Comparative Analysis of Optical Coherence Tomography and Ultrasound Biomicroscopy Imaging Diagnosis of Anterior Segment of Keratoconus in Acute Stage

Background To compare and analyze the imaging differences between anterior segment optical coherence tomography (AS-OCT) and ultrasound biomicroscopy (UBM) in acute keratoconus. Method The features of AS-OCT and UBM images of acute keratoconus patients in Hebei Eye Hospital were analyzed retrospectively. Result From May 2016 to December 2020, 38 patients (38 eyes) with acute keratoconus were treated in Hebei Eye Hospital. Both AS-OCT and UBM images showed rupture of the posterior elastic layer of cornea, edema and thickening of stroma layer, and cracks formed in it to communicate with anterior chamber. AS-OCT image showed that corneal epithelium was intact, and 26 cases had subcutaneous effusion. UBM images showed delamination of the posterior elastic layer interface in 26 cases. Paired t test showed that there was signicant difference between preoperative and postoperative best corrected visual acuity in all patients (t = 11.861, P < 0.05). There was no signicant difference in preoperative and postoperative best corrected visual acuity between group 1 and group 2 by independent sample t test (t = 0.751, P> 0.05, t = 0.219, P > 0.05). There was signicant difference in the height of rupture between group 1 and group 2 (t =-7.533, P < 0.001). Pearson correlation analysis showed that there was a signicant correlation between fracture size and fracture height (r = 0.766, P < 0.001), but no correlation between fracture size and post-elastic layer delamination (r =-0.107, P > 0.05). No corneal perforation was found in 38 cases (38 eyes). In the healing stage after deep lamellar keratoplasty, obvious stroma scar appeared, accompanied by corneal stroma thickening and corneal attening. and UBM can complement each other, provide accurate corneal imaging images and parameter measurement, predict evaluate the of the disease.

0.751, P> 0.05, t = 0.219, P > 0.05). There was signi cant difference in the height of rupture between group 1 and group 2 (t =-7.533, P < 0.001). Pearson correlation analysis showed that there was a signi cant correlation between fracture size and fracture height (r = 0.766, P < 0.001), but no correlation between fracture size and post-elastic layer delamination (r =-0.107, P > 0.05). No corneal perforation was found in 38 cases (38 eyes). In the healing stage after deep lamellar keratoplasty, obvious stroma scar appeared, accompanied by corneal stroma thickening and corneal attening.
Conclusion AS-OCT and UBM can complement each other, provide accurate corneal imaging images and parameter measurement, and predict and evaluate the development of the disease.

Background
Keratoconus is a keratopathy characterized by corneal dilation, which causes the central part of the cornea to bulge forwards and become thinner into a cone shape and produces highly irregular astigmatism. The disease mostly occurs in teenagers, and progressive onset is often noticed in both eyes successively. Acute hydrops of keratoconus (ACUTE HYDROPS) is a special condition of keratoconus, and the incidence rate is about 3% [1]. Suddenly, the vision is often decreased in the late stage of keratoconus, and the central cornea is obviously edematous and cloudy, with a large number of blisters under the skin. The obvious edematous person is the water-drop anterior process in the central cornea (Figs. <link rid=" g1">1</link>-A and 1-B). At present, there are few studies on the anatomical features of keratoconus in the acute phase, and even fewer studies on the anatomical features using comparative analysis of AS-OCT and UBM images. In this study, AS-OCT and UBM were used to nd out that the images of keratoconus in the acute phase both showed corneal posterior elastic layer rupture and discontinuous echo. Locally obvious thickening of the matrix layer; The stromal layers of the corresponding parts formed vertical fractures with different sizes to communicate with the anterior chamber. The thickness of the stromal layer was thin, and no de nite perforation was seen. AS-OCT image can show corneal epithelium; The UBM image shows the interfacial delamination of the corneal posterior elastic layer, considering the possibility of separation of the corneal posterior elastic layer from the endodermis.  axial not more than 0.04mm, lateral not more than 0.04 mm. Digital scan converter (DSC) capacity: 1024×768×8bit×64; Gray scale: 256 levels; Focal length: 9 mm; Window display (width×depth): 8.25mm×5.5mm (narrow view mode) Delay depth 3mm, 16.5mm×11mm (panoramic mode) Delay depth 6 mm. The patient lay at on the hard board bed in the indoor natural light, and the appropriate optic cup was selected and placed in the eyelid of the patient. An appropriate amount of sterile normal saline was poured into the eyelid, and the UBM probe was placed in a suitable position in the eye cup for image scanning. The scanned images with good image de nition and position were collected to complete the corresponding data measurement and stored.
1.4 Statistical analysis:Statistical analysis was performed using SPSS23.0 statistical software. The enumeration data were expressed as percentages, and the measurement data of normal distribution were expressed as mean addition and subtraction of standard deviation ( X ± S). The one-way analysis of variance(one-way ANOVA) was used for intra-group comparison, and paired t-test was used for intergroup comparison. Test level: α = 0.05.

