Quantitative analysis of in vivo confocal microscopy images: A review

Multimodal Imaging Characterization of Butterfly Hair-Induced Keratitis

PURPOSE

Multimodal imaging was performed to characterize butterfly hair-induced keratitis based on anterior segment optical coherence tomography and in vivo confocal microscopy.

METHODS

This study was a case report.

RESULTS

A 6-year-old girl presented with acute keratitis induced by multiple butterfly hairs. Severe itching and pain developed immediately after rubbing her left eye, leading to significant pain and moderate vision loss, even after undergoing twice removal of the corneal epithelium. The hair-like foreign bodies were distributed at various depths inside the corneal stroma, even extending into the anterior chamber. The symptoms and corneal infiltration gradually decreased within 6 months with the use of topical steroids and immunosuppressors. The hairs located in the superficial and middle stromal layers of the cornea disappeared at the 6-month follow-up, but the hairs in the deep stromal layer tended to move deeper. The diagnosis was confirmed by in vivo confocal microscopy and microphotography. The migration tendency of the hairs into the intraocular space was observed using anterior segment optical coherence tomography (AS-OCT).

CONCLUSIONS

Butterfly hair-induced keratitis can be controlled by the treatment with topical steroids and immunosuppressors, but the hairs tend to move into the eyes. To the best of our knowledge, this is the first case of corneal in vivo confocal imaging of butterfly hairs.

Non-invasive in vivo imaging of human corneal microstructures with optical coherence microscopy

Non-invasive imaging systems with cellular-level resolution offer the opportunity to identify biomarkers of the early stage of corneal diseases, enabling early intervention, monitoring of disease progression, and evaluating treatment efficacy. In this study, a non-contact polarization-dependent optical coherence microscope (POCM) was developed to enable non-invasive in vivo imaging of human corneal microstructures. The system integrated quarter-wave plates into the sample and reference arms of the interferometer to enable deeper penetration of light in tissues as well as mitigate the strong specular reflection from the corneal surface. A common-path approach was adopted to enable control over the polarization in a free space configuration, thus alleviating the need for a broadband polarization-maintained fiber. The POCM achieved volumetric imaging of corneal microstructures, including endothelial cells over a field of view 0.5 × 0.5 mm2 with an almost isotropic resolution of ∼2.2 µm and a volume (500 × 500 × 2048 voxels) rate of 1 Hz. A self-interference approach between the corneal surface and underlying layers was also developed to lessen the corneal curvature and axial motion artifacts, thus enabling high-resolution imaging of microstructures in the anterior cornea, including squamous epithelial cells, wing epithelial cells, basal epithelial cells, sub-basal nerve plexus, and stromal keratocytes.

Interpretable deep learning for diagnosis of fungal and acanthamoeba keratitis using in vivo confocal microscopy images

Infectious keratitis refers to a group of corneal disorders in which corneal tissues suffer inflammation and damage caused by pathogenic infections. Among these disorders, fungal keratitis (FK) and acanthamoeba keratitis (AK) are particularly severe and can cause permanent blindness if not diagnosed early and accurately. In Vivo Confocal Microscopy (IVCM) allows for imaging of different corneal layers and provides an important tool for an early and accurate diagnosis. In this paper, we introduce the IVCM-Keratitis dataset, which comprises of a total of 4001 sample images of AK and FK, as well as non-specific keratitis (NSK) and healthy corneas classes. We use this dataset to develop multiple deep-learning models based on Convolutional Neural Networks (CNNs) to provide automated assistance in enhancing the diagnostic accuracy of confocal microscopy in infectious keratitis. Densenet161 had the best performance among these models, with an accuracy, precision, recall, and F1 score of 93.55%, 92.52%, 94.77%, and 96.93%, respectively. Our study highlights the potential of deep learning models to provide automated diagnostic assistance for infectious keratitis via confocal microscopy images, particularly in the early detection of AK and FK. The proposed model can provide valuable support to both experienced and inexperienced eye-care practitioners in confocal microscopy image analysis, by suggesting the most likely diagnosis. We further demonstrate that these models can highlight the areas of infection in the IVCM images and explain the reasons behind their diagnosis by utilizing saliency maps, a technique used in eXplainable Artificial Intelligence (XAI) to interpret these models.

Evaluation of the Cell Concentration in Suspensions of Human Leukocytes by Ultrasound Imaging: The Influence of Size Dispersion and Cell Type

This work focuses on the use of ultrasound imaging to evaluate the cell concentration of dilute leukocyte suspensions in the range of 10-3000 cells/µL. First, numerical simulations were used to study the influence of the size dispersion and the leukocyte type on the performance of the concentration estimation algorithms, which were developed in previous works assuming single-sized scatterers. From this analysis, corrections to the mentioned algorithms were proposed and then the performance of these corrections was evaluated from experiments. For this, ultrasound images were captured from suspensions of lymphocytes, granulocytes, and their mixtures. These images were obtained using a 20 MHz single-channel scanning system. Results confirmed that concentration estimates provided by conventional algorithms were affected by the size dispersion of cells, leading to a remarkable underestimation of results. The proposed correction to compensate for cell size dispersion obtained from simulations improved the concentration estimation of these algorithms, for the cell suspensions tested, approaching the results to the reference optical characterization. Moreover, it was shown that these models provided a total leukocyte concentration from the ultrasound images which was independent of the relative populations of different white blood cell types.

Corneal Confocal Microscopy as a Quantitative Imaging Biomarker of Diabetic Peripheral Neuropathy: A Review

Distal symmetric polyneuropathy (DPN), particularly chronic sensorimotor DPN, represents one of the most frequent complications of diabetes, affecting 50% of diabetic patients and causing an enormous financial burden. Whilst diagnostic methods exist to detect and monitor this condition, they have significant limitations, mainly due to their high subjectivity, invasiveness, and non-repeatability. Corneal confocal microscopy (CCM) is an in vivo, non-invasive, and reproducible diagnostic technique for the study of all corneal layers including the sub-basal nerve plexus, which represents part of the peripheral nervous system. We reviewed the current literature on the use of CCM as an instrument in the assessment of diabetic patients, particularly focusing on its role in the study of sub-basal nerve plexus alterations as a marker of DPN. CCM has been demonstrated to be a valid in vivo tool to detect early sub-basal nerve plexus damage in adult and pediatric diabetic patients, correlating with the severity of DPN. Despite its great potential, CCM has still limited application in daily clinical practice, and more efforts still need to be made to allow the dissemination of this technique among doctors taking care of diabetic patients.

Technical Report: Automatic Measurement of Corneal Nerve Fiber Area Using Versatile Software

SIGNIFICANCE

In vivo confocal microscopy has become a popular method to observe the details of corneal structures. We consider the area of corneal structures to be a versatile index and have measured the areas of various corneal structures using commercially available software.

