INTRODUCTION

Glaucoma comprises a heterogeneous group of eye diseases. It leads to irreversible loss of ganglion cells and optic neuropathy with typical visual fi eld loss[1]. Despite the various medical and surgical treatment options, glaucoma is the third most common reason for blindness in the western countries[2] and the second most common reason worldwide[3].With regard to the pathogenesis of the disease, it has long been assumed that mechanical force resulting from increased intraocular pressure (IOP) is exerted on the peripapillary nerve fibers, depressing the fibers and leading to glaucomatous changes (mechanical hypothesis[4]). The vascular hypothesis,on the other hand, states that perfusion is disturbed by excessive IOP. However, 16% of all people have IOP＞22 mm Hg and do not develop any optic nerve damage. These findings indicate that there is a pressure-independent factor. Flammer and Weinreb[4] proposed that the change in blood flow is part of the cause and not, as long thought, a consequence of the disease.Examination of the ocular vasculature has always been dif fi cult. Many techniques have been devised to measure the hemodynamics of the eye. Due to the specific limitations of each technique, none of them completely satisfied the requirements[5-6]. With optical coherence tomography angiography (OCTA) it is now possible to assess the blood flow at the optic nerve head (ONH) non-invasively and quantitatively. In 2012, Jia et al[7] were the first to describe measurement of reduced vessel density in glaucomatous eyes by OCTA.

The purpose of our study was to validate the hypothesis of a difference in vessel density (VD) of the ONH between normal and glaucomatous eyes. Furthermore, we investigated the correlations of VD with various structural and functional parameters and rated the accuracy of OCTA in distinguishing between healthy and diseased eyes.

SUBJECTS AND METHODS

In this prospective monocentric study conducted at the Department of Ophthalmology, St. Franziskus Hospital Muenster (Germany), data from 97 glaucomatous eyes were evaluated. The healthy control group consisted of 74 eyes. The control subjects were either departmental staff or persons who attended for a routine ophthal mological examination.

The study was approved by the Ethics Committee of the Aerztekammer Westfalen-Lippe, Germany. All patients and volunteers were treated according to stipulations of the Helsinki Declaration and voluntarily participated in the study.Written informed consent was obtained from all participants.The glaucoma group and the control subjects both had a mean IOP of ＜21 mm Hg on the day of examination. The participants had no prior history of intraocular surgery, except for cataract extraction, and all were aged ＞18y. In addition, they all had a refractive error within ±6 D sphere and ±2 D cylinder. The control subjects had no family history of glaucoma, a normal appearing ONH, and intact neuroretinal rim and retinal nerve fiber layer (RNFL).

The inclusion criteria for the patient group were the diagnosis of glaucoma and glaucomatous optic neuropathy (cup:disc ratio ≥0.5) with corresponding RNFL defects detected by optical coherence tomography (OCT). Exclusion criteria in both groups were significant media opacity preventing highquality imaging, any ocular disease other than glaucoma or cataract, and previous intraocular operations other than cataract surgery. Persons with systemic hypertension, diabetes, or other vascular diseases such as status post heart failure, apoplexy, or thrombosis were also excluded. In particular, care was taken that no systemic drugs were taken, which could have an effect of vascular diameter either dilation or construction.

In addition to assessment of the relevant medical history and documentation of the current glaucoma therapy, all participants underwent a comprehensive ocular examination including bestcorrected visual acuity testing (BCVA), slit-lamp biomicroscopy,and IOP measurement (Goldmann applanation tonometry),plus visual field examination for the glaucoma patients (mode 30-2, Humphrey Field Analyzer; Zeiss, Jena, Germany). After pupil dilatation with tropicamide 0.5% and phenylephrine hydrochloride 2.5%, the subjects underwent binocular examination of the posterior segment. Subsequently, spectraldomain (SD)-OCT examination of the eye was performed using the OCT System AngioVue™ (RTVue-XR, Optovue,Inc.; Fremont; California, USA; software version 2016.2.035).Prior to the angiography we measured the RNFL, the ganglion cell complex (GCC) thickness, and the rim area at the ONH.The OCTA uses the split-spectrum amplitude-decorrelation angiography algorithm[8] to capture moving red blood cells.Two volumetric raster scans (one horizontal priority and one vertical priority) were obtained, each with a 4.5 mm×4.5 mm field of view centered on the ONH (302×302 pixels).

Statistical Analysis All analyses were performed using MedCalc® Version 12.4 (Ostend, Belgium) and R Version 3.2.5. (Dormagen, Germany). Normal distribution of the data was checked by means of the Kolmogorov-Smirnov test and shown as mean±standard deviation (SD; range) for Gaussian distributed values (t-test), and average medians (interquartile range) for non-Gaussian distribution (Wilcoxon rank sum test).Correlations were analyzed by Spearman’s rank correlation.To check potential discrimination between glaucomatous and control eyes, a receiver operating characteristics (ROC) curve was used. Moreover, we used Fisher’s linear discriminant analysis (LDA) to find a consensus classifier. LDA is an orthogonal transformation and data reduction technique to minimize intragroup variance and maximize intergroup variance.

