AIM:To propose an algorithm for automatic detection of diabetic retinopathy(DR)lesions based on ultra-widefield scanning laser ophthalmoscopy(SLO).METHODS:The algorithm utilized the FasterRCNN(Faster Regions with CNN ...AIM:To propose an algorithm for automatic detection of diabetic retinopathy(DR)lesions based on ultra-widefield scanning laser ophthalmoscopy(SLO).METHODS:The algorithm utilized the FasterRCNN(Faster Regions with CNN features)+ResNet50(Residua Network 50)+FPN(Feature Pyramid Networks)method for detecting hemorrhagic spots,cotton wool spots,exudates,and microaneurysms in DR ultra-widefield SLO.Subimage segmentation combined with a deeper residual network FasterRCNN+ResNet50 was employed for feature extraction to enhance intelligent learning rate.Feature fusion was carried out by the feature pyramid network FPN,which significantly improved lesion detection rates in SLO fundus images.RESULTS:By analyzing 1076 ultra-widefield SLO images provided by our hospital,with a resolution of 2600×2048 dpi,the accuracy rates for hemorrhagic spots,cotton wool spots,exudates,and microaneurysms were found to be 87.23%,83.57%,86.75%,and 54.94%,respectively.CONCLUSION:The proposed algorithm demonstrates intelligent detection of DR lesions in ultra-widefield SLO,providing significant advantages over traditional fundus color imaging intelligent diagnosis algorithms.展开更多
Cone photoreceptor cell identication is important for the early diagnosis of retinopathy.In this study,an object detection algorithm is used for cone cell identication in confocal adaptive optics scanning laser ophtha...Cone photoreceptor cell identication is important for the early diagnosis of retinopathy.In this study,an object detection algorithm is used for cone cell identication in confocal adaptive optics scanning laser ophthalmoscope(AOSLO)images.An effectiveness evaluation of identication using the proposed method reveals precision,recall,and F_(1)-score of 95.8%,96.5%,and 96.1%,respectively,considering manual identication as the ground truth.Various object detection and identication results from images with different cone photoreceptor cell distributions further demonstrate the performance of the proposed method.Overall,the proposed method can accurately identify cone photoreceptor cells on confocal adaptive optics scanning laser ophthalmoscope images,being comparable to manual identication.展开更多
An automated superpixels identification/mosaicking method is presented for the analysis of cone photoreceptor cells with the use of adaptive optics scanning laser ophthalmoscope(AO-SLO) images. This is an image overse...An automated superpixels identification/mosaicking method is presented for the analysis of cone photoreceptor cells with the use of adaptive optics scanning laser ophthalmoscope(AO-SLO) images. This is an image oversegmentation method used for the identification and mosaicking of cone photoreceptor cells in AO-SLO images.It includes image denoising, estimation of the cone photoreceptor cell number, superpixels segmentation, merging of superpixels, and final identification and mosaicking processing steps. The effectiveness of the presented method was confirmed based on its comparison with a manual method in terms of precision, recall, and F1-score of 77.3%, 95.2%, and 85.3%, respectively.展开更多
A bimorph deformable mirror (DM) with a large stroke of more than 30 μm using 35 actuators is presented and characterized for an adaptive optics (AO) confocal scanning laser ophthalmoscope application. Facilitate...A bimorph deformable mirror (DM) with a large stroke of more than 30 μm using 35 actuators is presented and characterized for an adaptive optics (AO) confocal scanning laser ophthalmoscope application. Facilitated with a Shack-Hartmann wavefront sensor, the bimorph DM-based AO operates closed-loop AO corrections for hu- man eyes and reduces wavefront aberrations in most eyes to below 0.1 μm rms. Results from living eyes, including one exhibiting ~5D of myopia and ~2D of astigmatism along with notable high-order aberrations, reveal a prac- tical efficient aberration correction and demonstrate a great benefit for retina imaging, including improving resolution, increasing brightness, and enhancing the contrast of images.展开更多
Background:With a large portion of older adults living longer,the number of individuals diagnosed with low vision is increasing.The use of optical coherence tomography/scanning laser ophthalmoscope(OCT/SLO)to diagnose...Background:With a large portion of older adults living longer,the number of individuals diagnosed with low vision is increasing.The use of optical coherence tomography/scanning laser ophthalmoscope(OCT/SLO)to diagnose retinal disease has become common place in the last 10 years,yet currently there are no OCT/SLO databases for pathological vision.Our aim is to develop a clinical database of individuals who have drusen(i.e.,lipid deposits found under the retina),or have been diagnosed with age-related macular degeneration(AMD),with information as to how the structure of the diseased retina changes over time,as well as measures of visual and cognitive functional performance.Methods:Fundus photographs and retinal scans will be taken using the same model of optos OCT/SLO located in three test sites(MAB-Mackay Rehabilitation Centre,School of Optometry Clinic at the University of Montreal,and the Lighthouse Institute,New York,USA).For each individual entry in the database,demographic