PURPOSE

To evaluate the accuracy of software used to measure the corneal nerve fiber area.

METHODS

The corneal structures of 11 healthy volunteers were visualized using in vivo confocal microscopy. The image that most clearly depicted the corneal nerve fibers of each participant was selected for analysis. The corneal nerve fiber area was automatically measured by the software. An experienced ophthalmologist then manually measured the corneal nerve fiber area in each image assessed by the software. The Pearson correlation test was used to determine the correlation coefficient between the corneal nerve fiber areas measured automatically and those measured manually. The correlation between the corneal nerve fiber area and the participant's age was also evaluated.

RESULTS

A strong correlation was found between the corneal nerve fiber area measured automatically and the corneal nerve fiber area measured manually (r = .98, P = 2.4 × 10-7). The corneal nerve fiber area was not correlated with participant age, regardless of whether the area was measured automatically (r = -.26, P = .44) or manually (r = -.13, P = .71).

CONCLUSIONS

The software used in this study automatically measures the corneal nerve fiber area with accuracy similar to that of manual measurement by an experienced ophthalmologist. This software has potential for use in quantifying the areas of various corneal structures.

Neuroimmune crosstalk in the cornea: The role of immune cells in corneal nerve maintenance during homeostasis and inflammation

In the cornea, resident immune cells are in close proximity to sensory nerves, consistent with their important roles in the maintenance of nerves in both homeostasis and inflammation. Using in vivo confocal microscopy in humans, and ex vivo immunostaining and fluorescent reporter mice to visualize corneal sensory nerves and immune cells, remarkable progress has been made to advance our understanding of the physical and functional interactions between corneal nerves and immune cells. In this review, we summarize and discuss recent studies relating to corneal immune cells and sensory nerves, and their interactions in health and disease. In particular, we consider how disrupted corneal nerve axons can induce immune cell activity, including in dendritic cells, macrophages and other infiltrating cells, directly and/or indirectly by releasing neuropeptides such as substance P and calcitonin gene-related peptide. We summarize growing evidence that the role of corneal intraepithelial immune cells is likely different in corneal wound healing versus other inflammatory-dominated conditions. The role of different types of macrophages is also discussed, including how stromal macrophages with anti-inflammatory phenotypes communicate with corneal nerves to provide neuroprotection, while macrophages with pro-inflammatory phenotypes, along with other infiltrating cells including neutrophils and CD4⁺ T cells, can be inhibitory to corneal re-innervation. Finally, this review considers the bidirectional interactions between corneal immune cells and corneal nerves, and how leveraging this interaction could represent a potential therapeutic approach for corneal neuropathy.

Segmentation methods and morphometry of confocal microscopy imaged corneal epithelial cells

PURPOSE

To develop and explore automated cell identification and segmentation methods for morphometry of confocal microscopy imaged corneal epithelial cells using ImageJ software.

METHODS

In vivo confocal microscopy images of the intermediate (wing) and basal cell layers of the central and peripheral corneas of 20 healthy participants were analysed. The intermediate and basal cell areas obtained using the two new techniques (i.e., manual- and auto- thresholding) were compared with the widely used manual tracing technique. A predefined range of epithelial cell morphometric parameters was used as image descriptors to improve cell identification and segmentation.

RESULTS

The mean intermediate cell area obtained using the manual tracing (central; 120 ± 14 µm², peripheral; 123 ± 15 µm²) was statistically similar (p > 0.05) to the manual thresholding (central; 119 ± 7 µm², peripheral; 119 ± 8) but not with the auto thresholding technique (central; 101 ± 8 µm², peripheral; 101 ± 7 µm²). Bland-Altman limits of agreement for the mean difference (measurement bias) in central and peripheral intermediate cell area determined via manual tracing and manual thresholding techniques were 1 µm² (+25 to - 23 µm²) and 4 µm² (+29.8 to - 21.9 µm²). There were statistically significant differences in basal cell area between the three methods.

CONCLUSION

The manual thresholding technique may be used for automated identification and segmentation of corneal epithelial intermediate cells (central and peripheral) for assessing various morphometric parameters. However, measurement of the corneal epithelial basal cells is less reliable using thresholding techniques.

Corneal in vivo Confocal Microscopy for Assessment of Non-Neurological Autoimmune Diseases: A Meta-Analysis

Purpose

This study aimed to evaluate the features of corneal nerve with in vivo confocal microscopy (IVCM) among patients with non-neurological autoimmune (NNAI) diseases.

Methods

We systematically searched PubMed, Web of Science, and Cochrane Central Register of Controlled Trials for studies published until May 2021. The weighted mean differences (WMDs) of corneal nerve fiber length (CNFL), corneal nerve fiber density (CNFD), corneal nerve branch density (CNBD), tortuosity, reflectivity, and beadings per 100 μm with a 95% CI between NNAI and control group were analyzed using a random-effects model.

Results

The results showed 37 studies involving collective totals of 1,423 patients and 1,059 healthy controls were ultimately included in this meta-analysis. The pooled results manifested significantly decreased CNFL (WMD: −3.94, 95% CI: −4.77–−3.12), CNFD (WMD: −6.62, 95% CI: −8.4–−4.85), and CNBD (WMD: −9.89, 95% CI: −14–−5.79) in NNAI patients. In addition, the NNAI group showed more tortuous corneal nerve (WMD: 1.19, 95% CI:0.57–1.81). The comparison between NNAI patients and healthy controls in beadings per 100 μm corneal nerve length was inconsistent. No significant difference was found in the corneal nerve fiber reflectivity between NNAI and the control group (WMD: −0.21, 95% CI: −0.65–0.24, P = 0.361).

Conclusions

The parameters and morphology of corneal nerves observed by IVCM proved to be different in NNAI patients from healthy controls, suggesting that IVCM may be a non-invasive technique for identification and surveillance of NNAI diseases.

Real-time large-area imaging of the corneal subbasal nerve plexus

The morphometric assessment of the corneal subbasal nerve plexus (SNP) by confocal microscopy holds great potential as a sensitive biomarker for various ocular and systemic conditions and diseases. Automated wide-field montages (or large-area mosaic images) of the SNP provide an opportunity to overcome the limited field of view of the available imaging systems without the need for manual, subjective image selection for morphometric characterization. However, current wide-field montaging solutions usually calculate the mosaic image after the examination session, without a reliable means for the clinician to predict or estimate the resulting mosaic image quality during the examination. This contribution describes a novel approach for a real-time creation and visualization of a mosaic image of the SNP that facilitates an informed evaluation of the quality of the acquired image data immediately at the time of recording. In cases of insufficient data quality, the examination can be aborted and repeated immediately, while the patient is still at the microscope. Online mosaicking also offers the chance to identify an overlap of the imaged tissue region with previous SNP mosaic images, which can be particularly advantageous for follow-up examinations.