近年来，地市级级媒体顺应时代的发展，不断改革创新，将工作的重点放在探索发展路径及创新体制机制上，使传统媒体与新兴媒体在渠道上、平台上、经营上及管理上均深度融合，并取得阶段性的成果。当前，媒体融合进入向纵深推进的关键阶段，媒体融合，不是选择，是趋势，更是使命。地市级媒体更应抢抓机遇，乘势而上，构建全媒体传播新格局。

Two segmentation levels were considered, designated by the manufacturer as “nerve head” (NH) segmentation (Figure 1A) and“radial peripapillary capillaries (RPC)” segmentation (Figure 1B). The NH segment extends from 2000 μm above the inner limiting membrane (ILM) to 150 μm below the ILM. The RPC segment is a slab from the ILM to the RNFL posterior boundary. All participants underwent SD-OCT and OCTA imaging on the same day. Poor-quality images-defined by a signal strength index (SSI) ≤40 were excluded from analysis.The AngioAnalytics™ software automatically calculates the VD by first extracting a binary image of the vessels from the gray scale of the en-face image and then computing the percentage of pixels of the vessels in the defined sectors or the entire en-face image based on the binary image. Thus, a quotient is formed from the fraction of the pixels with flux perception divided by all the detected pixels. The peripapillary region is defined as a 0.75-mm-wide elliptical annulus. The software calculates the value first as whole vessel density(wVD %) for the ONH and then divides it into an intrapapillary vessel density (iVD) value and an average peripapillary vessel density (pVD) value with further subdivision into six different peripapillary sectors based on the Garway-Heath map (Figure 2).

科学技术的不断发展，云计算技术和大数据平台的突飞猛进、电子商务的高速发展，让生活发生了巨大的改变，愈发凸显的智能化趋势，潜移默化的改变着人们的生活方式和思维方式。以智能家居行业为例，大量的智能单品、智能家居控制系统和智能家居设备，从概念走向了现实。

RESULTS

The glaucoma group comprised 97 eyes with an average age of 63.17±12.83y and the control group contained 74 eyes with an average age of 60.78±15.10y. There were no significant differences in age or refraction (all P＞0.05).

Since the first description by Jia et al[7] in 2012, there have been numerous publications on this topic. Starting from reports with a low number of cases[7,17-18], more and more studies are being carried out, con firming in the context of larger numbers of patients that VD is diminished in glaucomatous eyes[19-21].

With respect to GCC (Wang et al[27] r=0.4/our study rho 0.67)and RNFL (Wang et al[27] r=0.46/our study rho 0.72), Wang et al[27] found a high correlation but the results in our study are even better correlated. However, the correlation values with respect to rim area (Wang et al[27] r=0.147/our study rho 0.641) differ. Lévêque et al[28] also compared ONH perfusion with structural parameters and showed that these were correlated not only with RNFL and GCC but also with rim area. The differing results for correlation with rim area may be explained by different measuring methods. Lévêque et al[28] and our study group measured the rim area automatically with an OCT device, whereas Wang et al[27] do not specify the method used.Furthermore, our study demonstrates that measurement wVD and pVD shows high correlation with the severity of glaucoma,determined by the measurement of the GCC, RNFL, and rim area. This corresponds to recent publications such as that by Chen et al[29]. Some studies have already investigated the diagnostic utility of pVD measured by OCTA, usually compared with the diagnostic utility of the RNFL measured by OCT. Liu et al[18] found that the average pVD and the average thickness of the RNFL have similar sensitivity and speci ficity.These findings are supported by the results of Yamohammadi et al[30].

As expected, the glaucoma group had lower average values(P＜0.0001) for rim area, RNFL, and GCC than the control group (Table 1).

Table 1 Demographics and ocular characteristics of the study population mean±SD; median and interquartile range

Glaucoma group Control group P No. of eyes 97 74 Age (y) 63.17±12.83 60.78±15.10 0.5 Entity POAG: 41; PEX: 26; NTG:24; CPACG: 6 Time since first diagnosis of glaucoma (mo) 91.13±92.58 Refractive error (D) -0.44±2.27 -0.05±2.01 0.42 Pseudophakia (n) 32 14 Cup:disc ratio 0.69±0.18 0.27±0.16 ＜0.0001 Mean deviation in visual field (dB) -ber layer.

Figure 1 En-face OCT angiography of the ONH segmentation(A), the RPC segmentation (B) and the corresponding B-scan segmentations (C, D) of a right eye.

In both segmentation layers there was a statistically significant difference between glaucomatous and healthy eyes for all results of VD. In all tested sectors, the VD in glaucomatous eyes was significantly lower (P＜0.0001) than in the control group (Tables 2, 3). Comparing the numerical values of peripapillary VD in both segmentation layers, VD in the RPC layer was significantly higher than in the NH layer. The higher wVD value in the RPC layer is due to a low iVD value, which reduced the total wVD RPC.

