Abstract：
*Correspondence, Stephen Kaye: s.b.kaye@liverpool.ac.uk PURPOSE: To assess the accuracy, repeatability, and performance limits of in vivo Mirau ultrahigh axial resolution (UHR) line field spectral domain (LF-SD) optical coherence tomography (OCT) for the clinical in vivo measurement of Bowman\'s layer thickness in subjects with and without keratoconus.
METHODS: Patients with keratoconus and volunteers with no corneal disease were included. The thickness of Bowman\'s layer was measured in the clinic. An in vivo graph search image segmentation of the central cornea was obtained at the normal interface vector orientation. The Mirau-UHR-LF-SD-OCT system used has an axial resolution down to 2.4 µm in air (1.7 µm in tissue), with an A-scan speed of 204.8 kHz and a signal to noise ratio (sensitivity) of 69 (83) dB.
RESULTS: 40 patients with keratoconus and 20 healthy volunteers were included. The repeatability of mean Bowman\'s and epithelial thicknesses were 0.3 and 1.0 µm, respectively. The measured 95% population range for Bowman\'s layer thickness was 13.7 to 19.6 µm for healthy (mean 16.65, SD 1.48) and 10.94 to 16.99 for 23 of the keratoconics (mean 13.96 SD 1.51) (p<0.05).
CONCLUSIONS: The measured thicknesses of Bowman\'s layer using the Mirau-UHR-LF-SD-OCT were both accurate, with the range for healthy in vivo thicknesses matching prior confocal and OCT systems of varying axial resolutions and repeatable. Bowman\'s layer was significantly thinner in patients with keratoconus. Bowman\'s layer can be accurately measured in the clinical setting using a Mirau-UHR-LF-SD-OCT and can be useful for disease monitoring.
METHODS: Patients with keratoconus and volunteers with no corneal disease were included. The thickness of Bowman\'s layer was measured in the clinic. An in vivo graph search image segmentation of the central cornea was obtained at the normal interface vector orientation. The Mirau-UHR-LF-SD-OCT system used has an axial resolution down to 2.4 µm in air (1.7 µm in tissue), with an A-scan speed of 204.8 kHz and a signal to noise ratio (sensitivity) of 69 (83) dB.
RESULTS: 40 patients with keratoconus and 20 healthy volunteers were included. The repeatability of mean Bowman\'s and epithelial thicknesses were 0.3 and 1.0 µm, respectively. The measured 95% population range for Bowman\'s layer thickness was 13.7 to 19.6 µm for healthy (mean 16.65, SD 1.48) and 10.94 to 16.99 for 23 of the keratoconics (mean 13.96 SD 1.51) (p<0.05).
CONCLUSIONS: The measured thicknesses of Bowman\'s layer using the Mirau-UHR-LF-SD-OCT were both accurate, with the range for healthy in vivo thicknesses matching prior confocal and OCT systems of varying axial resolutions and repeatable. Bowman\'s layer was significantly thinner in patients with keratoconus. Bowman\'s layer can be accurately measured in the clinical setting using a Mirau-UHR-LF-SD-OCT and can be useful for disease monitoring.
摘要：
*通信，斯蒂芬·凯：s.b.凯@利物浦。AC.目的：为了评估准确性，重复性,以及体内Mirau超高轴向分辨率（UHR）线场谱域（LF-SD）光学相干断层扫描（OCT）的性能极限，用于有和没有圆锥角膜的受试者的Bowman层厚度的临床体内测量。
方法：包括圆锥角膜患者和无角膜疾病的志愿者。在临床上测量了Bowman层的厚度。在正常界面矢量方向上获得了中央角膜的体内图搜索图像分割。使用的Mirau-UHR-LF-SD-OCT系统在空气中的轴向分辨率低至2.4µm（在组织中为1.7µm），A扫描速度为204.8kHz，信噪比（灵敏度）为69（83）dB。
结果：纳入40例圆锥角膜患者和20例健康志愿者。平均Bowman's和上皮厚度的可重复性为0.3和1.0µm，分别。测量的Bowman层厚度的95％人群范围为健康者为13.7至19.6µm（平均16.65，SD1.48），而23例角膜角膜炎患者为10.94至16.99（平均13.96SD1.51）（p<0.05）。
结论：使用Mirau-UHR-LF-SD-OCT测得的Bowman层厚度均准确，与健康的体内厚度范围相匹配，现有的共聚焦和OCT系统具有不同的轴向分辨率和可重复性。圆锥角膜患者的Bowman层明显变薄。Bowman层可以在临床环境中使用Mirau-UHR-LF-SD-OCT进行精确测量，可用于疾病监测。
方法：包括圆锥角膜患者和无角膜疾病的志愿者。在临床上测量了Bowman层的厚度。在正常界面矢量方向上获得了中央角膜的体内图搜索图像分割。使用的Mirau-UHR-LF-SD-OCT系统在空气中的轴向分辨率低至2.4µm（在组织中为1.7µm），A扫描速度为204.8kHz，信噪比（灵敏度）为69（83）dB。
结果：纳入40例圆锥角膜患者和20例健康志愿者。平均Bowman's和上皮厚度的可重复性为0.3和1.0µm，分别。测量的Bowman层厚度的95％人群范围为健康者为13.7至19.6µm（平均16.65，SD1.48），而23例角膜角膜炎患者为10.94至16.99（平均13.96SD1.51）（p<0.05）。
结论：使用Mirau-UHR-LF-SD-OCT测得的Bowman层厚度均准确，与健康的体内厚度范围相匹配，现有的共聚焦和OCT系统具有不同的轴向分辨率和可重复性。圆锥角膜患者的Bowman层明显变薄。Bowman层可以在临床环境中使用Mirau-UHR-LF-SD-OCT进行精确测量，可用于疾病监测。