Artificial Intelligence-Based Approaches to Reflectance Confocal Microscopy Image Analysis in Dermatology

Reflectance confocal microscopy (RCM) is a non-invasive imaging method designed to identify various skin diseases. Confocal based diagnosis may be subjective due to the learning curve of the method, the scarcity of training programs available for RCM, and the lack of clearly defined diagnostic criteria for all skin conditions. Given that in vivo RCM is becoming more widely used in dermatology, numerous deep learning technologies have been developed in recent years to provide a more objective approach to RCM image analysis. Machine learning-based algorithms are used in RCM image quality assessment to reduce the number of artifacts the operator has to view, shorten evaluation times, and decrease the number of patient visits to the clinic. However, the current visual method for identifying the dermal-epidermal junction (DEJ) in RCM images is subjective, and there is a lot of variation. The delineation of DEJ on RCM images could be automated through artificial intelligence, saving time and assisting novice RCM users in studying the key DEJ morphological structure. The purpose of this paper is to supply a current summary of machine learning and artificial intelligence’s impact on the quality control of RCM images, key morphological structures identification, and detection of different skin lesion types on static RCM images.

Experimental Designs to Study the Aggregation and Colonization of Biofilms by Video Microscopy With Statistical Confidence

The goal of this study was to quantify the variability of confocal laser scanning microscopy (CLSM) time-lapse images of early colonizing biofilms to aid in the design of future imaging experiments. To accomplish this a large imaging dataset consisting of 16 independent CLSM microscopy experiments was leveraged. These experiments were designed to study interactions between human neutrophils and single cells or aggregates of Staphylococcus aureus (S. aureus) during the initial stages of biofilm formation. Results suggest that in untreated control experiments, variability differed substantially between growth phases (i.e., lag or exponential). When studying the effect of an antimicrobial treatment (in this case, neutrophil challenge), regardless of the inoculation level or of growth phase, variability changed as a frown-shaped function of treatment efficacy (i.e., the reduction in biofilm surface coverage). These findings were used to predict the best experimental designs for future imaging studies of early biofilms by considering differing (i) numbers of independent experiments; (ii) numbers of fields of view (FOV) per experiment; and (iii) frame capture rates per hour. A spreadsheet capable of assessing any user-specified design is included that requires the expected mean log reduction and variance components from user-generated experimental results. The methodology outlined in this study can assist researchers in designing their CLSM studies of antimicrobial treatments with a high level of statistical confidence.

A Novel Rat Model of Dry Eye Induced by Aerosol Exposure of Particulate Matter

Purpose

The purpose of this study was to introduce a novel dry eye rat model induced by aerosol exposure of particulate matter (PM).

Methods

A total of 30 female Sprague Dawley (SD) rats divided into 3 groups: the control group, the low-level exposed group, and the high-level exposed group. The rats in the experience groups were directly exposed to PM samples in the exposure chamber over 14 days. The clinical observation, including tear volume, corneal fluorescein staining, breakup time (BUT), inflammation index, corneal irregularity score, and corneal confocal microscopy. Eyeballs were collected on day 14 for hematoxylin and eosin (H&E) staining and PAS staining. TUNEL assay, CD45, and Ki67 immunostaining was performed and corneal ultrastructural changes were detected by electron microscopy. IL-1β, TNF-α, IFN-γ, and NF-κB Western blot analysis were used to observe the possible pathogenesis.

Results

In the PM-treated groups, the number of layers in the corneal epithelium and corneal nerve fiber length were significantly decreased compared with that of the control group. The number of corneal epithelial microvilli and chondriosome/desmosomes were drastically reduced in PM-treated groups. Confocal microscopy and CD45 immunohistochemistry showed inflammatory cell infiltration in the PM-treated groups. PM caused apoptosis of corneal and conjunctival epithelial cells while leading to abnormal epithelial cell proliferation, meanwhile, conjunctival goblet cells in the PM-treated group were also significantly reduced. PM significantly increased the levels of IL-1β, TNF-α, IFN-γ, and p-NF-κB-p65 in the cornea.

Conclusions

Aerosol exposure of PM can reduce the stability of tear film and cause the change of ocular surface, which is similar to the performance of human dry eye, suggesting a novel animal model of dry eye.

Corneal Confocal Microscopy in Type 1 Diabetes Mellitus: A Six-Year Longitudinal Study

Purpose

The current study describes corneal nerve morphology using in vivo confocal microscopy (IVCM) in patients with type 1 diabetes (T1D) who were followed up for 6 years, and it examines the relationship between corneal parameters and metabolic control of glucose and peripheral neuropathy.

Methods

Sixty-two participants (37 with T1D and 25 control participants) were assessed in 2011 and 2017. Participants with bilateral cataract surgery or controls who developed diabetes were excluded. All underwent HbA1c, IVCM, and central corneal sensitivity measurements at both time points in the eye previously examined. A modified total neuropathy score was obtained.

Results

Participants were age and sex matched. The mean duration of diabetes was 32.1 ± 12.0 years at the follow-up visit. The sub-basal nerve density in participants with T1D was lower than that of the controls and did not change (mean ± SD, 11.07 ± 4.0 to 11.41 ± 4.1 mm/mm²; P = 0.71), but it showed a marginal change in controls (19.5 ± 3.7 to 21.63 ± 4.03 mm/mm²; P = 0.06). The corneal sensitivity in T1D did not change (1.3 ± 1.5 to 1.4 ± 1.0 mbar; P = 0.8), and it declined in the controls (0.2 ± 0.3 to 0.6 ± 0.3 mbar; P < 0.001). There were no significant changes in HbA_1c (60.5 ± 12.5 to 61.6 ± 13.7 mmol/mol) or in modified total neuropathy scores (2.4 ± 3.2 to 3.4 ± 3.8; P = 0.2).

Conclusions

The corneal nerve damage and poorer corneal sensitivity reported in the patients with T1D did not change and displayed improvement with good glycemic control.

Translational Relevance

The corneal nerve changes may be of more value in those with a shorter duration of diabetes for the timely prediction of at-risk individuals likely to develop peripheral neuropathy, particularly in type 1 diabetes.

In vivo confocal microscopy qualitative investigation of the relationships between lattice corneal dystrophy deposition and corneal nerves

Background

To investigate the corneal neurotropic phenomenon in patients with lattice corneal dystrophy (LCD) with in vivo laser scanning confocal microscopy (IVCM).

Methods

IVCM was performed on a total of 15 patients (28 eyes) with LCD annually at a follow-up. A collection of the data was acquired to be analyzed.

Results

As indicated by the analysis, the LCD patients’ normal corneal stromal nerves (Grade 0) presented a decline with the prolongation of the follow-ups, corresponding to a gradual increase in grade I and II involving amyloid-wrapped nerve fibers, which demonstrated that the growing amount of amyloid deposit due to the corneal nerve invasion increased slowly over time.