Figure 2 OCTA of the ONH with sector classification of a right eye.

On analysis of the clinical findings in the glaucoma group(Table 4), wVD or pVD showed no significant correlation for refractive error and IOP. With regard to age, mean deviation in visual field testing, GCC, RNFL, and rim area, however, wVD and pVD were significantly correlated.

There was a strong correlation between the three parameters GCC/RNFL/rim area and wVD/pVD on both segmentation layers. Thinner RNFL, GCC, and rim area were correlated with lower VD (Figures 3, 4).

To check the accuracy of differentiation between eyes with and without glaucoma we calculated the ROC and the area under the curve (AUC). We initially evaluated the following six individual parameters: wVD (RPC/NH), iVD (RPC/NH), and pVD(RPC/NH) (Figure 4). The diagnostic accuracy was best for the iVD in the RPC layer (92.9%), followed by wVD RPC (84.0%),pVD Avg NH (79.4%), wVD NH (78.7%), iVD NH (76.7%),and pVD Avg RPC (74.9%). Moreover, the consensus classifi er analyzed using Fisher’s LDA increased the sensitivity and speci fi city to 94.2%.

Table 2 OCTA results for vessel density in the RPC layer mean±SD; median (interquartile range)

SSI: Signal strength index; RPC: Radial peripapillary capillaries; wVD: Whole vessel density; iVD:Intrapapillary vessel density; pVD: Peripapillary vessel density.

Segmentation RPC Glaucoma group Control group P SSI 54.52±8.97 61.06±11.12 0.009 wVD 46.66 (38.21, 52.35) 54.59 (50.77, 56.01) ＜0.0001 iVD 24.0 (16.50, 31.10) 43.29 (38.32, 52.58) ＜0.0001 pVD average 57.24 (46.93, 61.11) 61.58 (58.17, 63.85) ＜0.0001 pVD nas 53.10 (46.02, 59.07) 58.75 (55.18, 61.68) ＜0.0001 pVD infnas 56.30 (45.61, 63.81) 56.30 (45.61, 63.80) ＜0.0001 pVD inftemp 58.78 (44.64, 66.26) 66.03 (62.72, 68.55) ＜0.0001 pVD suptemp 57.49 (47.57, 68.77) 65.25 (60.83, 68.73) ＜0.0001 pVD supnas 51.36±11.01 59.22±5.73 ＜0.0001 pVD temp 60.66 (51.29, 64.17) 61.12 (58.33, 63.96) ＜0.0001

Table 3 OCTA results for vessel density in the NH mean±SD; median (interquartile range)

SSI: Signal strength index; NH: Nerve head; wVD: Whole nerve head vessel density; iVD: Intrapapillary vessel density; pVD: Peripapillary vessel density.

?

Table 4 Spearman correlations between wVD, pVD and clinical parameters

wVD: Whole vessel density; pVD: Peripapillary vessel density; RPC: Radial peripapillary capillaries; NH:Nerve head; IOP: Intraocular pressure; MD: Mean deviation; GCC: Ganglion cell complex; RNFL: Retinal nerve fiber layer; rho: Spearman’s correlation coefficient.

Clinical parametersw VD RPC (rho/P)pVD Avg RPC (rho/P)wVD NH (rho/P)pVD Avg NH (rho/P)Refractive error -0.05/0.65 -0.11/0.3001 -0.03/0.7604 -0.08/0.4202 IOP 0.01/0.96 -0.08/0.4130 -0.02/0.8075 -0.07/0.5044 Age -0.55/5.88×10-9 -0.51/7.63×10-8 -0.56/2.15×10-9 -0.50/1.08×10-7 MD 0.48/0.0004 0.50/0.0001 0.39/0.0039 0.46/0.00059 GCC total 0.75/1.66×10-18 0.72/9.05×10-17 0.67/7.33×10-14 0.68/1.33×10-14 RNFL average 0.81/9.68×10-24 0.76/1.41×10-19 0.72/6.37×10-17 0.73/1.99×10-17 Rim area 0.70/9.00×10-16 0.63/6.89×10-12 0.61/2.60×10-11 0.64/1.93×10-12

To get more detailed results, we increased the number of individual features to 14. Again using Fisher’s LDA as linear classifier, this yielded a 14-dimensional feature vector of 97 glaucomatous and 76 healthy eyes. Table 5 shows the statistics and performances of the single features and the linear classifier.The P-value was computed using a single-sided t-test. The LDA weights for each single feature were the weighting terms for computing the consensus feature. In the column “AUC ROC”the area under the ROC curve is given. Another characteristic to measure the performance is the true-positive rate (TPR) at a significance level of 5%. In the last column of the table the minimal error number for the classifier is presented.

Figure 3 Scatter plots for average pVD and ONH characteristics A thinner RNFL or GCC correlates significantly with a lower pVD.