Results
2.1 General results There were 33 males and 5 females. 18 in the right eye and 20 in the left eye. The maximum age was 34 years old and the minimum age was 11 years old, with an average age of 21.2 years old. Vision: there were 12 cases with manual operation at present, 19 cases with immediate index, and 7 cases with 0.01-0.1, none of which was improved after refraction correction. All the 38 patients presented with sudden visual acuity decline, and six of them presented with sudden visual acuity decline after hard eye rolling. Five patients presented with an acute attack of keratoconus in one eye of both eyes, one of whom had undergone corneal transplantation in the contralateral eye; Six cases had myopic history of both eyes. The AS-OCT and UBM images of keratoconus in the acute phase are characterized by corneal posterior elastic layer rupture and discontinuous echo. Locally obvious thickening of the matrix layer; The stromal layers of the corresponding parts formed vertical fractures with different sizes to communicate with the anterior chamber. The thickness of the stromal layer was thin, and no de nite perforation was seen. AS-OCT images showed intact corneal epithelium and subepithelial uid in 26 cases (Figures 1-2D, E and F); UBM images showed corneal posterior elastic layer interface strati cation in 26 cases (Figures 1-4G, H and I). All patients were treated with deep lamellar keratoplasty by the same operator under general anesthesia. In the healing stage after keratoplasty, there appeared obvious stromal scar accompanied by corneal stromal thickening and corneal attening (Figures 1-4J and K). A total of 24 cases with corneal posterior elastic layer rupture less than 2mm were included, among which 13 cases had corrected visual acuity less than 0.1, 9 cases had visual acuity of 0.1-0.2, and 2 cases had visual acuity of 0.3-0.5 one week after surgery. There were 14 cases with the rupture of the posterior elastic layer larger than 2mm, including 6 cases with corrected visual acuity less than 0.1, 7 cases with 0.1-0.2, and 1 case with 0.3-0.5 one week after surgery. 2.2 Statistical results According to paired t-test, the difference between the preoperative best corrected visual acuity and the postoperative best corrected visual acuity was statistically signi cant (t = 11.861, P < 0.05) ( Table 1). The differences in preoperative best corrected visual acuity and postoperative best corrected visual acuity between Group 1 and Group 2 after independent sample t-test were not statistically signi cant (t = 0.751, P > 0.05, T = 0.219, P > 0.05) ( Table 1). The differences in the heights of the rupture ports between Group 1 and Group 2 were statistically signi cant (t = − 7.533, P < 0.001) ( Table 1). Pearson correlation analysis showed that the size of fracture opening was signi cantly correlated with the height of fracture opening (r = 0.766, P < 0.001) ( Table 1), but there was no correlation between the size of fracture opening and posterior elastic layer strati cation (r = − 0.107, P > 0.05) ( Table 1).