Conclusions

The neurotropic phenomenon could increase with its severity in the corneal lesion of the patients with LCD, and also reflect the distribution of the corneal nerves, to some extent. IVCM provides a rapid, noninvasive way to observe the corneal nerves, which can be an efficient means of better understanding the development of LCD.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12886-021-02149-1.

Segmentation of endothelial cells of the cornea from the distance map of confocal microscope images

We propose a novel algorithm for segmenting cells of the cornea endothelium layer on confocal microscope images. To get an inter-cellular space with minimum gray-scale value and to enhance cell borders, we apply a difference of Gaussian filter before image binarization by thresholding with the minimum gray-scale value. Removal of segmented noise and artifacts is performed by automatic thresholding (using an image frequency analysis to obtain a global threshold value per image). Final segmentation of cells is achieved by fitting the largest inscribed circles into the centers of cell regions defined by the distance map of the binary images. Parameters of interest such as cell count and density, pleomorphism, polymegathism, and F-measure are computed on a publicly available data-set (Confocal Corneal Endothelial Microscopy Data Set - Rotterdam Ophthalmic Data Repository) and compared against the results of the segmentation methods included with the data set, and the results of state of the art automatic methods. The obtained results achieve higher accuracy compared to the results of the segmentation included with the data set (e.g., -proposed versus dataset in R² and mean relative error-, cell count: 0.823, - 0.241 versus 0.017, 0.534; cell density: 0.933, - 0.067 versus 0.154, 0.639; cell polymegathism: 0.652, - 0.079 versus 0.075, 0.886; cell pleomorphism: 0.242, - 0.128 versus 0.0352, - 0.222, respectively), and are in good agreement with the results of the state of the art method.

A Review On digital image processing techniques for in-Vivo confocal images of the cornea

This work reviews the scientific literature regarding digital image processing for in vivo confocal microscopy images of the cornea. We present and discuss a selection of prominent techniques designed for semi- and automatic analysis of four areas of the cornea (epithelium, sub-basal nerve plexus, stroma and endothelium). The main context is image enhancement, detection of structures of interest, and quantification of clinical information. We have found that the preprocessing stage lacks of quantitative studies regarding the quality of the enhanced image, or its effects in subsequent steps of the image processing. Threshold values are widely used in the reviewed methods, although generally, they are selected empirically and manually. The image processing results are evaluated in many cases through comparison with gold standards not widely accepted. It is necessary to standardize values to be quantified in terms of sensitivity and specificity of methods. Most of the reviewed studies do not show an estimation of the computational cost of the image processing. We conclude that reliable, automatic, computer-assisted image analysis of the cornea is still an open issue, constituting an interesting and worthwhile area of research.

Clinical Features, Risk Factors, and Therapy of Epithelial Keratitis after Cataract Surgery

Purpose

The study aimed to assess the clinical characteristics, risk factors, and therapy of epithelial keratitis after cataract surgery.

Methods

Medical data of 89 consecutive patients who developed epithelial keratitis after cataract surgery, including 37 patients with diabetes mellitus (37 eyes) and 52 patients without diabetes mellitus (52 eyes), were retrospectively reviewed. The clinical characteristics, risk factors, and therapy in those patients were evaluated.

Results

The preoperative tear film function determined by the tear breakup time, meibomian gland atrophy score, and low tear meniscus height in diabetic patients was poorer than nondiabetic patients (P < 0.001). Of diabetic patients, 83.78% (31/37) had been diagnosed with meibomian gland dysfunction before cataract surgery and treated with topical nonsteroidal anti-inflammatory drugs after cataract surgery for 44.69 ± 10.51 days, compared to 42.31% (22/52) of nondiabetic patients receiving the topical nonsteroidal anti-inflammatory treatment for 33.35 ± 5.16 days (both P < 0.001). Epithelial lesions progressed within three to four days following cataract surgery in 59.46% (22/37) of diabetic patients, versus 30.77% (16/52) of the nondiabetic patients (P=0.025). Patients with combined meibomian gland dysfunction and epithelial defects accounted for 48.65% (18/37) in the diabetic group and 25.00% (13/52) in the nondiabetic group (P < 0.001). In vivo confocal microscopy showed absence of subbasal never fibers in eyes with epithelial defects, and central corneal sensation was also significantly depressed in those eyes, but there was no significant difference between the two groups (P=0.227). Corneal ulceration and herpes simplex keratitis were found in 2.70% (1/37) and 5.41% (2/37) of diabetic patients, respectively. Amniotic membrane transplantation was required in 32.43% (12/37) of patients in the diabetic group, and the proportion was higher than 1.92% (1/52) in the nondiabetic group (P < 0.001). Average healing time of the corneal epithelium in the diabetic group was 40.62 ± 20.0 days, much longer than 21.74 ± 6.94 days in the nondiabetic group (P=0.002).

Conclusion

Epithelial keratitis after cataract surgery in diabetic patients has the characteristics of rapid development, severe epithelial damage, and slow repair of the corneal epithelium. Amniotic membrane transplantation is a good choice for persistent epithelial defects associated with such epithelial keratitis. Attention should be paid to the tear film function and use of topical nonsteroidal anti-inflammatory drugs in patients undergoing cataract surgery.

Generative Adversarial Network Based Automatic Segmentation of Corneal Subbasal Nerves on In Vivo Confocal Microscopy Images

Purpose

In vivo confocal microscopy (IVCM) is a noninvasive, reproducible, and inexpensive diagnostic tool for corneal diseases. However, widespread and effortless image acquisition in IVCM creates serious image analysis workloads on ophthalmologists, and neural networks could solve this problem quickly. We have produced a novel deep learning algorithm based on generative adversarial networks (GANs), and we compare its accuracy for automatic segmentation of subbasal nerves in IVCM images with a fully convolutional neural network (U-Net) based method.

Methods

We have collected IVCM images from 85 subjects. U-Net and GAN-based image segmentation methods were trained and tested under the supervision of three clinicians for the segmentation of corneal subbasal nerves. Nerve segmentation results for GAN and U-Net-based methods were compared with the clinicians by using Pearson's R correlation, Bland-Altman analysis, and receiver operating characteristics (ROC) statistics. Additionally, different noises were applied on IVCM images to evaluate the performances of the algorithms with noises of biomedical imaging.

Results

The GAN-based algorithm demonstrated similar correlation and Bland-Altman analysis results with U-Net. The GAN-based method showed significantly higher accuracy compared to U-Net in ROC curves. Additionally, the performance of the U-Net deteriorated significantly with different noises, especially in speckle noise, compared to GAN.

