Algorithm and software for field distortion correction in a commercial SD-OCT for corneal curvature assessment.

Accurate assessment of corneal curvatures using frequency domain optical coherence tomography (OCT) with galvanometer scanners remains challenging due to the well-known scan field distortion. This paper presents an algorithm and software for correcting the distortion using only two simple measurements in which a readily available standard sphere is positioned in different depths in front of the OCT scanner. This offers a highly accessible and easily reproducible method for the field distortion correction (FDC). The correction was validated by measuring different spherical phantoms and conducting corneal curvature measurements of ex vivo porcine corneas using a commercial spectral-domain OCT system and a clinically approved swept-source OCT as a reference instrument. Thus, the error in radius measurements of spherical phantoms was reduced by >90% and astigmatism by >80% using FDC. In explanted porcine eyes, the error in astigmatism measurements with the Telesto was reduced by 75% for power and 70% for angle. The best fitting sphere radius was determined up to a deviation of 0.4% from the Anterion. This paper describes a correction algorithm for OCT immanent distortion that is applicable to any scanning OCT setup and enables precise corneal curvature measurements. The MATLAB software for the FDC is publicly available on GitHub.

Quantitative assessment of corneal elasticity distribution after FS-LASIK using optical coherence elastography.

Quantifying corneal elasticity after femtosecond laser-assisted in situ keratomileusis (FS-LASIK) procedure plays an important role in improving surgical safety and quality, since some latent complications may occur ascribing to changes in postoperative corneal biomechanics. Nevertheless, it is suggested that current research has been severely constrained due to the lack of an accurate quantification method to obtain postoperative corneal elasticity distribution. In this paper, an acoustic radiation force optical coherence elastography system combined with the improved phase velocity algorithm was utilized to realize elasticity distribution images of the in vivo rabbit cornea after FS-LASIK under various intraocular pressure levels. As a result, elasticity variations within and between the regions of interest could be identified precisely. This is the first time that elasticity imaging of in vivo cornea after FS-LASIK surgery was demonstrated, and the results suggested that this technology may hold promise in further exploring corneal biomechanical properties after refractive surgery.

A new method for in vivo assessment of corneal transparency using spectral-domain OCT.

Corneal transparency is essential to provide a clear view into and out of the eye, yet clinical means to assess such transparency are extremely limited and usually involve a subjective grading of visible opacities by means of slit-lamp biomicroscopy. Here, we describe an automated algorithm allowing extraction of quantitative corneal transparency parameters with standard clinical spectral-domain optical coherence tomography (SD-OCT). Our algorithm employs a novel pre-processing procedure to standardize SD-OCT image analysis and to numerically correct common instrumental artifacts before extracting mean intensity stromal-depth (z) profiles over a 6-mm-wide corneal area. The z-profiles are analyzed using our previously developed objective method that derives quantitative transparency parameters directly related to the physics of light propagation in tissues. Tissular heterogeneity is quantified by the Birge ratio Br and the photon mean-free path (ls) is determined for homogeneous tissues (i.e., Br~1). SD-OCT images of 83 normal corneas (ages 22-50 years) from a standard SD-OCT device (RTVue-XR Avanti, Optovue Inc.) were processed to establish a normative dataset of transparency values. After confirming stromal homogeneity (Br <10), we measured a median ls of 570 µm (interdecile range: 270-2400 µm). By also considering corneal thicknesses, this may be translated into a median fraction of transmitted (coherent) light Tcoh(stroma) of 51% (interdecile range: 22-83%). Excluding images with central saturation artifact raised our median Tcoh(stroma) to 73% (interdecile range: 34-84%). These transparency values are slightly lower than those previously reported, which we attribute to the detection configuration of SD-OCT with a relatively small and selective acceptance angle. No statistically significant correlation between transparency and age or thickness was found. In conclusion, our algorithm provides robust and quantitative measurements of corneal transparency from standard SD-OCT images with sufficient quality (such as 'Line' and 'CrossLine' B-scan modes without central saturation artifact) and addresses the demand for such an objective means in the clinical setting.