Table 5 Linear classifier of 14 dimensional features using Fisher’s LDA

ROC: Receiver operating characteristic; AUC: Area under curve; LDA: Linear discriminant analysis (Fisher); TPR: True-positive rate; FPR:False-positive rate; iVD: Intrapapillary vessel density; pVD: Peripapillary vessel density; RPC: Radial peripapillary capillaries; NH: Nerve head.

No. of minimal errors Consensus -2.0±1.2 1.3±1.4 9.9×10-35 95.6 84.5 17 iVD RPC 45.1%±9.3% 23.9%±10.5% 4.5×10-28 -0.53 92.9 74.2 20 pVD NH inftemp 63.6%±4.1% 54.5%±10.6% 1.4×10-12 0.36 77.8 51.5 45 pVD RPC inftemp 65.4%±4.3% 55.5%±12.4% 1.7×10-11 -0.38 75.4 50.5 48 pVD RPC supnas 59.2%±5.7% 51.4%±11.0% 9.1×10-9 -0.15 70.9 39.2 54 pVD RPC infnas 62.5%±6.6% 54.3%±11.4% 2.5×10-8 0.23 71.1 38.1 55 pVD NH infnas 61.1%±6.6% 52.9%±10.4% 3.4×10-9 -0.21 74.6 36.1 51 pVD NH supnas 58.2%±6.4% 49.9%±11.0% 3.9×10-9 0.14 72.5 35.1 53 pVD RPC nas 57.9%±5.5% 51.3%±8.9% 1.7×10-8 -0.00 73.5 32.0 51 pVD NH nas 57.6%±5.4% 49.8%±8.6% 2.7×10-11 -0.24 78.7 32.0 42 pVD NH temp 57.9%±5.5% 52.1%±8.6% 3.2×10-7 -0.17 69.8 28.9 54 iVD NH 51.6%±6.2% 44.7%±6.8% 2.6×10-10 0.30 76.7 27.8 47 pVD RPC temp 60.5%±4.6% 56.8%±8.8% 2.7×10-4 0.36 60.1 26.8 66 pVD RPC suptemp 64.0%±6.8% 56.6%±12.3% 9.0×10-7 0.06 68.6 25.8 55 pVD NH suptemp 61.7%±7.3% 54.0%±11.0% 1.3×10-7 0.03 71.7 25.8 50 Feature Glaucoma group Control group P LDA weight AUC ROC(%)TPR (%)FPR=5%

Figure 4 Comparison between whole vessel density, inferior vessel density and peripapillary vessel density in the radial papillary capillaries layer and the nerve head layer in terms of ROC curve,AUC, and in addition, LDA as feature reduction method.

As shown in Table 5 the consensus classifier of the 14 features achieved a good performance for the discrimination of glaucoma and normal eyes. At a significance level of 5%, a TPR of 84.5% was achieved. The area under the ROC curve was 95.6%. The importance of this classifier for the consensus classifier is confirmed by the highest absolute weight of -0.53 for computing the consensus and lowest P-value of the single classifier.

DISCUSSION

此外，在服务意识为导向的内部管理环境建设过程中，还应加强信息化管理手段升级及高校新文化建设工作开展力度。信息化管理手段升级能够确保管理层更加及时地进行财务数据收集与分析，确保管理与实际活动开展始终保持一致，避免管理层与执行层因为信息不对称而产生的矛盾冲突。高校新文化建设则应致力于将经费管理制度与具体执行细节融入文化建设与意识培养之中，将制度内化为一种文化，让管理人员与执行人员从被动执行转为主动参与，从而大大提升他们在制度执行方面的主动性、自觉性以及创新性，充分发挥其主观能动性。

Figure 5 Magnified temporal sector A: NH segmentation; B: RPC segmentation; C: Difference RPC-NH; D: Black pixel: RPC at least 10 gray levels brighter than NH; white pixel: NH at least 10 gray levels brighter than RPC; gray pixel: NH and RPC have same value ±10 gray levels.

The OCTA with SSADA has great advantages over previous examination methods. It offers non-invasive, high-resolution depiction of the vascular structure in three dimensions.OCTA is not very sensitive to the reflected signal of nonmoving tissue and therefore gives a good representation of the individual vessel density[7]. Taking advantage of light absorption by the erythrocytes, AngioVue™ scans the areas of interest several times. Recording of the moving erythrocytes allows three-dimensional representation of the blood flow and the network of the capillaries in the retina. It supplies scans sized 3 mm×3 mm or 4.5 mm×4.5 mm for the optic nerve. The analysis is presented together with the structural OCT B-scans and the en-face image of the same data set. The details have been described previously[7].

