OCT不同扫描模式检测视网膜神经纤维层厚度的可重复性研究  

Reproducibility of different scan protocols of optical coherence tomography in scanning retinal nerve fiber layer

作　　者：田润[1] 唐罗生[1] 王玲[1] 袁苑[1] 杨晖[1] 黄娟[1] 

机构地区：[1]中南大学湘雅二医院眼科,湖南省长沙市410011

出　　处：《眼科新进展》2008年第5期358-360,共3页Recent Advances in Ophthalmology

摘　　要：目的采用光学相干断层成像扫描(optical coher-ence tomography,OCT)不同扫描模式检测视网膜神经纤维层(retinal nerve fiber layer,RNFL)厚度的可重复性。方法采用双盲法,2名检查者对20名(20眼)正常人分别以OCT3种不同扫描模式:RNFL厚度(3.4)扫描模式,视神经盘圆(0·98+视盘半径)扫描模式,RNFL厚度(2.27×disc)扫描模式,连续3d测量RNFL厚度,每天每眼记录5幅图像,取其中3幅图像的RNFL厚度值的平均值。计算3d中全周平均及各象限RNFL厚度的变异系数。比较3种扫描模式检测RNFL的可重复性。结果OCT3种扫描模式所测全周平均及4象限RNFL厚度的变异系数范围分别是:RNFL厚度(3·4)扫描模式:(4.57±2·14)%～(9.57±6.28)%,视神经盘圆(0.98+视盘半径)扫描模式:(1.14±0.59)%～(4.81±2.22)%,RNFL厚度(2·27×disc)扫描模式:(2.77±1·49)%～(9.52±4.50)%,且均以鼻侧象限最大。2名操作者间变异系数对比差异无统计学意义(P>0.05)。结论OCT3种扫描模式检测RNFL可重复性均较好。以视神经盘圆(0.98+视盘半径)扫描模式最优。3种扫描模式检测各象限RNFL均以鼻侧象限的可重复性最差。Objective To study the reproducibility in measure the thickness of retinal nerve fibre layer （RNFL） by the use of different optical coherence tomography （OCT） scan protocols. Methods Using two-blind method, 20 normal persons （ 20 eyes） were imaged with three different kinds of OCT scan protocols separately, RNFL thickness （ 3.4 ） scan protocol, nerve head circle （0.98 ＋ optic radius） scan protocol, RNFL thickness （ 2. 27 × disc ） scan protocol. Images were obtained on three consecutive days, and each eye was scanned five images everyday, and three highquality images were picked up to form mean RNFL thickness. Variance and coefficient of variation were calculated for 360°average RNFL thickness and quadrant RNFL thickness. The reproducibility of three scan protocols were compared. Results The range of coefficient of variation of the 360 ° average and four quadrants RNFL thickness measured by three OCT scan protocols were separately （ 4.57 ± 2. 14 ） % - （ 9.57 ± 6.28 ）% for RNFL thickness（ 3.4 ） scan protocol, （ 1.14 ±0. 59）% - （4.81 ±2.22）% for nerve head circle （0.98 ＋ optic radius） scan protocol, and （ 2.77 ± 1. 49 ） % - （9.52 ± 4.50 ） % for RNFL thickness （ 2. 27 × disc ） scan protocol. The coefficient of variation of the nasal quadrant is the biggest one. The coefficient of variation between the two operators had no significant difference in statistics （ P 〉 0.05 ）. Conclusion The reproducibility of three different scan protocols was good enough. Nerve head circle （ 0.98 ＋ optic radius ） scan protocol was the best one among them. The reproducibility of nasal quadrant RNFL thickness measurement was the worst one within three scan protocols.

关 键 词：光学相干断层扫描 视网膜神经纤维层 模式 可重复性 

分 类 号：R770.43[医药卫生—眼科]