期刊文献+
共找到2篇文章
< 1 >
每页显示 20 50 100
青光眼影像人工智能深度学习研究现状与展望 被引量:6
1
作者 carol y.cheung 冉安然 《山东大学学报(医学版)》 CAS 北大核心 2020年第11期24-32,38,共10页
青光眼是一组异质性神经退行性疾病,其特征是视网膜神经节细胞及其轴突逐渐消失,现已成为全球不可逆性失明的主要原因。人工智能(AI)是由机器展示的智能,而深度学习(DL)是其中一个基于深度神经网络的分支,在医学成像领域取得了重大突破... 青光眼是一组异质性神经退行性疾病,其特征是视网膜神经节细胞及其轴突逐渐消失,现已成为全球不可逆性失明的主要原因。人工智能(AI)是由机器展示的智能,而深度学习(DL)是其中一个基于深度神经网络的分支,在医学成像领域取得了重大突破。在青光眼影像方面,已有越来越多的研究将DL应用于眼底图像以及光学相干断层扫描(OCT),以检测青光眼性视神经病变。有很好的结果显示,将DL技术整合到影像中进行青光眼评估是高效、准确的,这可能会解决当前实践和临床工作流程中的一些难题。但是,未来进一步的研究对于解决现存挑战至关重要,例如为不同研究之间的图像标记建立标准,将"黑匣子"的学习过程进行可视化,提高模型在未知数据集上的泛化能力,开发基于DL的实际应用程序,以及建立合理的临床工作流程,进行前瞻性验证和成本效益分析等。本文总结了AI应用于青光眼影像的最新研究现状,并讨论了对临床的潜在影响和未来的研究方向。 展开更多
关键词 青光眼影像 人工智能 光学相干断层扫描 眼底照相 深度学习
原文传递
Clinically relevant factors associated with quantitative optical coherence tomography angiography metrics in deep capillary plexus in patients with diabetes 被引量:10
2
作者 Fang Yao Tang Erica O.Chan +8 位作者 Zihan Sun Raymond Wong Jerry Lok Simon Szeto Jason C.Chan Alexander Lam Clement C.Tham Danny S.Ng carol y.cheung 《Eye and Vision》 SCIE CSCD 2020年第1期64-74,共11页
Background:To test clinically relevant factors associated with quantitative artifact-free deep capillary plexus(DCP)metrics in patients with diabetes mellitus(DM).Methods:563 eligible eyes(221 with no diabetic retinop... Background:To test clinically relevant factors associated with quantitative artifact-free deep capillary plexus(DCP)metrics in patients with diabetes mellitus(DM).Methods:563 eligible eyes(221 with no diabetic retinopathy[DR],135 with mild DR,130 with moderate DR,and 77 with severe DR)from 334 subjects underwent optical coherence tomography-angiography(OCT-A)with a swept-source OCT(Triton DRI-OCT,Topcon,Inc.,Tokyo,Japan).Strict criteria were applied to exclude from analysis those DCP images with artifacts and of poor quality,including projection artifacts,motion artifacts,blurriness,signal loss,B-scan segmentation error,or low-quality score.A customized MATLAB program was then used to quantify DCP morphology from the artifact-free DCP images by calculating three metrics:foveal avascular zone(FAZ),vessel density(VD),and fractal dimension(FD).Results:166(29.5%)eyes were excluded after quality control,leaving in the analysis 397 eyes(170 with no DR,101 with mild DR,90 with moderate DR,36 with severe DR)from 250 subjects.In the multiple regression models,larger FAZ area was associated with more severe DR(β=0.687;p=0.037),shorter axial length(AL)(β=−0.171;p=0.003),thinner subfoveal choroid thickness(β=−0.122;p=0.031),and lower body mass index(BMI)(β=−0.090;p=0.047).Lower VD was associated with more severe DR(β=−0.842;p=0.001),shorter AL(β=0.107;p=0.039),and poorer visual acuity(VA)(β=−0.133;p=0.021).Lower FD was associated with more severe DR(β=−0.891;p<0.001)and with older age(β=−0.142;p=0.004).Conclusions:Quantitative artifact-free DCP metrics are associated with VA,DR severity,AL,subfoveal choroidal thickness,age,and BMI in diabetic patients.The effects of ocular and systemic factors should be considered for meaningful interpretations of DCP changes in DM patients. 展开更多
关键词 Optical coherence tomography angiography Diabetic retinopathy Deep capillary plexus Visual acuity
原文传递
上一页 1 下一页 到第
使用帮助 返回顶部