自此，盘山的名胜、人物、诗文、建置、物产因此稿得以记载。 而盘山有此志，与其他名川大山争雄有据，再无愧色。 《盘山志》既是史实记载，也是文学汇编，具有藏之名山的价值，是智朴大师的功绩。 《盘山志》因此流传开来，成为人们争相阅读的书目，在当时文人圈子里产生了不小的影响，由此对于智朴与方内人物的交游也起到推动作用。

With regard to the types of glaucoma, 41 eyes had primary open angle glaucoma (POAG), 26 eyes had pseudoex foliation(PEX) glaucoma, 24 eyes had normal-tension glaucoma(NTG), and 6 eyes had chronic primary angle-closure glaucoma(CPACG). Further clinical data are presented in Table 1.

In order to better understand the pathogenesis of glaucoma and to investigate the hypothesis of a vascular genesis[9], the blood circulation of the optic nerve head has been investigated with various examination techniques, including fluorescein angiography[10-11], laser Doppler flowmetry[12-13], Heidelberg retina flowmetry[14], and the Retina Vessel Analyzer[15]. Since these methods are partly invasive procedures with potential adverse effects[16] and are also not always available for routine use they have not become established as standard examinations for glaucoma patients.

By analyzing the VD separately in different sectors and layers of the optic nerve, OCTA adds a valuable diagnostic tool to distinguish a healthy optic nerve from a damaged one. In our study we found that the iVD of the RPC layer has the best diagnostic validity for this purpose (Figure 4). Also, the wVD in both the RPC layer and the NH layer are significantly different in healthy and glaucomatous eyes. Especially the significant difference of the iVD in the RPC layer between control and glaucoma eyes is noticeable. This finding may possibly be explained by the typical excavation form of a glaucomatous ONH. Bayoneting of circumlinear vessels occurs in areas where neuroretinal rim tissue has been lost and nasalization of blood vessels occurs in very advanced glaucoma[23]. The blood vessels run through a glaucomatous ONH with a steep gradient, so that the erythrocytes are moving nearly perpendicular to the OCT optics. This kind of flow cannot be detected reliably by state-of-the-art OCTA devices.Due to these effects the average iVD RPC of glaucoma eyes is not much more than half of the normal density (24.0% in glaucoma eyes vs 43.3% in control eyes; Table 2). Thus, the highest diagnostic accuracy of intrapapillary values could be an artifact. However, the wVD in the RPC layer, next in the evaluation (Figure 4), also has an excellent accuracy with 84.0%.In order to examine this observation in more detail, we looked at 14 different parameters using Fisher’s LDA. Table 5 shows the statistics and performances of the single features and the linear classifier. All 14 features achieved good performance with regard to discrimination of glaucomatous and normal eyes. Once again, the feature iVD in the RPC layer attained the highest diagnostic value. Standing alone, it was nearly as good as the consensus of all features taken together. The importance of this classifier for the consensus classifier is con fi rmed by the highest absolute weight of -0.53 for computing the consensus and lowest P-value of the single classifier. No other classifier reached the performance of the iVD feature. Although the P-values are below 0.05, the AUC, the TPR (at 5% significance level) and the minimal error value performed notably worse than the best feature, iVD. Again, it must be critically noted that there may have been measurement error due to the recording technology. However, the inferior temporal pVD, in particular, shows a high degree of accuracy in distinguishing glaucomatous eyes from healthy ones.

Comparing the two segmentation layers, we found higher VD for all of the six peripapillary sectors in the RPC layer. This result is inconsistent with the visual inspection, which suggests higher VD in the NH layer (Figure 5C, 5D). Furthermore, the RPC segmentation is included within the NH segmentation.The average pixel density, however, is slightly higher within the RPC (74.6 vs 73.1; computed using the Fiji software).

A potential explanation for the lower VD in the NH layer is that the en-face image is computed by agglomerating a thicker slice (Figure 1C, 1D). This leads to a lower pixel intensity(= flow) for thin structures (yelloWellipse, Figure 1A, 1B and Figure 5) in the NH layer than in the RPC layer. Computing the flow area, thin structures have a higher contrast on RPC,which gives a more detailed segmentation of the flow area.Thus AngioVue™ computes higher flow area and density,even though physically less flow is agglomerated in the RPC segmentation and both layers (RPC and NH) have the same size.To obtain a linear correlation between several variables we used the Pearson correlations (Table 4). According to Cohen’s classification[25] neither the IOP nor the refractive error correlated with wVD/pVD. For age and MD we found a moderate correlation with wVD/pVD, while GCC total, RNFL average, and rim area showed a high correlation.

With regard to MD in visual field examination, we had expected a high correlation, as preliminary descriptions showed a significant relationship between wVD/pVD and severity of visual field damage independent of structural loss[26]. Our results showed only a moderate correlation (Table 4). In the RPC layer there is a slightly better correlation between VD in OCTA and MD in visual fi eld testing. Comparing our values for the wVD NH layer with the results of Wang et al[27] the results are similar (Wang r=0.404/our study pVD NH rho 0.39). Liu et al[18] demonstrated a stronger correlation of the peripapillary VD with MD, whereby their work was based on a small sample (n=12).