Corneal remodeling after SMILE for moderate and high myopia: short-term assessment of spatial changes in corneal volume and thickness.

PURPOSE: To evaluate the early corneal remodeling and its influencing factors after Small incision lenticule extraction (SMILE) for moderate and high myopia. METHODS: This was a retrospective study. Pre- and post-operative (1 week and 1, 3, 6 months) corneal volume (CV), mean keratometry (Km), and corneal thickness (CT) were measured by Scheimpflug tomography. CT at the central, thinnest point, and on concentric circles of 2, 4, and 6 mm diameter was recorded to assess corneal thickness spatial profile (CTSP) and percentage of thickness increase (PTI) in the moderate and high myopia groups, and to explore possible influencing factors. RESULTS: After SMILE, the peripheral CT decreased in the moderate myopia group and central corneal thickness (CCT) increased in the high myopia group at 1 month compared to 1 week (all P < 0.05). The CV, Km and CT were significantly increased at 3 months compared to 1 month (all P < 0.05), but there was no significant change at 6 months compared to 3 months for both groups (all P > 0.05). Patients with high myopia showed greater corneal thickness changes (△CT) and higher PTI than moderate myopia (all P < 0.05). Regression analysis revealed that in addition to refraction, peripheral PTI was negatively correlated with CCT in the moderate myopia group (4 mm: ß = -0.023, P = 0.001; 6 mm: ß = -0.050, P < 0.001), as well as in the high myopia group (4 mm: ß = -0.038, P < 0.001; 6 mm: ß = -0.094, P < 0.001). Moreover, peripheral PTI in the moderate myopia group was negatively correlated with age (4 mm: ß = -0.071, P = 0.003; 6 mm: ß = -0.162, P < 0.001). CONCLUSIONS: After SMILE, the CV, Km, and CTSP showed dynamic changes in the early stage, which stabilized after 3 months. Compared to the moderate myopia group, the high myopia group experienced slower corneal stabilization. The change in PTI at 6 months after SMILE may be related to higher preoperative refraction, thinner CCT and younger age.

[Direct assessment of changes in biomechanical properties of the cornea after ReLEx SMILE and FemtoLASIK surgeries]. / Neposredstvennaya otsenka izmenenii biomekhanicheskikh svoistv rogovitsy posle provedeniya operatsii ReLEx SMILE i FemtoLASIK.

The aim of this study was to evaluate corneal biomechanical properties in myopic patients after ReLEx SMILE and FemtoLASIK surgeries using the Corvis ST analyzer. MATERIAL AND METHODS: The SMILE group comprised 23 patients (46 eyes) with spherical refraction -3.8±1.8 diopters (D), the FemtoLASIK group included 18 patients (36 eyes) with spherical refraction -3.5±1.3 D. Analysis of the biomechanical properties of the cornea was carried out using the CORVIS ST device (Oculus, Germany) before and seven days after corneal refractive surgery. RESULTS: In the SMILE group, with intraoperative decrease in corneal thickness of 91.43±19.43 µm, a significant increase in the following parameters was observed: deformation coefficient (DA ratio; p=0.0001), peak distance (PD; p=0.02) and inverse concave radius (ICR; p=0.003); a decrease in the stiffness parameter at first applanation (SP-A1; p=0.0001), Corvis biomechanical index (CBI; p=0.0001), intraocular pressure (IOP; p=0.001). In the FemtoLASIK group, with intraoperative decrease in corneal thickness of 75.33±32.3 µm, we observed a significant increase in DA ratio (p=0.0002), PD (p=0.04), ICR (p=0.0002), a decrease in SP-A1 (p<0.0001), IOP values (p=0.0003). In comparison with the FemtoLASIK group, the deformation amplitude DA changed significantly less in the SMILE group (p=0.04). In the FemtoLASIK group, in comparison with the SMILE group, the DA ratio (p=0.0009) and SP-A1 (p=0.0003) significantly increased. Intraoperative corneal thickness change correlates with ICR both in SMILE (R=0.52) and in FemtoLASIK (R=0.65). CONCLUSION: Corneal biomechanical properties determined with CORVIS ST in eyes with mild to moderate myopia change to a lesser extent after ReLEx SMILE compared to FemtoLASIK.