Discussion
With the advancement of science and technology, there are a variety of ophthalmic examination equipment, among which UBM and AS-OCT are two important examination items. UBM scans the tissue at the depth of 5mm in the anterior segment of the eye through high-frequency ultrasound (50-100 MHz). In the panoramic mode, the exploration depth can reach about 6-7 mm, obtaining high-resolution and high-de nition images of any section, as well as quantitative analysis of many structures in the anterior segment. As a new ophthalmic imaging device, AS-OCT uses 1310nm infrared light for scanning, which can penetrate the cornea and scleral tissue, and complete the imaging and measurement of the anterior segment anatomy and pathology after cataract, refractive surgery, glaucoma, corneal transplantation and ocular trauma in a non-contact manner, providing the basis for anterior segment analysis. The scanning angle is measured from 0 to 360 degrees at a full angle, with the maximum depth of 6 mm. AS-OCT is a form of cross-sectional imaging based on light that does not require eye contact. Compared with UBM, AS-OCT is simpler and faster in operation, and more acceptable to patients [4]. The imaging principles of the two methods are similar. AS-OCT has higher resolution, while UBM detects biological tissues through ultrasonic waves, with stronger penetration, which is much higher than that of light waves. For a normal transparent cornea, the imaging ndings of the two are not much different, but for the examination of pathological changes that affect light penetration, such as obvious edema and thickening of corneal epithelium and scar in stroma layer, UBM has advantages over AS-OCT. Many articles have reported that under the condition of normal clear cornea, there is no signi cant difference between the measured values of UBM and AS-OCT in the comparative study of the measurement of biological parameters in the anterior segment of the eye [5][6][7][8]. However, there are few reports on the study of UBM and OCT in the anterior segment of keratoconus in the acute phase, and the comparative analysis of the two examinations is even less.
Esteban Fuentes et al. applied anterior segment OCT to analyze the anatomical features such as the thickness of epithelial and stroma layers, the hyperre exia of anterior elastic layer, Vogt striation, and the opacity of stroma layer of keratoconus in the acute phase, and they found that the thickening of corneal epithelial layer, thinning of stroma layer, the hyperre exia of anterior elastic layer, and the lack of stroma scar layer are the risk factors for inducing keratoconus edema in the acute phase [2]. Wang et al. found in their study that the imaging resolution of the central and peripheral cornea by UBM was low [9], and the epithelial boundary was unclear in the case of corneal edema. In this study, the imaging features of AS-OCT and UBM in acute keratoconus were corneal posterior elastic layer rupture and discontinuous echo. Localized obvious edema and thickening of the matrix layer; Vertical ssures with different sizes were formed in the matrix layer of the corresponding parts to communicate with the anterior chamber. The thickness of the matrix layer was thin, and no de nite perforation was seen. AS-OCT images clearly showed corneal epithelial integrity, subepithelial effusions, and epithelial separation from the anterior elastic layer. The UBM image shows that the interface of corneal posterior elastic layer is split into two layers. As there is no hemidesmosomes between the corneal posterior elastic layer and endothelial cells, the attachment relationship between the two depends on the endothelial cells themselves permanently producing new posterior elastic layer bers, and the endothelial cells are easily separated from the posterior elastic layer. When the posterior elastic layer is ruptured, there are usually wavy bending and re ecting double lines with a darker place in the middle [10]. Therefore, when the aqueous humor ows back due to the ssure formation of the matrix layer, it may break through the interface of the posterior elastic layer and separate from the endothelium (no related research report has been found in the literature yet)! ) Acute edema of keratoconus often occurs just below the central part of the cornea or under the nose, so we may not nd any subepithelial effusion in UBM examination of the same case, which is separated from the anterior elastic layer, but it may also be caused by different angles of incidence of ultrasonic waves. Corneal edema presented as uniform and weak re ection on AS-OCT images [11][12][13][14], and the corneal boundary was relatively clear. The near-infrared light of AS-OCT can transmit the edematous corneal stroma to the endothelial surface, anterior chamber, iris and lens for cross-sectional imaging, making up for the shortage that the posterior cornea and anterior chamber cannot be observed by slit-lamp microscope when the cornea is cloudy. As the corneal stroma layer was obviously thickened during acute corneal edema, the thickest edema area in this study was about 2.63 mm. Considering the imaging principle of AS-OCT, the permeability of light wave was relatively weak, which may cause unclear display at the posterior part of the cornea, and no interfacial layer of corneal posterior elastic layer was found. On the other hand, AS-OCT had a higher resolution, and we observed that the posterior elastic layer rupture opening in our patient was at least about 0.1mm, and the basal layer rupture resulted in an increased contact area with anterior aqueous humor, which also explains the fact that acute corneal edema is less likely to occur when the posterior elastic layer is simply avulsionruptured.
Although the difference between the pre-operative best corrected visual acuity and the post-operative best corrected visual acuity in all patients with acute keratoconus studied in this group is statistically signi cant, the post-operative visual acuity is often poor, which is related to various factors such as the scope of corneal edema, the size and location of posterior elastic layer hiatus, and the degree of stromal opacity of the patient's cornea. All the patients were treated with deep lamellar keratoplasty under general anesthesia by the same operator. In this study, it was found that there was no statistical difference between the preoperative and postoperative best corrected visual acuity of all the groups 1 and 2. Deep lamellar keratoplasty can reduce postoperative rejection to a great extent, reduce postoperative astigmatism and relieve the pressure of insu cient fresh corneal material. However, based on the above research, in order to achieve a better visual quality for patients, we are reminded that for patients with acute keratoconus, can partial penetrating keratoplasty be used? The size of fracture opening is signi cantly correlated with the height of fracture opening, while it has no correlation with the strati cation of back elastic layer. AS-OCT and UBM can provide us with accurate corneal imaging and measurement of ocular biological parameters, accurate positioning and indication of the size of the posterior elastic layer hiatus, and provide a very valuable reference for clinical surgery.
Esteban Fuentes et al. found that the lack of scar in the keratoconus stromal layer might be an inducing factor for the acute edema of keratoconus [2]. Due to the small number of cases in this study and the failure to observe the pre-acute corneal edema image of our patient, as well as the lack of research on its in uencing factors, we will strengthen the observation of relevant factors in future studies. The stroma ssure in keratoconus at the acute stage is rarely seen under slit-lamp microscope, and it has been only rarely reported that in our case AS-OCT and UBM examinations both showed stroma ssure of different sizes communicating with the anterior chamber. There are two causes for it, primary and secondary. The primary one is caused by the cracking of corneal stroma layer itself, and structural change and ssure appearance are caused by corneal deformation. Secondary because of rupture of the back elastic layer. The corneal edema is aggravated by the ssures in the stroma layer, because the contact area between the stroma layer and anterior aqueous humor is enlarged due to the ssures. The ssures in the stroma layer also affect the healing of the posterior elastic layer and the recovery of corneal edema [15][16][17]. In our case, the matrix fractures were more clearly demonstrated on AS-OCT images than on UBM images, with clearer margins. When corneal perforation occurs due to the thin corneal stroma of patients, UBM, as a contact test, is not suitable. Compared with UBM, AS-OCT is simple and fast to operate, and easy to be accepted by patients.   Figure 1 1-ABC is photograph of the anterior segment of the eye before and after acute keratoconus transplantation; 2-DEF is an AS-OCT image of keratoconus in the acute phase. 3-GHI is an UBM image of acute keratoconus; 4-JK is an AS-OCT image after deep lamellar keratoplasty.