与此同时，经济增长同样是天津市物流业能耗增长最重要的推动因素，推动作用一直比较强，因此，伴随城镇化的发展，合理引导居民消费，对天津市物流业能耗下降具有积极作用。此外，与北京市不同，空间结构对天津市物流业能耗增长具有较大的推动作用，这说明强化天津市各区功能定位规划，合理布局城市空间结构对物流业能耗降低具有一定的积极作用，而交通基础设施建设对物流业能耗的增长起到了重要的推动作用。

在航空航天用柔性复合材料技术基础上，NASA LRC还与美国农业部森林服务处 (USDAFS)合作，正在开发一种能够轻便携行，并在火灾中为消防员提供保护屏障的应急防火棚CHIEFS(Convective Heating Improvement for Emergency Fire Shelters)。

In our study we examined a total of 97 glaucomatous eyes.We found that the VD in the radial papillary capillaries layer and in the nerve head layer of glaucomatous eyes was significantly lower than in age-matched control eyes. This is in agreement with our group’s earlier study, which, however,had a smaller number of participants[22]. In contrast to our previous work, this time we measured the VD in six different peripapillary sectors. Clinically, glaucomatous damage occurs predominantly in the temporal superior and temporal inferior sectors of the nerve head and advanced disease leads to nasal shifting of central vessels[23]. For this reason we had expected the lowest VD in the temporal sector. However, our OCTA results show lower pVD in the nasal and nasal superior ONH sectors compared with the other five sectors (all P≤1.5×10-4).The VD is also significantly reduced in these two sectors compared with healthy eyes (P＜0.0001). If the VD quotient nasal/temporal is formed for both segmentation levels (RPC and NH), the glaucomatous eyes have a greater reduction of flow than the healthy ones (RPC: P=0.0002; NH: P=0.010).Differing results have been described in the literature. Jia et al[7] reported a greater reduction of VD in the temporal ONH sector, whereas Rao et al[24] did not fi nd a diminished VD in the temporal sector compared with the entire ONH. We currently have no explanation for the different results; therefore, further investigations are necessary.

2013年6月,习近平总书记在全国组织工作会议上强调：面对复杂多变的国际形势和艰巨繁重的国内改革发展任务,关键在党,关键在人,关键在于建设一支宏大的高素质干部队伍。为此,我们必须解决好“怎样是好干部,怎样成长为好干部,怎样把好干部用起来”这三个问题,按照“信念坚定、为民服务、勤政务实、勇于担当、清正廉洁”五条标准,培养选拔党和人民需要的好干部。习近平总书记提出的“20字标准”便成为新时期好干部的主要内容,成为好干部成长的方向和选拔任用考核的指挥棒。中国传统文化关于从政者正己修身的教育资源十分丰富，且具有强大生命力，对于党员领导干部对标新时代好干部“20字标准”仍有重要的指导意义。

Our study has limitations. It does not show whether the diminished VD is the cause or the consequence of the glaucomatous ONH damage. A study by Holló[31] showed that peripapillary angioflow density measurements can identify decreased peripapillary perfusion in early glaucoma prior to the development of significant RNFL damage and visual field deterioration. We did not perform any blood pressure measurements during the OCTA scanning. A connection between blood pressure and the ONH blood flow is conceivable. In previous studies, however, no significant correlation between blood pressure and ONH VD was shown[17-18].

3）课后阶段。教师要求学生对本次教学内容进行总结归纳。比如采用表格的形式将各种单相整流电路的电路拓扑、控制角的移相范围、输出电压和电流的计算公式等一一列出，并提交到平台上；并完成课后自测，进一步将知识内化。此外，学有余力的学生可以结合教师提供的仿真模型，通过改变电路和参数，对单相半控及不同负载等进行仿真实验研究，获得拓展提高的机会。

A further limitation is the use of topical glaucoma treatment.Studies using older methods to measure the blood flow under topical medication have yielded varying results. Fuchsjäger-Mayrl et al[32], for example, showed improved blood flow and local carbohydrate inhibitors, whereas Pillunat et al[33] did not demonstrate any change. Takusagawa et al[34] even found an association between β-blocker eyedrop use and decreased macular VD in glaucomatous eyes measured by OCTA.Ninety-two percent of our glaucoma patients were using IOP-lowering eyedrops, and for ethical and medical reasons we did not discontinue the local medication before the study.

Furthermore, a limitation on the use of phenylephrine for pupil dilation has to be discussed. It is a selective α1-adrenergic receptor agonist with vasoconstrictive effect. In a study by Vandewalle et al[35] however, it has been shown that addition of phenyephrine (in this work even at 5% concentration) to tropicamide 0.5% does not in fl uence retinal vessel diameter in glaucoma patients. Nevertheless should avoid using vasoactive substances in future work whenever possible.

