【作者】 张帅；

【导师】 汤欣；

【作者基本信息】 天津医科大学 ， 眼科学， 2009， 博士

【摘要】 目的：探讨利用Pentacam三维眼前节分析诊断系统观察年龄相关性白内障患者手术前后前房直径、前房深度(anterior chamber depth，ACD)、角膜前表面波前像差的变化。比较Pentacam与全景超声生物显微镜(UltrasoundBiomicroscopy，UBM)测量前房直径、ACD的差异；Pentacam与iTrace像差仪测量术后角膜前表面波前像差的差异。探讨采用人工神经网络(artifical nueral network，ANN)的方法实现人眼前房直径的预测。方法：临床选取行白内障超声乳化吸除折叠型人工晶状体(intraocular lens，IOI)植入术后的年龄相关性白内障患者60例(79只眼)。应用Pentacam对全部患者进行检查，分别测量术前、术后1周、1个月、2个月的水平前房直径和ACD。应用全景UBM对全部患进行检查，测量术前、术后2个月的水平、垂直前房直径和ACD。分别于术前、术后2个月对全部患者应用Pentacam测量角膜前表面的像差，于术后2个月对全部患者行iTrace视觉功能分析仪检查。对99例(121只眼)行超声乳化白内障吸除折叠型IOL植入术的年龄相关性白内障患者，利用全景UBM、角膜曲率计、角膜地形图仪、A型超声诊断仪分别测量术前、术后2个月的水平前房直径及术前前房深度(anterior chamber depth，ACD)，角膜曲率，3mm、5mm、7mm区的角膜曲率半径，眼轴长度。选择88例(176只眼)青年人作为对照组，进行以上检查。采用反向传播(back-propagation，BP)ANN的方法，利用以上参数分别预测两组的前房直径。随机选择38例(38只眼)行超声乳化白内障吸除联合折叠型IOL植入术的年龄相关性白内障患者及15例(30只眼)青年人分别验证前房直径预测的有效性和准确性。结果：应用Pentacam测量年龄相关性白内障超声乳化吸除联合折叠型IOL植入术后水平前房直径增加(F=7．170，P=0．009)、ACD加深(F=26．439，P＜0．05)，术后1周、1个月、2个月前房直径、ACD进行两两比较，差别无统计学意义P＞0．05)。Pentacam测量术前、术后2个月的水平前房直径与全景UBM测量值相比，差别无统计学意义(t=0．231，P=0．818；t=1．017，P=0．311)。测量术前ACD，与UBM相比差别无统计学意义(t=0．153，P=0．995)；而测量术后ACD，差别有统计学意义(P＜0．05)。应用Pentacam测量角膜前表面总的高阶像差均方根值(Root-mean-square，RMS)术后2个月较术前轻度增加，但手术前后变化差别无统计学意义(t=0．717，P＞0．05)；其余各项球差、彗差、三叶草以及2～6阶像差的RMS手术前后的变化与总的高阶像差的变化相似。分别应用Pentacam和iTrace测量白内障患者术后角膜前表面的角膜波前像差，其中彗差、三叶草像差的RMS，Pentacam测量结果与iTrace相比，差别无统计学意义(P＞0．05)；而总的高阶像差、球差、2～6阶像差RMS两者比较差别均有统计学意义(P＜0．05)。全部年龄相关性白内障患者术后水平前房直径与术前呈显著正相关(r=0．805，P＜0．05)，回归方程为：y=5．326+0．556x；(x为术前前房直径，y为术后前房直径)。使用Matlab7．0．1软件加载训练好的BP-ANN，设计出软件的操作平台，可以预测出前房直径。白内障组与青年组分别用输入4个参数、输入14个参数的神经网络软件、W1、W2回归方程预测前房直径与全景UBM测得的真实值比较，差别均无统计学意义(P＞0．05)。结论：Pentacam测量结果显示，年龄相关性白内障患者术后水平前房直径增加，ACD加深，术后1周前房直径、ACD保持不变。UBM与Pentacam都可以用于前房直径和ACD的测量，而且对于年龄相关性白内障患者术前、术后前房直径以及术前ACD的测量结果相似。但测量白内障患者术后IOL眼的ACD时，两种仪器的测量结果不同。超声乳化白内障吸除术后角膜前表面高阶像差的变化不影响术后患者的视觉质量。对于Pentacam三维眼前节分析诊断系统和iTrace像差仪测量角膜波前像差，结果存在部分差别。是因为两种仪器测量角膜部位不同从而产生差别。但是每一种仪器都有各自的优缺点，使用时应根据不同的目的选用不同的仪器。在不具备全景UBM的医院，使用常规的临床检查数据，运用Matlab-ANN软件准确的预测前房直径，为选择适合的前房型人工晶状体(anterior chamberintraocular lenses，AC-IOL)提供可靠依据，从而减少术后并发症。更多还原