Simultaneous Assessment of the Whole Eye Biomechanics Using Ultrasonic Elastography.

OBJECTIVE: Current elastography techniques in the field of ophthalmology usually target one specific tissue, such as the cornea or the sclera. However, the eye is an inter-related organ, and some ocular diseases can alter the biomechanical properties of multiple anatomical structures. Hence, there is a need to develop an imaging tool that can non-invasively, quantitatively, and accurately characterize dynamic changes among these biomechanical properties. METHODS: A high resolution ultrasound elastography system was developed to achieve this goal. The efficacy and accuracy of the system was first validated on tissue-mimicking phantoms while mechanical testing measurements served as the gold standard. Next, an in vivo elevated intraocular pressure (IOP) model was established in rabbits to further test our system. In particular, elastography measurements were obtained at 5 IOP levels, ranging from 10 mmHg to 30 mmHg in 5 mmHg increments. Spatial-temporal maps of the multiple ocular tissues (cornea, lens, iris, optic nerve head, and peripapillary sclera) were obtained. RESULTS: The spatial-temporal maps were acquired simultaneously for the ocular tissues at the 5 different IOP levels. The statistical analysis of the elastic wave speed was presented for ocular tissues. Finally, the mapping for the elastic wave speed of each ocular component was acquired at each IOP level. CONCLUSION: Our elastography system can concurrently assess the biomechanical properties of multiple ocular structures and detect changes in biomechanical properties associated with changes in IOP. SIGNIFICANCE: This system provides a novel tool to measure and quantify the biomechanical properties of the whole eye.

Corneal Hysteresis for the Diagnosis of Glaucoma and Assessment of Progression Risk: A Report by the American Academy of Ophthalmology.

PURPOSE: To review the current published literature on the utility of corneal hysteresis (CH) to assist the clinician in the diagnosis of glaucoma or in the assessment of risk for disease progression in existing glaucoma patients. METHODS: Searches of the peer-reviewed literature in the PubMed database were performed through July 2022. The abstracts of 423 identified articles were examined to exclude reviews and non-English articles. After inclusion and exclusion criteria were applied, 19 articles were selected, and the panel methodologist rated them for level of evidence. Eight articles were rated level I, and 5 articles were rated level II. The 6 articles rated level III were excluded. RESULTS: Corneal hysteresis is lower in patients with primary open-angle glaucoma, primary angle-closure glaucoma, pseudoexfoliative glaucoma, and pseudoexfoliation syndrome compared with normal subjects. Interpretation of low CH in patients with high intraocular pressure (IOP) or on topical hypotensive medications is complicated by the influence of these parameters on CH measurements. However, CH is also lower in treatment-naïve, normal-tension glaucoma patients compared with normal subjects who have a similar IOP. In addition, lower CH is associated with an increased risk of progression of glaucoma based on visual fields or structural markers in open-angle glaucoma patients, including those with apparently well-controlled IOP. CONCLUSIONS: Corneal hysteresis is lower in glaucoma patients compared with normal subjects, and lower CH is associated with an increased risk of disease progression. However, a causal relationship remains to be demonstrated. Nevertheless, measurement of CH complements current structural and functional assessments in determining disease risk in glaucoma suspects and patients. FINANCIAL DISCLOSURE(S): Proprietary or commercial disclosure may be found after the references.

Convolutional neural network-based common-path optical coherence tomography A-scan boundary-tracking training and validation using a parallel Monte Carlo synthetic dataset.

We present a parallel Monte Carlo (MC) simulation platform for rapidly generating synthetic common-path optical coherence tomography (CP-OCT) A-scan image dataset for image-guided needle insertion. The computation time of the method has been evaluated on different configurations and 100000 A-scan images are generated based on 50 different eye models. The synthetic dataset is used to train an end-to-end convolutional neural network (Ascan-Net) to localize the Descemet's membrane (DM) during the needle insertion. The trained Ascan-Net has been tested on the A-scan images collected from the ex-vivo human and porcine cornea as well as simulated data and shows improved tracking accuracy compared to the result by using the Canny-edge detector.