and diagnosis information will be available.All OCT/SLO images will be graded according to the Age-related Eye Disease Study standard,in addition to number and size of drusen,severity of geographic atrophy,severity of pigment mottling and presence of choroidal neovascularization.Retinal topography and Raster scans from the OCT/SLO will provide a cross-sectional look at affected retinas.Fixation stability will be recorded using the SLO function,and present four different tasks that are designed to reproduce typical tasks of daily vision,with each task lasting for 10 seconds.The tasks are cross fixation,face recognition,visual search,and reading.These tasks in addition to the retinal scans will be used to determine the eccentricity of a preferred retinal locus from the anatomical fovea,and can be used as an outcome measure for clinical interventions in visually impaired patients.Results:The database will be available to professors training eye-care practitioners and rehabilitation specialists as a teaching tool.Students will be able to familiarize themselves with the retina and a variety of AMD-related pathologies before they start working with patients.The database will also be accessible by researchers interested in studying AMD from basic science to epidemiology,to investigate how drusen and AMD impact visual and cognitive functional performance.Conclusions:The common infrastructure is easily accessible to all VHRN members on request.The database will also be accessible online in 2018(see http://cvl.concordia.ca for more information).展开更多
基金Supported by Hunan Provincial Science and Technology Department Clinical Medical Technology Innovation Guidance Project(No.2021SK50103)。
文摘AIM:To propose an algorithm for automatic detection of diabetic retinopathy(DR)lesions based on ultra-widefield scanning laser ophthalmoscopy(SLO).METHODS:The algorithm utilized the FasterRCNN(Faster Regions with CNN features)+ResNet50(Residua Network 50)+FPN(Feature Pyramid Networks)method for detecting hemorrhagic spots,cotton wool spots,exudates,and microaneurysms in DR ultra-widefield SLO.Subimage segmentation combined with a deeper residual network FasterRCNN+ResNet50 was employed for feature extraction to enhance intelligent learning rate.Feature fusion was carried out by the feature pyramid network FPN,which significantly improved lesion detection rates in SLO fundus images.RESULTS:By analyzing 1076 ultra-widefield SLO images provided by our hospital,with a resolution of 2600×2048 dpi,the accuracy rates for hemorrhagic spots,cotton wool spots,exudates,and microaneurysms were found to be 87.23%,83.57%,86.75%,and 54.94%,respectively.CONCLUSION:The proposed algorithm demonstrates intelligent detection of DR lesions in ultra-widefield SLO,providing significant advantages over traditional fundus color imaging intelligent diagnosis algorithms.
基金the Natural Science Foundation of Jiangsu Province(BK20200214)National Key R&D Program of China(2017YFB0403701)+5 种基金Jiangsu Province Key R&D Program(BE2019682 and BE2018667)National Natural Science Foundation of China(61605210,61675226,and 62075235)Youth Innovation Promotion Association of Chinese Academy of Sciences(2019320)Frontier Science Research Project of the Chinese Academy of Sciences(QYZDB-SSW-JSC03)Strategic Priority Research Program of the Chinese Academy of Sciences(XDB02060000)and Entrepreneurship and Innova-tion Talents in Jiangsu Province(Innovation of Scienti¯c Research Institutes).
文摘Cone photoreceptor cell identication is important for the early diagnosis of retinopathy.In this study,an object detection algorithm is used for cone cell identication in confocal adaptive optics scanning laser ophthalmoscope(AOSLO)images.An effectiveness evaluation of identication using the proposed method reveals precision,recall,and F_(1)-score of 95.8%,96.5%,and 96.1%,respectively,considering manual identication as the ground truth.Various object detection and identication results from images with different cone photoreceptor cell distributions further demonstrate the performance of the proposed method.Overall,the proposed method can accurately identify cone photoreceptor cells on confocal adaptive optics scanning laser ophthalmoscope images,being comparable to manual identication.
基金supported by the Jiangsu Provincial Key R&D Program (Nos. BE2019682 and BE2018667)National Natural Science Foundation of China(Nos. 61605210,61675226,and 61378090)+3 种基金Youth Innovation Promotion Association of Chinese Academy of Sciences (No. 2019320)National Key R&D Program of China(Nos. 2016YFC0102500 and 2017YFB0403700)Frontier Science Research Project of the Chinese Academy of Sciences (No. QYZDB-SSW-JSC03)Strategic Priority Research Program of the Chinese Academy of Sciences(No. XDB02060000)
文摘An automated superpixels identification/mosaicking method is presented for the analysis of cone photoreceptor cells with the use of adaptive optics scanning laser ophthalmoscope(AO-SLO) images. This is an image oversegmentation method used for the identification and mosaicking of cone photoreceptor cells in AO-SLO images.It includes image denoising, estimation of the cone photoreceptor cell number, superpixels segmentation, merging of superpixels, and final identification and mosaicking processing steps. The effectiveness of the presented method was confirmed based on its comparison with a manual method in terms of precision, recall, and F1-score of 77.3%, 95.2%, and 85.3%, respectively.