Conclusions

This study is the first application of GAN-based algorithms on IVCM images. The GAN-based algorithms demonstrated higher accuracy than U-Net for automatic corneal nerve segmentation in IVCM images, in patient-acquired images and noise applied images. This GAN-based segmentation method can be used as a facilitating diagnostic tool in ophthalmology clinics.

Translational Relevance

Generative adversarial networks are emerging deep learning models for medical image processing, which could be important clinical tools for rapid segmentation and analysis of corneal subbasal nerves in IVCM images.

Optical Coherence Tomography for Ophthalmology Imaging

Optical coherence tomography (OCT) is a depth-resolved imaging modality, which is able to achieve micrometer-scale resolution within biological tissue noninvasively. In the past 30 years, researchers all around the world had made several essential efforts on techniques relevant to OCT. OCT has become a routine process for eye diseases with different types. In this chapter, the three important stages in the development of OCT are briefly illustrated, including the time domain OCT (TD-OCT), the frequency domain OCT (FD-OCT) and the optical coherence tomography angiography (OCTA). Each of the technique has made great progress for use on living human eye imaging in clinical applications. TD-OCT was first proposed and commercialized, which is able to achieve acceptable 2D depth-resolved cross-sectional images of human retina in vivo. FD-OCT was the upgraded OCT technique compared with TD-OCT. By capturing the coherent signal within the Fourier space, the FD-OCT could improve the image sensitivity compared with TD-OCT, and achieve dozens of kilo hertz imaging speed. OCTA is the newest developments of OCT technique, which is able to visualize the micro vasculature networks of human retina in vivo. With OCTA technique, the newest ophthalmologic OCT system is able to achieve detailed diagnosis for both micro-structure and vasculature abnormalities for clinical applications. The further development of OCT technique on imaging speed, contrast, resolution, field of view, and so on will make OCT to be a more powerful tool for clinical usages.

Comparison of parameter agreement for characterization of corneal subbasal nerve plexus in the whorl-like region and central cornea using in vivo confocal microscopy

PURPOSE

To compare the reliability of the whorl-like region with that of the central cornea for accurate assessment of corneal subbasal nerve plexus (SNP) by analyzing the parameter variability of these two anatomical regions in repeated measurements.

METHODS

Participants were scanned in the central cornea and whorl-like region with in vivo confocal microscopy on three occasions by two examiners within a time span of one week. Coefficients of repeatability (CoR), intra-class correlation coefficient (ICC), and Bland-Altman scatter plots with 95% limits of agreement (LOA) in the central cornea and whorl-like region were calculated, respectively, based on the nerve fiber length, then the inter-observer and intra-observer agreement were compared between these two anatomical regions.

RESULTS

The inter-observer ICC was 0.945, the inter-observer CoR was 0.052, the intra-observer ICC was 0.936, and the inter-observer CoR was 0.046, with narrow 95% LOA within 1 standard deviation in the whorl-like region, whereas the inter-observer ICC was 0.600, the inter-observer CoR was 0.207, the intra-observer ICC was 0.206, and the intra-observer CoR was 0.253, with 95% LOA nearly threefold wider than the standard deviation in the central cornea.

CONCLUSIONS

Nerve parameter in the whorl-like region showed higher inter-observer and intra-observer agreement than that of the central cornea. The whorl-like region is a more reliable site for accurate assessment of SNP.

Recent Advances in Stem Cell Therapy for Limbal Stem Cell Deficiency: A Narrative Review

Destruction of the limbus and depletion of limbal stem cells (LSCs), the adult progenitors of the corneal epithelium, leads to limbal stem cell deficiency (LSCD). LSCD is a rare, progressive ocular surface disorder which results in conjunctivalisation and neovascularisation of the corneal surface. Many strategies have been used in the treatment of LSCD, the common goal of which is to regenerate a self-renewing, transparent, and uniform epithelium on the corneal surface. The development of these techniques has frequently resulted from collaboration between stem cell translational scientists and ophthalmologists. Direct transplantation of autologous or allogeneic limbal tissue from a healthy donor eye is regarded by many as the technique of choice. Expansion of harvested LSCs in vitro allows smaller biopsies to be taken from the donor eye and is considered safer and more acceptable to patients. This technique may be utilised in unilateral cases (autologous) or bilateral cases (living related donor). Recently developed, simple limbal epithelial transplant (SLET) can be performed with equally small biopsies but does not require in vitro cell culture facilities. In the case of bilateral LSCD, where autologous limbal tissue is not available, autologous oral mucosa epithelium can be expanded in vitro and transplanted to the diseased eye. Data on long-term outcomes (over 5 years of follow-up) for many of these procedures is needed, and it remains unclear how they produce a self-renewing epithelium without recreating the vital stem cell niche. Bioengineering techniques offer the ability to re-create the physical characteristics of the stem cell niche, while induced pluripotent stem cells offer an unlimited supply of autologous LSCs. In vivo confocal microscopy and anterior segment OCT will complement impression cytology in the diagnosis, staging, and follow-up of LSCD. In this review we analyse recent advances in the pathology, diagnosis, and treatment of LSCD.

Corneal fibroblasts: Function and markers

Corneal stromal keratocytes contribute to the maintenance of corneal transparency and shape by synthesizing and degrading extracellular matrix. They are quiescent in the healthy cornea, but they become activated in response to insults from the external environment that breach the corneal epithelium, with such activation being associated with phenotypic transformation into fibroblasts. Corneal fibroblasts (activated keratocytes) act as sentinel cells to sense various external stimuli-including damage-associated molecular patterns derived from injured cells, pathogen-associated molecular patterns of infectious microorganisms, and inflammatory mediators such as cytokines-under pathological conditions such as trauma, infection, and allergy. The expression of various chemokines and adhesion molecules by corneal fibroblasts determines the selective recruitment and activation of inflammatory cells in a manner dependent on the type of insult. In infectious keratitis, the interaction of corneal fibroblasts with various components of microbes and with cytokines derived from infiltrated inflammatory cells results in excessive degradation of stromal collagen and consequent corneal ulceration. Corneal fibroblasts distinguish between type 1 and type 2 inflammation through recognition of corresponding cytokines, with their activation by type 2 cytokines contributing to the pathogenesis of corneal lesions in severe ocular allergic diseases. Pharmacological targeting of corneal fibroblast function is thus a potential novel therapeutic approach to prevention of excessive corneal stromal inflammation, damage, and scarring.