Quantitative Assessment of Biomechanical Properties of the Human Keratoconus Cornea Using Acoustic Radiation Force Optical Coherence Elastography.

Purpose: Quantification of biomechanical properties of keratoconus (KC) corneas has great significance for early diagnosis and treatment of KC, but the corresponding clinical measurement remains challenging. Here, we developed an acoustic radiation force (ARF) optical coherence elastography technique and explored its potential for evaluating biomechanical properties of KC corneas. Methods: An ARF system was used to induce the tissue deformation, which was detected by an optical coherence tomography system, and thus the localized point-by-point Young's modulus measurements were achieved. Then, two healthy rabbit eyes were imaged to test the system, after which the human keratoconus cornea was evaluated by using the same method. Three regions were selected for biomechanics analysis: the conical region, the transitional region, and the peripheral region. Results: Young's moduli of transitional region ranged from 53.3 to 58.5 kPa. The corresponding values for the peripheral region were determined to be 58.6 kPa and 63.2 kPa, respectively. Young's moduli of the conical region were gradually increased by 18.3% from the center to the periphery, resulting in the minimum and maximum values of 44.9 kPa and 53.1 kPa, respectively. Furthermore, Young's moduli of the anterior and posterior of the center were determined to be 44.9 kPa and 50.7 kPa, respectively. Conclusions: Differences in biomechanical properties between the three regions and slight variations within the conical region were clearly distinguished. Biomechanical weakening of the keratoconus cornea was mainly localized in the conical region, especially in the vertex position. Translational Relevance: The system may provide a promising clinical tool for the noninvasive evaluation of local corneal biomechanics and thus may have potential applications in early keratoconus detection with further optimization.

Longitudinal assessment of the effect of alkali burns on corneal biomechanical properties using optical coherence elastography.

Eye injury due to alkali burn is a severe ocular trauma that can profoundly affect corneal structure and function, including its biomechanical properties. Here, we assess the changes in the mechanical behavior of mouse corneas in response to alkali-induced injury by conducting longitudinal measurements using optical coherence elastography (OCE). A non-contact air-coupled ultrasound transducer was used to induce elastic waves in control and alkali-injured mouse corneas in vivo, which were imaged with phase-sensitive optical coherence tomography. Corneal mechanical properties were estimated using a modified Rayleigh-Lamb wave model, and results show that Young's modulus of alkali-burned corneas were significantly greater than that of their healthy counterparts on days 7 (p = 0.029) and 14 (p = 0.026) after injury. These findings, together with the changes in the shear viscosity coefficient postburn, indicate that the mechanical properties of the alkali-burned cornea are significantly modulated during the wound healing process.

Detection of subclinical keratoconus using a novel combined tomographic and biomechanical model based on an automated decision tree.

Early detection of keratoconus is a crucial factor in monitoring its progression and making the decision to perform refractive surgery. The aim of this study was to use the decision tree technique in the classification and prediction of subclinical keratoconus (SKC). A total of 194 eyes (including 105 normal eyes and 89 with SKC) were included in the double-center retrospective study. Data were separately used for training and validation databases. The baseline variables were derived from tomography and biomechanical imaging. The decision tree models were generated using Chi-square automatic interaction detection (CHAID) and classification and regression tree (CART) algorithms based on the training database. The discriminating rules of the CART model selected metrics of the Belin/Ambrósio deviation (BAD-D), stiffness parameter at first applanation (SPA1), back eccentricity (Becc), and maximum pachymetric progression index in that order; On the other hand, the CHAID model selected BAD-D, deformation amplitude ratio, SPA1, and Becc. Further, the CART model allowed for discrimination between normal and SKC eyes with 92.2% accuracy, which was higher than that of the CHAID model (88.3%), BAD-D (82.0%), Corvis biomechanical index (CBI, 77.3%), and tomographic and biomechanical index (TBI, 78.1%). The discriminating performance of the CART model was validated with 92.4% accuracy, while the CHAID model was validated with 86.4% accuracy in the validation database. Thus, the CART model using tomography and biomechanical imaging was an excellent model for SKC screening and provided easy-to-understand discriminating rules.

