RTVue傅里叶光学相干断层扫描仪测量角膜参数的重复性和准确性评价

Repeatability and accuracy of corneal parameters measured by RTVue Fourier-domain optical coherence topography

背景角膜各参数（如曲率、厚度等）对于诊断角膜源性疾病、角膜接触镜的验配及角膜屈光手术至关重要。目的研究RTVue傅里叶光学相干断层扫描仪（FD—OCT）评估角膜各参数的重复性与准确性，为FD—OCT的临床应用提供参考依据。方法本研究为前瞻性研究。接受准分子激光角膜原位磨镶术（LASIK）术前检查者43例77眼，屈光度为-1．25—-10．00D，散光度〈2D。使用RTVueFD—OCT测量角膜中央3miTl直径范围内前表面曲率半径Ranterior、后表面曲率半径Rposterier、角膜后前表面曲率半径之比Rposterion／Ranterier、中央角膜厚度（CCT）、通过厚透镜公式获得角膜总屈光力Knet、模拟角膜屈光力SimK、角膜前表面中央直径3mm范围内屈光力Kanterior、角膜后表面中央直径3mm范围内屈光力Kposierior，及基于Placido环原理的Topolyzer获得的角膜屈光力K。，每只受试眼连续测量3次。采用变异系数（CV）、CronbachAlpha系数及组内相关系数（ICC）评估重复性，重复测量方差分析比较SimK与Knet、Knet与Km、SimK与Km之间的差异性，Pearson相关分析SimK与Knet、Knet，与Km、SimK与Km之间的相关性，Bland—Altman分析SimK与Knet、Knet与Km、SimK与K。之间的一致性。结果Ranterior,Rposterior、Rposterior／Ranterior、Kanterior、Kposterior SimK、Knet和CCT分别为（7．691±0．302）mm、（6．532±0．276）mill、0．849±0．014、（48．97±1．92）D、（-6．13±0．26）D、（43．95±1．72）D、（42．95±1．68）D和（545．20±35．04）mm，各参数的CV均〈1％，CronbachAlpha系数均〉O．9；Rposterior／Ranterier的ICC为0．802，其余参数ICC均〉O．9。SimK比Km大（0．27＋0．34）D，Knet比Km小（0．73±0．37）D，SimK比Knet大（1．01±0．11）D，差异均有统计学意义（P〈O．001）。SimK与K。、K。与K。、SimK与K。均呈正相关（r=0．981、0．976、0．998，P〈0．001）。Bland—Altman分析显示，SimK、Knet、Km之间的一致性差。结论RTVueFD—OCT评估角膜曲率和CCT具有很好的重复性；RTVueFD—OCT获得的SimK比角膜地形图仪略大；RTVueFD—OCT获得的Rposterior／Ranterior，比Gullstrand模型眼小，可能为建立更准确的标准化模型眼提供依据。

Background Corneal parameters （such as curvature, thickness, etc） are essential to the diagnosis of corneal related diseases, contact lenses fitting and corneal refractive surgery. Objective The aim of this study was to assess the repeatability and accuracy of corneal parameters obtained by RTVue Fourier-domain optical coherence topography （ FD-OCT）. Methods In this prospective study, 77 eyes of 43 subjects with the refraction from -1.25 D to -10.00 D and astigmatism 〈2 D were enrolled in keratorefractive surgery center, Eye Hospital of Wenzhou Medical College. The anterior and posterior corneal curvature in 3 mm central zone （ R nor and Rpo,oror） ,the ratio of posterior and anterior curvature （Rpote,~or/Rrior）, corneal central thickness （CCT）, total corneal power（ Kno, ） , the simulated corneal power （SlinK） , the anterior and posterior corneal power （ Kanterlo, Kpo ） were measured by FD-OCT. Corneal power （ Km ） was obtained by Topolyzer topography based on Placido ring. Three consecutive scans were acquired in every tested eye. Repeatability of FD-OCT measurement was assessed using the coefficient of variation （ CV）, Cronbach Alpha and intra-elass correlation coefficient （ICC）. Repeated measure ANOVA was used to analyze the differences among SimK, Koo, and K,,. Pearson correlation analysis was used to analysis the correlation between SimK and K o,, K e, and Km , SimK and K. The agreement between SimK and K,, K o, and Km,SimK and K was assessed by Bland-Altman plots analysis. All the subjects understood the purpose of this investigation and written informed consent was obtained prior to the medical examination. Results The Ranterior, Rposterior , 1~posterior/Ranterior, K ,~rior, Kpo~,~rio~ , SimK, K,ot and CCT were （ 7. 691 ±0. 302） ram, （ 6. 532-±0. 276） ram, 0. 849±0. 014, （48.97±1.92） D, （-6. 13±0.26）D,（43.95±1.72） D, （42.95±1.68）D and （545.20± 35.04） μm, respectively. The CV of all parameters was less than 1% . The Cronbach Alpha and ICC were both more than 0. 9 except 0. 802 （]CC）of the Rooor/R erlo. SimK was （0. 27±0. 34） D higher than Kin, and Koo was （0. 73 ±0.37 ）D lower than Km. SimK was （1.01 ±0. 11 ）D higher than K and all the differences showed a statistical significance （P〈0. 001 ）. There were close positive correlations between SimK and K, Kno, and Kin, SimK and Kno,（ all r I〉 0. 976, P〈0. 001 ）. The Bland-Altman analysis showed a poor agreement between SimK and Km ,Ko and K ,SlinK and K,. Conclusions The RTVue FD-OCT measuring corneal curvature and CCT performes good repeatability. The SimK obtained by RTVue FD-OCT is slightly greater than that obtained by corneal topography, and the Roorio/R obtained by RTVue FD-OCT is slightly lower than that from Gullstrand model eye. These results offer a basis for establishing a more accurate standardized model eye.