作者：
Wang, Ruikang K.
影响因子：
4.4
作者：
Schmitz-Valckenberg, Steffen;Brinkmann, Christian K.;Holz, Frank G.
通讯作者：
Holz, Frank G.
影响因子：
4.6
作者：
Saha, Sajib;Nassisi, Marco;Hu, Zhihong Jewel
通讯作者：
Hu, Zhihong Jewel
影响因子：
4.2
作者：
Klein, Ronald;Klein, Barbara E. K.;Gangnon, Ronald E.
通讯作者：
Gangnon, Ronald E.