Translation model for anterior segment tomographic data to corneal spherical aberration derived from a Monte-Carlo simulation based on raytracing.

BACKGROUND: Intraocular lenses with a negative aspherical design for correction of corneal spherical aberration (SA) have gained popularity in recent decades. In most cases, a 'one size fits all' concept is followed, where all eyes receive lenses with the same SA correction. The purpose of this study is to develop a strategy based on raytracing using anterior segment tomography data to extract corneal SA and to provide simple multivariable linear models for prediction of corneal SA. METHODS: The analysis was based on a large dataset of 8737 measurements of 8737 eyes from 1 clinical centre, using the Casia2 anterior segment tomographer. An optical model based on: corneal front and back surface radius Ra and Rp, asphericities Qa and Qp, corneal thickness CCT, anterior chamber depth ACD, and pupil centre position (X-Y position: PupX and PupY ), was defined for each measurement. Corneal SA was derived using a 6-mm aperture perpendicular to the incident ray and centred on the chief ray, and linear prediction models were derived for SA using biometric data. Cross-validation was used for model performance evaluation. RESULTS: Using raytracing, the wavefront error within an aperture (6-mm diameter centred on the intersection of the chief ray with the cornea) was calculated and corneal SA was extracted. After identifying the relevant effect sizes (Ra, Qa, Rp Qp, ACD, PupX and PupY ) using stepwise linear regression, linear mixed-effects models (model 1: all effect sizes, model 2: Ra, Qa, Rp and Qp, model 3: Ra and Qa) were set up on the training data in terms of a Monte-Carlo simulation. On the test data (training data), model 1 with a mean absolute/root-mean-squared prediction error of 0.0095/0.0130 (0.0095/0.0127) performed similarly to model 2 with 0.0097/0.0131 (0.0096/0.0127), and both outperformed model3 with 0.0152/0.0197 (0.0148/0.0190). CONCLUSION: Based on the Casia2 anterior segment tomographer, corneal SA could be derived using shape data (curvature and asphericities) of both corneal surfaces (model 2). This information could easily be used for selection of the appropriate negative aspherical lens design in cataract surgery.

New algorithm for corneal densitometry assessment based on anterior segment optical coherence tomography.

PURPOSE: To describe a new algorithm to measure corneal densitometry based on images obtained by swept source anterior segment ocular coherence tomography (SS-AS-OCT) and establish standard densitometry values in a group of normal eyes. METHODS: A total of 111 healthy participants (195 eyes) were enrolled in this study. Using a MATLAB designed algorithm, the cornea was segmented into three layers: anterior, posterior and mid-stroma, and it was divided into two concentric areas, 0-2 and 2-4 mm, resulting in nine areas for the analysis. The mean corneal densitometry values were calculated and expressed as grayscale units (GSU). RESULTS: The mean age was 57 years (range 22-87), with 100 (51.3%) right eyes and 95 (48.7%) left eyes. The total corneal densitometry was 86.9 ± 12.1 GSU. The mid-stroma layer had the highest densitometry values, 87.4 ± 12.1 GSU, and the anterior layer had the lowest values, 81.9 ± 14.2 GSU. Densitometry differences between the anterior layer and the mid-stroma layer (P < 0.001), as well as the anterior layer and the posterior layer (P < 0.05) were statistically significant. The 0-2 mm concentric area had higher mean densitometry values, 97.8 ± 12.7 GSU, and the differences were significant compared to the 2-4 mm concentric area (P < 0.001). No correlation was found between the corneal densitometry values and gender or age. CONCLUSIONS: The new MATLAB segmentation algorithm for the analysis of corneal SS-AS-OCT images is capable to objectively assess corneal densitometry. We provide standard and normal data for better clinical and research approach.

Objective Assessment of Corneal and Lens Clarity in Patients with COVID-19.