Micromirror-Embedded Coverslip Assembly for Bidirectional Microscopic Imaging

3D imaging of a biological sample provides information about cellular and subcellular structures that are important in cell biology and related diseases. However, most 3D imaging systems, such as confocal and tomographic microscopy systems, are complex and expensive. Here, we developed a quasi-3D imaging tool that is compatible with most conventional microscopes by integrating micromirrors and microchannel structures on coverslips to provide bidirectional imaging. Microfabricated micromirrors had a precisely 45° reflection angle and optically clean reflective surfaces with high reflectance over 95%. The micromirrors were embedded on coverslips that could be assembled as a microchannel structure. We demonstrated that this simple disposable device allows a conventional microscope to perform bidirectional imaging with simple control of a focal plane. Images of microbeads and cells under bright-field and fluorescent microscopy show that the device can provide a quick analysis of 3D information, such as 3D positions and subcellular structures.

In vivo imaging of corneal nerves and cellular structures in mice with Gabor-domain optical coherence microscopy

Gabor-domain optical coherence microscopy (GDOCM) demonstrated in vivo corneal imaging with cellular resolution and differentiation in mice over a field of view of 1 mm². Contact and non-contact imaging was conducted on six healthy and six hyperglycemic C57BL/6J mice. Cellular resolution in the 3D GDOCM images was achieved after motion correction. Corneal nerve fibers were traced and their lengths and branches calculated. Noncontact, label-free imaging of corneal nerves has clinical utility in health and disease, and in transplant evaluation. To the authors' knowledge, this is the first report of in vivo 3D corneal imaging in mice with the capability to resolve nerve fibers using a non-contact imaging modality.

In Vivo Confocal Microscopy of Cornea in Patients with Terrien's Marginal Corneal Degeneration

This study was aimed at observing the morphological changes of the cornea with ocular in vivo confocal microscopy (IVCM) in patients with Terrien's marginal degeneration (TMD). Ten patients (20 eyes) with TMD treated in the Department of Ophthalmology, Xiangya Hospital, and 10 healthy controls (20 eyes) were included in the current study. A detailed slit lamp microscopy, anterior segment photography, and corneal IVCM examination were performed for each eye. The density of central and marginal corneal epithelial cells, stromal cells, and subepithelial nerve fibers was compared between the two groups using the Wilcoxon rank sum test. Compared with the control group, the corneal epithelial and endothelial cells in the TMD group showed granular highly reflective substances and thinner subepithelial nerve fibers. The uneven dot-like highly reflective substances without cell structures appeared in the stromal layer of the cornea. The density of central and marginal corneal epithelial cells, stromal cells, and subepithelial nerve fibers was lower in the TMD group (p < 0.05), and they were negatively correlated with severity of the disease (p < 0.05). Our study demonstrated that the density of corneal epithelial cells, stromal cells, and sensory plexus nerve fibers was significantly reduced in the TMD group. The pathological changes were more obvious in the marginal cornea, and it is correlated with severity of the disease.

Efficacy of autologous serum tears for treatment of neuropathic corneal pain

OBJECTIVE

Corneal nerve damage may result in neuropathic corneal pain (NCP). Autologous serum tears (AST) have been shown to results in nerve regeneration and may help alleviate corneal pain. This study aimed to evaluate the efficacy of AST in the treatment of NCP.

METHODS

This was a retrospective case-control study. Sixteen patients suffering from severe NCP and no current ocular surface disease were compared to 12 controls. In vivo confocal microscopy (IVCM) (HRT3/RCM; Heidelberg Engineering GmbH, Germany) of the central corneas was performed bilaterally. Change in pain severity (scale of 0-10), corneal nerve density, tortuosity, reflectivity and presence of beading and micro-neuromas before and after treatment were recorded.

RESULTS

All patients had severe pain, with a mean of 9.1 ± 0.2 (range 8-10). Subbasal nerves were significantly decreased before treatment as compared to controls, including total nerve length (10,935.5 ± 1264.3 vs. 24,714.4 ± 1056.2 μm/mm²; p < 0.0001) and total number of nerves (10.5 ± 1.4 vs. 28.6 ± 2.0; p < 0.0001), respectively. Morphologically, significantly increased reflectivity (2.9 ± 0.2 vs. 1.2 ± 0.1; p = 0.00008) and tortuosity (2.4 ± 0.2 vs. 1.7 ± 0.1; p = 0.001), both graded on a scale of 0-4, were noted. After a mean of 3.8 ± 0.5 months (range 1-8 months) of AST treatment, pain severity decreased to 3.1 ± 0.3 (range 0-4), (p < 0.0001). Further, IVCM demonstrated a significant improvement (p < 0.005) in total nerve length (17,351.3 ± 1395.6  μm/mm²) and number (15.1 ± 1.6), as well as significant decrease in reflectivity (2.4 ± 0.2; p = 0.001) and tortuosity (2.2 ± 0.2; p = 0.001).

CONCLUSION

IVCM demonstrates underlying alterations of the subbasal corneal nerve plexus in patients suffering from debilitating NCP. AST-induced nerve regeneration is seen following treatment with AST, which correlates with improvement in patient symptoms of NCP.

Review and analysis of grading scales for ocular surface staining

Vital dye staining has been used for over a century to assess the severity of ocular surface disease. However, despite common usage, a universally accepted "gold standard" grading scale does not exist for corneal and conjunctival staining, which can impact the ability to diagnose and monitor ocular surface conditions such as dry eye. The Food and Drug Administration (FDA) and other international regulatory agencies rely on ocular surface staining as a primary endpoint for new drug approvals, so that absence of a "gold standard" scale may affect approval of new drug treatments. To begin to address this problem, we review existing, published grading scales in an integrated fashion, highlighting their differences and similarities to emphasize common themes and the methods and elements that are important in creating a standardized scale. Our goal is to aid the field in moving towards an accepted standardized grading scale for ocular surface staining that can be applied in clinic and research settings for a variety of ocular conditions.

Pathogenesis of Keratoconus: The intriguing therapeutic potential of Prolactin-inducible protein

Keratoconus (KC) is the most common ectatic corneal disease, with clinical findings that include discomfort, visual disturbance and possible blindness if left untreated. KC affects approximately 1:400 to 1:2000 people worldwide, including both males and females. The aetiology and onset of KC remains a puzzle and as a result, the ability to treat or reverse the disease is hampered. Sex hormones are known to play a role in the maintenance of the structure and integrity of the human cornea. Hormone levels have been reported to alter corneal thickness, curvature, and sensitivity during different times of menstrual cycle. Surprisingly, the role of sex hormones in corneal diseases and KC has been largely neglected. Prolactin-induced protein, known to be regulated by sex hormones, is a new KC biomarker that has been recently proposed. Studies herein discuss the role of sex hormones as a control mechanism for KC onset and progression and evidence supporting the view that prolactin-induced protein is an important hormonally regulated biomarker in KC is discussed.

A contralateral eye study comparing characteristics of corneal endothelial cells in bilateral keratoconus patients with unilateral corneal Vogt's striae

Purpose

The aim of this study was to analyze and compare corneal endothelial cell morphology and characteristics in bilateral keratoconus (KCN) patients with unilateral Vogt's striae.

