Automated Segmentation of Autofluorescence Lesions in Stargardt Disease.
Automated Segmentation of Autofluorescence Lesions in Stargardt Disease.
复制标题
DOI：
10.1016/j.oret.2022.05.020
复制
发表时间：
2022-11
影响因子：
4.5
通讯作者：
Jayasundera, K. Thiran
中科院分区：
文献类型：
--
作者：
Zhao, Peter Y.;Branham, Kari;Schlegel, Dana;Fahim, Abigail T.;Jayasundera, K. Thiran
训练深度学习（DL）算法对Stargardt病中多种自身荧光病变类型进行全自动语义分割。回顾性影像学资料横断面研究。来自97名Stargardt病患者193只眼睛的193张图像。眼底自身荧光（FAF）图像从2013年至2020年患者就诊中获得，并使用ground-truth标签进行注释。模型训练和评估采用五重交叉验证。骰子相似系数、类内相关系数（ICCs）和Bland-Altman分析比较了算法预测和分级标记的分割。所有病变类别的总体Dice相似系数为0.78 （95%CI, 0.69-0.86）。自体荧光明显减弱区域（DDAF）的Dice系数为0.90 (95%CI, 0.85-0.94)，可疑自体荧光减弱区域（QDAF）的Dice系数为0.55 (95%CI, 0.35-0.76)，异常背景自体荧光区域（ABAF）的Dice系数为0.88 （95%CI, 0.73-1.00）。DDAF的ICCs为0.997 (95%CI, 0.996-0.998)， QDAF的ICCs为0.863 (95%CI, 0.823-0.895)， ABAF的ICCs为0.974 （95%CI, 0.966-0.980）。DL算法对Stargardt病的自身荧光病变进行了准确的分割，证明了全自动分割作为人工或半自动标记方法的替代方法的可行性。ResNet-UNet卷积神经网络可以准确标记Stargardt病自身荧光图像中的多种病变类型，促进疾病进展的自动化监测。
To train a deep learning (DL) algorithm to perform fully automated semantic segmentation of multiple autofluorescence lesion types in Stargardt disease. Cross-sectional study with retrospective imaging data. 193 images from 193 eyes of 97 patients with Stargardt disease. Fundus autofluorescence (FAF) images obtained from patient visits between 2013 and 2020 were annotated with ground-truth labels. Model training and evaluation were performed with five-fold cross-validation. Dice similarity coefficients, intraclass correlation coefficients (ICCs), and Bland-Altman analyses comparing algorithm-predicted and grader-labeled segmentations. The overall Dice similarity coefficient across all lesion classes was 0.78 (95%CI, 0.69–0.86). Dice coefficients were 0.90 (95%CI, 0.85–0.94) for areas of definitely decreased autofluorescence (DDAF), 0.55 (95%CI, 0.35–0.76) for areas of questionably decreased autofluorescence (QDAF), and 0.88 (95%CI, 0.73–1.00) for areas of abnormal background autofluorescence (ABAF). ICCs comparing the ground truth and automated methods were 0.997 (95%CI, 0.996–0.998) for DDAF, 0.863 (95%CI, 0.823–0.895) for QDAF, and 0.974 (95%CI, 0.966–0.980) for ABAF. A DL algorithm performed accurate segmentation of autofluorescence lesions in Stargardt disease, demonstrating the feasibility of fully automated segmentation as an alternative to manual or semi-automated labeling methods. A ResNet-UNet convolutional neural network can accurately label multiple lesion types in autofluorescence images for Stargardt disease, facilitating automated monitoring of disease progression.
登录
查看更多内容
Stargardt病：临床特征，分子遗传学，动物模型和治疗选择。
DOI：
10.1136/bjophthalmol-2016-308823
发表时间：
2017-01
期刊：
The British journal of ophthalmology
影响因子：
--
作者：
Tanna P;Strauss RW;Fujinami K;Michaelides M
通讯作者：
Michaelides M
DOI：
10.1007/978-3-319-67558-9_28
发表时间：
2017-09-09
期刊：
Deep learning in medical image analysis and multimodal learning for clinical decision support : Third International Workshop, DLMIA 2017, and 7th International Workshop, ML-CDS 2017, held in conjunction with MICCAI 2017 Quebec City, QC,..
影响因子：
--
作者：
Sudre CH;Li W;Vercauteren T;Ourselin S;Jorge Cardoso M
通讯作者：
Jorge Cardoso M
影响因子：
--
作者：
Jeffery, Rachael C. Heath;Thompson, Jennifer A.;Lo, Johnny;Lamey, Tina M.;McLaren, Terri L.;McAllister, Ian L.;Mackey, David A.;Constable, Ian J.;De Roach, John N.;Chen, Fred K.
通讯作者：
Chen, Fred K.
影响因子：
13.7
作者：
Strauss, Rupert W.;Ho, Alex;Scholl, Hendrik P. N.
通讯作者：
Scholl, Hendrik P. N.
影响因子：
3
作者：
Durham TA;Duncan JL;Ayala AR;Birch DG;Cheetham JK;Ferris FL 3rd;Hoyng CB;Pennesi ME;Sahel JA;Foundation Fighting Blindness Consortium Investigator Group
通讯作者：
Foundation Fighting Blindness Consortium Investigator Group