【Abstract】 Objective: To study the change in anterior chamber diameter, anterior chamberdepth (ACD), the optical aberrations of the cornea after phacoemlsification andfoldable intraocular lens (IOL) implantation with Pentacam. To compare anteriorchamber diameter and ACD measurements using Pentacam and panoramic ultrasoundbiomicroscopy (UBM) and the aberrations of the anterior surface of cornea measuredwith Pentacam and iTrace wavefront aberrometer.To study the implementation of anterior chamber diameter forecasting systembased on artificial nueral network.Methods: Seventy-nine eyes of 60 patients with age-related cataract which wereperformed small-incision phacoemulsification and foldable IOL implantation wereinvolved in this study. The anterior chamber diameter in horizontal meridian andACD were measured by Pentacam before surgery, and 1 week, 1 months, 2 monthsafter surgery. The anterior chamber diameter in horizontal and vertical meridian andACD were measured by panoramic UBM before and 2 months after surgery. All eyeswere measured with Pentacam and iTrace wavefront aberrometer two months aftercataract surgery at random order. We record Zernike coefficients, RMS of total highorder aberration (HOA), spherical, coma and trefoil aberration, from 2 order to 6order aberrations.One hundreds and twenty-one eyes of 99 patients with age-related cataractwhich were performed small-incision phacoemulsification and foldable IOLimplantation and 176 eyes of 88 normal youth persons are selected. The anteriorchamber diameter in horizontal meridian, anterior chamber depth (ACD), cornealcurvature, radius of corneal curvature, axial length are measured by panoramic UBM,A-ultrasonoscope, corneal topographer and corneal keratometer. Based onback-propagation (BP) neural network, we used a forecasting method to forecast thediameter of anterior chamber. 38 eyes of 38 patients and 22 eyes of 14 normal youthpersons were selected to evaluate the validity and acuity.Results: In all patients, the horizontal anterior chamber diameter was widenedsignificantly after surgery (F=7.170, P=0.009), and the ACD was deepened (F= 26.439,P＜0.05). There were no statistically significant difference among thehorizontal anterior chamber diameters and the ACD 1 week, 1 months, 2 months aftersurgery (P＞0.05). There were no statistically significant difference between Pentacamand UBM in the horizontal anterior chamber diameters before and 2 months aftersurgery (t=0.231, P=0.818; t=1.017, P=0.311). In the ACD before surgery, thedifference between the two devices was not statistically significant (t=0.153,P=0.995), but after surgery the difference was statistically significant (P＜0.05). In theRMS of the total HOA, spherical aberration, coma, trefoil and from 2 order to 6 orderaberrations of the anterior surface increased slightly from a mean of (1.025±0.213)μm preoperatively to (1.050±0.280)μm postoperatively (t= 0.717, P＞0.05). The RMSof total HOA, spherical aberration and from 2 order to 6 order aberrations showedstatistically difference between Pentacam and iTrace (P＜0.05).In all patients, There was a significant positive correlation between thepreoperative and postoperative anterior chamber diameters (r=0.805,P＜0.05). Theregression equation was y=5.326+0.556x. Using Matlab-BP ANN software, thediameter of anterior chamber can be forecasted. The differences between the resultsof different methods were not statistically significant (P＞0.05).Conclusion: After phacoemulsification and foldable IOL implantation,Pentacam revealed that the anterior chamber diameter was widened and the ACD wasdeepened. After one week, the anterior chamber diameter and ACD remainedunchanged. Pentacam and UBM can both be used for the anterior chamber diametermeasurements and yielded comparable results. But for the ACD, the difference wassignificantly lower when the ACD was measured with the Pentacam after surgery.Phacoemulsification cataract surgery does not systematically degrade the opticalquality of the anterior corneal surface. There were partly difference of Zernikecoefficients obtained by Pentacam and iTrace.Using Matlab-BP ANN software, the diameter of anterior chamber can bepredicted by input parameters of anterior chamber depth, corneal curvature or radiusof corneal curvature and axial length.更多还原