Page 54 - Handbook of Deep Learning in Biomedical Engineering Techniques and Applications

Article type Journal Journal Journal Conference Journal

References [84] [85] [47] [86] [87]

set 35 London per contains seven US-licensed senior graded
data from individually labels which retinal to diabetic gradeability; December 5600 set data database test stage, Messidor images
Canada images whereas expert used 1,28,175 three retinopathy, image 54 of ophthalmology and than more AREDS Kaggle images 53,576 DRP available 1200 sites

set used: usable exams, 106 of 4 with was images of graded and panel a and May using NIH used: training and stages undisclosed publicly of different
data were 1459 347 consisted database fundus consisted were diabetic edema, by done between done the were 35,126 consisted three

on sets of and set images color set that for was ophthalmologists was from sets of DRP five with the at
Remark data Two fundus consisted patients data labeled image DRiDB The 50 data The images times macular this residents 2015 Testing images data Two consists into images whereas database acquired

imaging.dcontinued Modality CNN Neonatal images Fundus photographs fundus Retinal photographs image Fundus Retinography and

medical of Type GoogleNet CNN Deep CNN Deep CNN Deep CNN SeS CNN NSeS

to detection of of photographs of retinal age- of hemorrhage
application automatic prematurity diagnosis early detection by fundus detection using detection degeneration detect

CNN for of for retinopathy color for retinopathy photographs for macular to images fundus

of designed retinopathy designed in developed designed designed
Literature Applications Model of System diabetic exudates System diabetic fundus System related Model color in

2.2 Year 2016
Table Body part