Our work demonstrates that spectral domain OCTA can measure reduced vessel density at the optic nerve head in glaucomatous eyes and that these findings are correlated to functional and structural glaucomatous alterations. In addition,the further the disease has advanced, the more the vessel density is reduced. To the best of our knowledge, there are currently no similar studies investigating the possibility of distinguishing between healthy and glaucomatous eyes in detail. On the basis of 14 different features it was shown that the OCTA technique is able to distinguish glaucomatous from healthy eyes with high reliability.

ACKNOWLEDGEMENTS

Conflicts of Interest: Lommatzsch C, lecture Optovue;Rothaus K, None; Koch JM, None; Heinz C, None; Grisanti S, None.

（二）让学生大胆尝试“小老师”的味道，创设学生主动参与教学的情景。成功是求知的有力支柱。在教学中的某个片段，教师可鼓励学生担任“小老师”，帮助老师进行辅导和帮助学习有困难的学生。这样训练了“小老师”们的逻辑表达能力，学生们也对所学知识和问题得到进一步的理解，尝试到了学习的甜头，增强了求知的欲望。

REFERENCES

1 Weinreb RN, KhawPT. Primary open-angle glaucoma. Lancet 2004;363(9422):1711-1720.

2 Dietlein TS, Hermann MM, Jordan JF. The medical and surgical treatment of glaucoma. Dtsch Arztebl Int 2009;106(37):597-605.

3 Resnikoff S, Pascolini D, Etya’ale D, Kocur I, Pararajasegaram R,Pokharel GP, Mariotti SP. Global data on visual impairment in the year 2002. Bull World Health Organ 2004;82(11):844-851.

4 Fechtner RD, Weinreb RN. Mechanisms of optic nerve damage in primary open angle glaucoma. Surv Ophthalmol 1994;39(1):23-42.

5 Harris A, Chung HS, Ciulla TA, Kagemann L. Progress in measurement of ocular blood flow and relevance to our understanding of glaucoma and age-related macular degeneration. Prog Retin Eye Res 1999;18(5):669-687.6 Schuman JS. Measuring blood flow: so what? JAMA Ophthalmol 2015;133(9):1052-1053.

7 Jia Y, Morrison JC, Tokayer J, Tan O, Lombardi L, Baumann B, Lu CD,Choi W, Fujimoto JG, Huang D. Quantitative OCT angiography of optic nerve head blood flow. Biomed Opt Express 2012;3(12):3127-3137.

8 Jia Y, Tan O, Tokayer J, Potsaid B, Wang Y, Liu JJ, Kraus MF, Subhash H, Fujimoto JG, Hornegger J, Huang D. Split-spectrum amplitudedecorrelation angiography with optical coherence tomography. Opt Express 2012;20(4):4710-4725.

9 Flammer J, Orgül S, Costa VP, Orzalesi N, Krieglstein GK, Serra LM,Renard J-P, Stefánsson E. The impact of ocular blood flow in glaucoma.Prog Retin Eye Res 2002;21(4):359-393.

10 Arend O, Plange N, Sponsel WE, Remky A. Pathogenetic aspects of the glaucomatous optic neuropathy: fluorescein angiographic findings in patients with primary open angle glaucoma. Brain Res Bull 2004;62(6):517-524.

11 Talusan E, Schwartz B. Speci fi city of fl uorescein angiographic defects of the optic disc in glaucoma. Arch Ophthalmol 1977;95(12):2166-2175.

12 Piltz-seymour JR, Grunwald JE, Hariprasad SM, Dupont J. Optic nerve blood flow is diminished in eyes of primary open-angle glaucoma suspects. Am J Ophthalmol 2001;132(1):63-69.

13 Hamard P, Hamard H, Dufaux J, Quesnot S. Optic nerve head blood flow using a laser Doppler velocimeter and haemorheology in primary open angle glaucoma and normal pressure glaucoma. Br J Ophthalmol 1994;78(6):449-453.

14 Logan JFJ, Rankin SJA, Jackson AJ. Retinal blood flow measurements and neuroretinal rim damage in glaucoma. Br J Ophthalmol 2004;88(8):1049-1054.

15 Ramm L, Jentsch S, Peters S, Augsten R, Hammer M. Investigation of blood flow regulation and oxygen saturation of the retinal vessels in primary open-angle glaucoma. Graefes Arch Clin Exp Ophthalmol 2014;252(11):1803-1810.

16 Stein MR, Parker CW. Reactions following intravenous fluorescein.Am J Ophthalmol 1971;72(5):861-868.

17 Jia Y, Wei E, Wang X, Zhang X, Morrison JC, Parikh M, Lombardi LH, Gattey DM, Armour RL, Edmunds B, Kraus MF, Fujimoto JG,Huang D. Optical coherence tomography angiography of optic disc perfusion in glaucoma. Ophthalmology 2014;121(7):1322-1332.

18 Liu L, Jia Y, Takusagawa HL, Pechauer AD, Edmunds B, Lombardi L, Davis E, Morrison JC, Huang D. Optical coherence tomography angiography of the peripapillary retina in glaucoma. JAMA Ophthalmol 2015;133(9):1045-1052.