Methods

Fifty patients aged 20–38 years were recruited in this cross-sectional contralateral eye study. In this study, corneal endothelial cell parameters were evaluated in patients with bilateral KCN and unilateral Vogt's striae using the Topcon SP2000P specular microscope (Topcon, Tokyo, Japan).

Results

In the current study, there were no significant differences in corneal endothelial cell parameters including endothelial cell density (ECD), hexagonal cell ratio (HEX), and coefficient of variance of cell size (CV) between the KCN groups with and without Vogt's striae, [(2968.34 ± 276.65 vs. 2980.05 ± 253.30, P = 0.618), (51.88 ± 13.57 vs. 53.24 ± 9.31, P = 0.658), and (32.50 ± 5.40 vs. 32.97 ± 4.07, P = 0.467), respectively]. Also, among study groups with and without Vogt's striae, ECD did not correlate with anterior chamber depth (ACD) [(P = 0.564, r = 0.09), (P = 0.219, r = −0.18), respectively], maximum keratometry (Kmax) [(P = 0.215, r = 0.18), (P = 0.898, r = 0.02), respectively], and central corneal thickness (CCT) [(P = 0.989, r = −0.02), (P = 0.643, r = −0.07), respectively].

Our results showed significant differences in corrected and uncorrected distance visual acuity (UDVA), cycloplegic refractive error components (calculated by vectorial analysis), CCT, and Kmax between two study groups (all P < 0.05) except for J₄₅ (Jackson cross cylinder, axes at 45 and 135°) (P = 0.131).

Conclusions

We were not able to find the statistically significant differences in ECD, HEX, and CV between KCN eyes with and without Vogt's striae. Despite clinical and tomographic results, it seems that Vogt's striae cannot cause deterioration in the corneal endothelial morphology.

Corneal confocal sub-basal nerve plexus evaluation: a review

The aim of this study was to review the most recent data about corneal sub-basal nerve plexus (SNP) evaluated with the use of corneal confocal microscopy (CCM). For this purpose, an electronic search was conducted based on PubMed and Google Scholar and Web of Science databases from 2008 up to the end of 2016. Ninety-eight articles in English were cited, as well as abstracts in other languages, concerning the morphology and function of corneal SNP in various diseases. Changes in corneal SNP as a result of local treatment were also introduced. Figures with scans from confocal microscopy from our Department were included. The main conclusion of this review was that both corneal SNP diminishment and high tortuosity as well as low sensitivity are in principle related to the presence or level of pathology. In addition, increased nerve tortuosity may represent a morphological determinant of nerve regeneration. However, the presented literature shows that SNP changes are not characteristic for one unified corneal pathology; rather, they reflect the non-specific pathological process present in many diseases. Future studies should use automatized biometric software and also examine the effects of new treatments on SNP.

Corneal nerve regeneration after herpes simplex keratitis: A longitudinal in vivo confocal microscopy study

PURPOSE

To evaluate the long-term alterations of corneal nerves in patients with herpes simplex virus (HSV) keratitis using in vivo confocal microscopy (IVCM).

DESIGN

Prospective, longitudinal, cross sectional.

METHODS

This study included 16 patients with a history of HSV keratitis and 15 age-matched normal controls. Slit-scanning IVCM was performed in all subjects at baseline and then after a mean follow-up of 37.3 ± 1.7 months in the patient group. Corneal subbasal nerve density and corneal sensation were compared between groups at baseline and follow-up.

RESULTS

At baseline, the mean subbasal nerve density was significantly lower in both affected eyes (1.4 ± 0.6 mm/mm²) and contralateral unaffected eyes (6.4 ± 0.7 mm/mm²) compared with the controls (14.1 ± 1.6 mm/mm²; all P < .001). At the end of follow-up, the mean nerve density in affected eyes increased to 2.8 ± 0.7 mm/mm² (P = .006), with no significant change in contralateral unaffected eyes (6.5 ± 1.0 mm/mm², P = .72). However, both eyes had lower nerve density than controls (all P < .001). Corneal sensation was significantly lower in affected eyes (2.6 ± 0.6 cm) than in the control group (6.0 ± 0.0, P < .001) and showed no significant change at the end of follow-up (2.5 ± 0.6 cm, P = .80). Corneal sensation in contralateral unaffected eyes was not different in comparison with controls at both baseline and follow up (all p > .05).

CONCLUSIONS

Our results demonstrate that although corneal nerve regeneration occurs in patients with HSV keratitis, this change is not clinically significant and does not results in changes of corneal sensation. Therefore, these patients need to be followed closely for complications of neurotrophic keratopathy and might benefit from neuro-regenerative therapies.

Automatic analysis of corneal nerves imaged using in vivo confocal microscopy

Interest has grown over the past decade in using in vivo confocal microscopy to analyse the morphology of corneal nerves and their changes over time. Advances in computational modelling techniques have been applied to automate the estimation of sub-basal nerve structure. These objective methods have the potential to quantify nerve density (and length), tortuosity, variations in nerve thickness, as well as temporal changes in nerve fibres such as migration patterns. Different approaches to automated nerve analysis, methods proposed and how they were validated in previous literature are reviewed. Improved understanding of these approaches and their limitations will help improve the diagnostic leverage of emerging developments for monitoring the onset and progression of a broad class of systemic diseases, including diabetes.

Corneal Nerve Fiber Structure, Its Role in Corneal Function, and Its Changes in Corneal Diseases

Recently, in vivo confocal microscopy is used to examine the human corneal nerve fibers morphology. Corneal nerve fiber architecture and its role are studied in healthy and pathological conditions. Corneal nerves of rats were studied by nonspecific acetylcholinesterase (NsAchE) staining. NsAchE-positive subepithelial (stromal) nerve fiber has been found to be insensitive to capsaicin. Besides, NsAchE-negative but capsaicin-sensitive subbasal nerve (leash) fibers formed thick mesh-like structure showing close interconnections and exhibit both isolectin B4- and transient receptor potential vanilloid channel 1- (TRPV1-) positive. TRPV1, TRPV3, TRPA (ankyrin) 1, and TRPM (melastatin) 8 are expressed in corneal nerve fibers. Besides the corneal nerve fibers, the expressions of TRPV (1, 3, and 4), TRPC (canonical) 4, and TRPM8 are demonstrated in the corneal epithelial cell membrane. The realization of the importance of TRP channels acting as polymodal sensors of environmental stresses has identified potential drug targets for corneal disease. The pathophysiological conditions of corneal diseases are associated with disruption of normal tissue innervation, especially capsaicin-sensitive small sensory nerve fibers. The relationships between subbasal corneal nerve fiber morphology and neurotrophic keratopathy in corneal diseases are well studied. The recommended treatment for neurotrophic keratopathy is administration of preservative free eye drops.