基金supported by the National Science Foundation of China(No.61605210)the National Instrumentation Program(NIP)(No.2012YQ120080)+4 种基金the National Key Research and Development Program of China(No.2016YFC0102500)the Jiangsu Province Science Fund for Distinguished Young Scholars(No.BK20060010)the Frontier Science Research Project of the Chinese Academy of Sciences(No.QYZDB-SSWJSC03)the Strategic Priority Research Program of the Chinese Academy of Sciences(No.XDB02060000)the Zhejiang Province Technology Program(No.2013C33170)
文摘A bimorph deformable mirror (DM) with a large stroke of more than 30 μm using 35 actuators is presented and characterized for an adaptive optics (AO) confocal scanning laser ophthalmoscope application. Facilitated with a Shack-Hartmann wavefront sensor, the bimorph DM-based AO operates closed-loop AO corrections for hu- man eyes and reduces wavefront aberrations in most eyes to below 0.1 μm rms. Results from living eyes, including one exhibiting ~5D of myopia and ~2D of astigmatism along with notable high-order aberrations, reveal a prac- tical efficient aberration correction and demonstrate a great benefit for retina imaging, including improving resolution, increasing brightness, and enhancing the contrast of images.
文摘Background:With a large portion of older adults living longer,the number of individuals diagnosed with low vision is increasing.The use of optical coherence tomography/scanning laser ophthalmoscope(OCT/SLO)to diagnose retinal disease has become common place in the last 10 years,yet currently there are no OCT/SLO databases for pathological vision.Our aim is to develop a clinical database of individuals who have drusen(i.e.,lipid deposits found under the retina),or have been diagnosed with age-related macular degeneration(AMD),with information as to how the structure of the diseased retina changes over time,as well as measures of visual and cognitive functional performance.Methods:Fundus photographs and retinal scans will be taken using the same model of optos OCT/SLO located in three test sites(MAB-Mackay Rehabilitation Centre,School of Optometry Clinic at the University of Montreal,and the Lighthouse Institute,New York,USA).For each individual entry in the database,demographic and diagnosis information will be available.All OCT/SLO images will be graded according to the Age-related Eye Disease Study standard,in addition to number and size of drusen,severity of geographic atrophy,severity of pigment mottling and presence of choroidal neovascularization.Retinal topography and Raster scans from the OCT/SLO will provide a cross-sectional look at affected retinas.Fixation stability will be recorded using the SLO function,and present four different tasks that are designed to reproduce typical tasks of daily vision,with each task lasting for 10 seconds.The tasks are cross fixation,face recognition,visual search,and reading.These tasks in addition to the retinal scans will be used to determine the eccentricity of a preferred retinal locus from the anatomical fovea,and can be used as an outcome measure for clinical interventions in visually impaired patients.Results:The database will be available to professors training eye-care practitioners and rehabilitation specialists as a teaching tool.Students will be able to familiarize themselves with the retina and a variety of AMD-related pathologies before they start working with patients.The database will also be accessible by researchers interested in studying AMD from basic science to epidemiology,to investigate how drusen and AMD impact visual and cognitive functional performance.Conclusions:The common infrastructure is easily accessible to all VHRN members on request.The database will also be accessible online in 2018(see http://cvl.concordia.ca for more information).
文摘目的探讨孤立性脉络膜血管瘤共焦激光扫描眼底血管造影的影像学特征及其临床意义。方法采用海德堡HRA2共焦激光扫描血管造影系统对21例21只眼孤立性脉络膜血管瘤进行眼底血管造影检查,其中12例行荧光素眼底血管造影(fundus fluorescein angiography,FFA)检查,9例行FFA和吲哚菁绿血管造影(indocyanine green angiography,ICGA)同步检查,分析比较孤立性脉络膜血管瘤2种检查的影像学特征。结果FFA检查动脉前期或动脉早期血管丛状或斑状强荧光,静脉期强荧光灶明显渗漏融合,晚期瘤体呈弥漫性强荧光。6例瘤体表面有明显的不规则点片状弱荧光,13例瘤体表面及周围视网膜毛细血管扩张。10例可见与视网膜色素上皮带状萎缩相应的透见荧光。ICGA检查影像特征为早期瘤体由不规则网状血管团样强荧光,中期瘤体呈强荧光团,晚期所有病例均有特征性的"冲洗现象"。FFA和ICGA同步检查同屏显示,成像清晰,有可比性,ICGA比FFA更清晰显示瘤体范围。结论孤立性脉络膜血管瘤ICGA影像比FFA更具有特征性,并能清晰显示肿瘤大小边界;FFA则可清晰显示视网膜血管及视网膜色素上皮的继发性损伤。激光扫描FFA和ICGA同步检查可为孤立性脉络膜血管瘤诊断、治疗及疗效监测提供更多的临床信息。