目的：比较偏振光学相干干涉(partial coherence interferometry，PCI)与接触式A超对眼前节参数的生物学测量结果，以评价两者在眼内人工晶状体屈光度计算的精确性和可重复性。

方法：施行白内障超声乳化摘除联合人工晶状体植入手术的患者121例137眼，分别用PCI与接触式A超于术前测量眼前节相关参数，术后测量最佳矫正视力和实际屈光度。

结果：PCI与接触式A超测量前房深度(anterior chamber depth，ACD)值分别为2.94±0.49mm，2.69±0.51mm，配对t检验，两者的差值为0.25±0.22mm(P<0.001)，有显著差异(r=0.823，P<0.001)。测量眼轴长度(axial length，AL)值分别为24.17±1.64mm，23.81±1.83mm，配对t检验，两者差值为0.36±0.24mm(P<0.001)，有显著差异(r=0.996，P<0.001)。两者测量ACD和AL的95%一致率分别为-0.08mm～+0.48mm，-0.09mm～+0.69mm。IOLMaster生物测量平均预期屈光误差为-0.15±0.38D，平均绝对屈光误差为0.29±0.27D，而优化后的接触式A超测量与IOLMaster相比，平均绝对屈光误差明显增大为0.41±0.38D。同一测量者应用IOLMaster进行AL、ACD、角膜曲率半径的连续测量获得的标准差分别为±25.6μm，±33.4μm，±12.9μm，变异系数分别为0.11%，0.52%，0.17%； 两位不同测量者间测量标准差分别为±21.5，±29.8，±15.9μm，变异系数分别为0.09%，0.62%，0.21%。各测量参数的可信度AL为99.9%，ACD为97.8%，角膜曲率半径(r₁/r₂)为99.8%/99.5%。

结论：PCI与接触式A超均可用于眼前节参数的生物测量，两者的相关性好。但基于光学原理的PCI与接触式A超相比，具有良好的精确性和可重复性，并可同时测量出其他相关参数，具有较好的临床应用前景。

AIM: To compare the measurement of anterior segment parameters by partial coherence interferometry(PCI)and contact ultrasonic(US)axial scan(A-scan). The accuracy in predicting postoperative refraction and the reproducibility of each biometry measurement were also estimated in a prospective study of eyes that underwent phacoemulsification with IOL implantation.

METHODS: Preoperative measurement of anterior segment parameters were prospectively obtained in 137 eyes of 121 subjects with the PCI compared with the US. The postoperative best corrected visual acuity and postoperative refraction were obtained and compared with each biometric method.

RESULTS: There was an excellent correlation between PCI and US measurements for the anterior chamber depth(ACD; r=0.823; P<0.001)and axial length(AL; r=0.996; P<0.001). The mean values of the parameters measured by IOLMaster and US were, respectively, as follows: ACD, 2.94±0.49mm, 2.69±0.51mm; AL, 24.17±1.64mm, 23.81±1.83mm. The mean differences of ACD and AL values between IOLMaster and US measurements were 0.25±0.22mm, 0.36±0.24mm respectively, proved to be statistically significant(P<0.001). With the 95% limits of agreement from -0.08mm to +0.48mm for ACD and from -0.09mm to+0.69mm for AL. For IOLMaster, the mean prediction error -0.15±0.38D, the mean absolute prediction error was 0.29±0.27D with 96% of the eyes within 1D from the predicted refraction. Applanation ultrasonography after optimisation yielded a greater absolute prediction error than the IOLMaster biometry, 0.41±0.38D with 88% of the eyes within 1D from the predicted refraction. For IOLMaster biometry, the intraobserver variability(SD)was ±25.6μm for AL, ±33.4μm for ACD and ±12.9μm for corneal radius. The coefficients of variation(COV)were 0.11%, 0.52%, and 0.17%, respectively. The interobserver variability(SD)was ±21.5μm for AL, ±29.8μm for ACD and ±15.9μm for corneal radius. The COV were 0.09%, 0.62%, and 0.21%, respectively. The reliability was 99.9% for AL, 97.8% for ACD, and 99.8%/99.5% for corneal radius(r₁/r₂).

CONCLUSION: PCI using the IOLMaster provides the more accurate and reliable anterior segment parameters measurement values. A high degree of agreement between US and IOLMaster was noted. The IOLMaster not only has the advantage of performing noncontact examinations, but also produces various additional data simultaneously and may thus obviate the need for multiple examinations.