19 Shin JW, Lee J, Kwon J, Choi J, Kook MS. Regional vascular densityvisual field sensitivity relationship in glaucoma according to disease severity. Br J Ophthalmol 2017;101(12):1666-1672.

20 Rao HL, Kadambi SV, Weinreb RN, Puttaiah NK, Pradhan ZS,Rao DAS, Kumar RS, Webers CAB, Shetty R. Diagnostic ability of peripapillary vessel density measurements of optical coherence tomography angiography in primary open-angle and angle-closure glaucoma. Br J Ophthalmol 2017;101(8):1066-1070.

21 Suh MH, Zangwill LM, Manalastas PI, Belghith A, Yarmohammadi A, Medeiros FA, Diniz-Filho A, Saunders LJ, Yousefi S, Weinreb RN. Optical coherence tomography angiography vessel density in glaucomatous eyes with focal lamina cribrosa defects. Ophthalmology 2016;123(11):2309-2317.

22 Lommatzsch C, Koch JM, Claußnitzer H, Heinz C. OCT angiography of the glaucoma optic nerve. Klin Monbl Augenheilkd 2018;235(2):205-211.

23 Gandhi M, Dubey S. Evaluation of the optic nerve head in glaucoma. J Curr Glaucoma Pract 2013;7(3):106-114.

24 Rao HL, Pradhan ZS, Weinreb RN, Reddy HB, Riyazuddin M, Dasari S,Palakurthy M, Puttaiah NK, Rao DA, Webers CA. Regional comparisons of optical coherence tomography angiography vessel density in primary open-angle glaucoma. Am J Ophthalmol 2016;171:75-83.

25 Cohen J. Statistical Power Analysis for the Behavioral Sciences. 2.edition. Hillsdale: Lawrence Erlbaum Associates; 1988.

26 Yarmohammadi A, Z angwill LM, Diniz-Filho A, Suh MH, Yousefi S, Saunders LJ, Belghith A, Manalastas PI, Medeiros FA, Weinreb RN.Relationship between optical coherence tomography angiography vessel density and severity of visual field loss in glaucoma. Ophthalmology 2016;123(12):2498-2508.

27 Wang X, Jiang C, Ko T, Kong X, Yu X, Min W, Shi G, Sun X. Correlation between optic disc perfusion and glaucomatous severity in patients with open-angle glaucoma: an optical coherence tomography angiography study. Graefes Arch Clin Exp Ophthalmol 2015;253(9):1557-1564.

28 Lévêque PM, Zéboulon P, Brasnu E, Baudouin C, Labbé A. Optic disc vascularization in glaucoma: value of spectral-domain optical coherence tomography angiography. J Ophthalmol 2016;2016:6956717.

29 Chen CL, Bojikian KD, Gupta D, Wen JC, Zhang Q, Xin C, Kono R,Mudumbai RC, Johnstone MA, Chen PP, Wang RK. Optic nerve head perfusion in normal eyes and eyes with glaucoma using optical coherence tomography-based microangiography. Quant Imaging Med Surg 2016;6(2):125-133.

30 Yarmohammadi A, Zangwill LM, Diniz-Filho A, Suh MH, Manalastas PI, Fatehee N, Youse fi S, Belghith A, Saunders LJ, Medeiros FA, Huang D,Weinreb RN. Optical coherence tomography angiography vessel density in healthy, glaucoma suspect, and glaucoma eyes. Invest Ophthalmol Vis Sci 2016;57(9):OCT451-459.

31 Holló G. Vessel density calculated from OCT angiography in 3 peripapillary sectors in normal, ocular hypertensive, and glaucoma eyes.Eur J Ophthalmol 2016;26(3):e42-e45.

32 Fuchsjäger-Mayrl G, Wally B, Rainer G, Buehl W, Aggermann T,Kolodjaschna J, Weigert G, Polska E, Eichler HG, Vass C, Schmetterer L.Effect of dorzolamide and timolol on ocular blood flow in patients with primary open angle glaucoma and ocular hypertension. Br J Ophthalmol 2005;89(10):1293-1297.

33 Pillunat LE, Böhm AG, Köller AU, Schmidt KG, Klemm M, Richard G. Effect of topical dorzolamide on optic nerve head blood flow. Graefes Arch Clin Exp Ophthalmol 1999;237:495-500.

34 Takusagawa HL, Liu L, Ma KN, Jia Y, Gao SS, Zhang M, Edmunds B,Parikh M, Tehrani S, Morrison JC, Huang D. Projection-resolved optical coherence tomography angiography of macular retinal circulation in glaucoma. Ophthalmology 2017;124(11):1589-1599.

35 Vandewalle E, Abegão Pinto L, Olafsdottir OB, Stalmans I. Phenylephrine 5% added to tropicamide 0.5% eye drops does not influence retinal oxygen saturation values or retinal vessel diameter in glaucoma patients.Acta Ophthalmol 2013;91(8):733-737.