SIGNIFICANCE: The severe acute respiratory syndrome coronavirus 2 virus, which causes the coronavirus disease 2019 (COVID-19) pandemic, can bind to epithelial cells in the human cornea and conjunctiva, which may result in changes of corneal and lenticular densitometry. PURPOSE: We aimed to report the corneal and lenticular clarity of patients who had been diagnosed previously with confirmed infection of COVID-19. METHODS: This was a prospective, cross-sectional study. Fifty-three patients who had recovered from COVID-19 and 51 healthy individuals who had not had COVID-19 (control) were included in this study. Measurement of the corneal (at a corneal diameter of 12 mm) and lens densitometry of participants was performed using a Pentacam HR Scheimpflug imaging system. Three different optimal depths as anterior, central, and posterior layers and four concentric zones, consisting of diameters of 0 to 2, 2 to 6, 6 to 10, and 10 to 12 mm, were chosen to perform the corneal densitometric measurements. The mean lenticular densitometric values were calculated in zone 1 at 2.0 mm, zone 2 at 4.0 mm, and zone 3 at 6.0 mm by taking the central part of the pupil as reference point. RESULTS: When compared with the individuals in the control group, the corneal densitometry measurements in the patients in the COVID-19 group were considerably higher in anterior 0 to 2 mm, 2 to 6 mm, 6 to 10 mm zones and total diameter; center 0 to 2 and 2 to 6 mm zones and total diameter; posterior 0 to 2 mm zone; and total corneal 0 to 2 and 2 to 6 mm zones and total diameter (P < .05, for each). When compared with the individuals in the control group, all of the lens densitometry measurements, except for those located in zone 1, were determined to be significantly higher among the patients in the COVID-19 group (P < .05, for each). CONCLUSIONS: Significant alterations were found in corneal and lenticular densitometric values in patients who had had COVID-19. The virus could adversely affect cornea and lens transparency.

The relationship of ocular parameters with clinical parameters and disease-related quality of life in patients with systemic sclerosis: A cross-sectional study.

OBJECTIVE: To evaluate choroidal thickness (CT), corneal parameters, and scleral thickness (ST) in patients with systemic sclerosis (SSc) and to determine their relationship with disease-related quality of life (QoL). METHODS: The study included 38 patients with SSc and 40 healthy controls. A detailed ocular examination was performed on all participants. Corneal parameters such as K1, K2, Km, corneal volume (CV), central corneal thickness (CCT), and ST at a distance of 1000, 2000, and 3000 µm from the scleral spur were measured. CT was measured at five points, including the subfoveal area and the temporal and nasal points at radii of 750.0 and 1500.0 µm. The scleroderma health assessment questionnaire (SHAQ) was administered to SSc patients to investigate the disease-related QoL. RESULTS: Individuals with SSc had thicker ST at all distances from the scleral spur (P=0.008, P=0.001, P=0.002, respectively). All corneal parameters were significantly lower in the SSc group than in the control group (P < 0.05). Moreover, SSc patients had significantly lower median CT at N750.0, N1500.0, T750.0, and T1500.0 points and thinner subfoveal CT than healthy controls (P < 0.05). There was a weak-moderate negative correlation between ST and the components of the SHAQ scale and SHAQ-global. CONCLUSION: Despite not having ocular involvement, SSc patients had thicker ST but thinner CT and corneal parameters than healthy controls. This may indicate subclinical inflammation in patients with SSc. Only ST was affected by organ involvement and QoL among the ocular parameters.

An inter-day assessment of the ABC parameters in the evaluation of progressive keratoconus.

The progression of keratoconus is commonly determined by comparing the results of corneal tomographic measurements on different occasions. However, investigations on the repeatability of measurements are commonly performed within the same day, thus not taking the inter-day variation into account. The effect of keratoconus disease severity on the measurement error is also seldom considered. In this post hoc investigation, the parameters A, B and C in the Belin ABCD Progression Display were evaluated in relation to disease severity in intra-day and inter-day measurements. Four consecutive measurements were performed on 61 patients with keratoconus on the same day (intra-day). In another cohort, four consecutive measurements were obtained and then repeated 3 days later in 25 patients with keratoconus and 25 healthy controls (inter-day). The results suggest that the diagnosis of disease progression would benefit from inter-day measurements, and the stratification of the parameters A and C according to disease severity. It is also recommended that tomographic systems such as the Pentacam HR be modified to allow the comparison of both single measurements and the mean of replicate measurements of the parameters used in the assessment of progression of keratoconus.

A deep transfer learning framework for the automated assessment of corneal inflammation on in vivo confocal microscopy images.

PURPOSE: Infiltration of activated dendritic cells and inflammatory cells in cornea represents an important marker for defining corneal inflammation. Deep transfer learning has presented a promising potential and is gaining more importance in computer assisted diagnosis. This study aimed to develop deep transfer learning models for automatic detection of activated dendritic cells and inflammatory cells using in vivo confocal microscopy images. METHODS: A total of 3453 images was used to train the models. External validation was performed on an independent test set of 558 images. A ground-truth label was assigned to each image by a panel of cornea specialists. We constructed a deep transfer learning network that consisted of a pre-trained network and an adaptation layer. In this work, five pre-trained networks were considered, namely VGG-16, ResNet-101, Inception V3, Xception, and Inception-ResNet V2. The performance of each transfer network was evaluated by calculating the area under the curve (AUC) of receiver operating characteristic, accuracy, sensitivity, specificity, and G mean. RESULTS: The best performance was achieved by Inception-ResNet V2 transfer model. In the validation set, the best transfer system achieved an AUC of 0.9646 (P<0.001) in identifying activated dendritic cells (accuracy, 0.9319; sensitivity, 0.8171; specificity, 0.9517; and G mean, 0.8872), and 0.9901 (P<0.001) in identifying inflammatory cells (accuracy, 0.9767; sensitivity, 0.9174; specificity, 0.9931; and G mean, 0.9545). CONCLUSIONS: The deep transfer learning models provide a completely automated analysis of corneal inflammatory cellular components with high accuracy. The implementation of such models would greatly benefit the management of corneal diseases and reduce workloads for ophthalmologists.

Longitudinal Assessment of Alkali Injury on Mouse Cornea Using Anterior Segment Optical Coherence Tomography.

Purpose: Chemical burns due to alkalis cause extensive damage to the ocular surface leading to blindness. Assessment of ocular burn could be challenging due to severe opacity, inflammation, and angiogenesis. Anterior segment optical coherence tomography (AS-OCT) and OCT angiography (OCTA) may provide fast, non-invasive deep tissue visualization of pathology with high sensitivity in conjunction with slit-lamp analysis. Methods: C57-BL/6J mice were anesthetized with ketamine/dexmedetomidine, and corneal alkali burn was induced (n = 6) by placing filter paper soaked in 1-M sodium hydroxide for 30 seconds on the right eye while the left eye was kept as control. Longitudinal imaging was done with AS-OCT/OCTA and fluorescein angiography at various time intervals for 14 days. Results: AS-OCT showed characteristic pathological changes in alkali-burned eyes with high sensitivity. Although OCT/OCTA showed three-dimensional and cross-sectional views of the anterior chamber and angiogenesis, fluorescein angiography showed nascent vessels with active leakage. Corneal swelling progressively increased by 125.26% on day 12 with a high prevalence of epithelial bullae, stromal cysts, stromal splitting, and Descemet's membrane detachment. Neovascularization was noted as early as day 4 in the burned eyes by both methods. Severe corneal opacity and anterior chamber inflammation were also detected by AS-OCT/OCTA. Conclusions: AS-OCT/OCTA is a promising, noninvasive, high-resolution imaging modality that can provide both qualitative and quantitative information regarding deep tissue pathology at a structural level. Translational Relevance: Noninvasive AS-OCT/OCTA and fluorescein methods show promise in clinical pathology evaluation for ocular injury management and prognostic indications, as the early presence of Descemet's membrane detachment and corneal swelling appears to be correlated with the severity and localization of corneal neovascularization.