Laser scanning in vivo confocal microscopy (IVCM) for evaluating human corneal sub-basal nerve plexus parameters: protocol for a systematic review

INTRODUCTION

Laser scanning in vivo confocal microscopy (IVCM) enables non-invasive, high-resolution imaging of the cornea. In recent years, there has been a vast increase in researchers using laser scanning IVCM to image and quantify corneal nerve parameters. However, a range of methodological approaches have been adopted. The primary aim of this systematic review is to critically appraise the reported method(s) of primary research studies that have used laser scanning IVCM to quantify corneal sub-basal nerve plexus (SBNP) parameters in humans, and to examine corneal nerve parameters in healthy individuals.

METHODS AND ANALYSIS

A systematic review of primary studies that have used laser scanning IVCM to quantify SBNP parameters in humans will be conducted. Comprehensive electronic searches will be performed in Ovid MedLine, Embase and the Cochrane Library. Two reviewers will independently assess titles and abstracts, and exclude studies not meeting the inclusion criteria. For studies judged eligible or potentially eligible, full texts will be independently assessed by two reviewers to determine eligibility. A third reviewer will resolve any discrepancies in judgement. Risk of bias will be assessed using a custom tool, covering five methodological domains: participant selection, method of image capture, method of image analysis, data reporting and other sources of bias. A systematic narrative synthesis of findings will be provided. A multilevel random-effects meta-analysis will be performed for corneal nerve parameters derived from healthy participants. This review will be reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement.

ETHICS AND DISSEMINATION

As this review considers published data, ethical approval is not required. We foresee that this synthesis will serve as a reference for future studies, and can be used to inform best practice standards for using IVCM in clinical research. A manuscript reporting the results of the review will be published and may also be presented at scientific conferences.

Corneal Nerve Regeneration after Self-Retained Cryopreserved Amniotic Membrane in Dry Eye Disease

Purpose

To evaluate the efficacy of self-retained cryopreserved amniotic membrane (CAM) in promoting corneal nerve regeneration and improving corneal sensitivity in dry eye disease (DED).

Methods

In this prospective randomized clinical trial, subjects with DED were randomized to receive CAM (study group) or conventional maximum treatment (control). Changes in signs and symptoms, corneal sensitivity, topography, and in vivo confocal microscopy (IVCM) were evaluated at baseline, 1 month, and 3 months.

Results

Twenty subjects (age 66.9 ± 8.9) were enrolled and 17 completed all follow-up visits. Signs and symptoms were significantly improved in the study group yet remained constant in the control. IVCM showed a significant increase in corneal nerve density in the study group (12,241 ± 5083 μm/mm² at baseline, 16,364 ± 3734 μm/mm² at 1 month, and 18,827 ± 5453 μm/mm² at 3 months, p = 0.015) but was unchanged in the control. This improvement was accompanied with a significant increase in corneal sensitivity (3.25 ± 0.6 cm at baseline, 5.2 ± 0.5 cm at 1 month, and 5.6 ± 0.4 cm at 3 months, p < 0.001) and corneal topography only in the study group.

Conclusions

Self-retained CAM is a promising therapy for corneal nerve regeneration and accelerated recovery of the ocular surface health in patients with DED. The study is registered at clinicaltrials.gov with trial identifier: NCT02764814.

Automated morphometric description of human corneal endothelium from in-vivo specular and confocal microscopy

Corneal images acquired by in-vivo specular and confocal microscopy provide clinical information on the cornea endothelium health state. Indeed, the normal hexagonal shape of endothelial cells is usually affected by age and pathologies. At present, the analysis is based on manual or semi-automatic methods and the segmentation of a large number of endothelial cells is required for a meaningful estimation of the clinical parameters. We propose an automated method that detects the centers of endothelial cells by convolving the original image with customized two-dimensional kernels, derives a structure made by connected vertices from the recognized centers using the Euclidean distance, and fine-tunes the derived structure through a genetic algorithm, which combines information about the typical regularity of endothelial cells shape with the pixels intensity of the actual image. The final structure of connected vertices forms a set of polygons that fit the underlying cells contours. From these contours the morphometric parameters of clinical interest can be easily computed. The procedure was applied to 15 images acquired with the SP-3000P (Topcon, Japan) specular microscope and 15 images acquired with the Confoscan4 (Nidek Technologies, Italy) confocal microscope, from both healthy and pathological subjects. Ground truth values for the morphometric parameters were obtained from manually carefully drawn cell contours. Results show that the mean percent absolute difference between the automated and the manual estimate of the clinical parameters is between 2 and 6%, and no statistically significant difference was found between them. The proposed totally automatic method appears capable of detecting contour of hundreds of cells covering a large area, and of obtaining a reliable estimation of the relevant morphometric parameters used in clinical practice.

In Vivo Confocal Microscopy of Corneal Nerves in Health and Disease

In vivo confocal microscopy (IVCM) is becoming an indispensable tool for studying corneal physiology and disease. Enabling the dissection of corneal architecture at a cellular level, this technique offers fast and noninvasive in vivo imaging of the cornea with images comparable to those of ex vivo histochemical techniques. Corneal nerves bear substantial relevance to clinicians and scientists alike, given their pivotal roles in regulation of corneal sensation, maintenance of epithelial integrity, as well as proliferation and promotion of wound healing. Thus, IVCM offers a unique method to study corneal nerve alterations in a myriad of conditions, such as ocular and systemic diseases and following corneal surgery, without altering the tissue microenvironment. Of particular interest has been the correlation of corneal subbasal nerves to their function, which has been studied in normal eyes, contact lens wearers, and patients with keratoconus, infectious keratitis, corneal dystrophies, and neurotrophic keratopathy. Longitudinal studies have applied IVCM to investigate the effects of corneal surgery on nerves, demonstrating their regenerative capacity. IVCM is increasingly important in the diagnosis and management of systemic conditions such as peripheral diabetic neuropathy and, more recently, in ocular diseases. In this review, we outline the principles and applications of IVCM in the study of corneal nerves in various ocular and systemic diseases.

Temporal and spatial analysis of stromal cell and extracellular matrix patterning following lamellar keratectomy

Extracellular matrix (ECM) supplies both physical and chemical signals to keratocytes which can impact their differentiation to fibroblasts and/or myofibroblasts. It also provides a substrate through which they migrate during wound repair. We have previously shown that following transcorneal freeze injury (FI), migrating corneal fibroblasts align parallel to the stromal lamellae during wound repopulation. In this study, we compare cell and ECM patterning both within and on top of the stroma at different time points following lamellar keratectomy (LK) in the rabbit. Twelve rabbits received LK in one eye. Rabbits were monitored using in vivo confocal microscopy at 3, 7, 21 and 60 days after injury. A subset of animals was sacrificed at each time point to further investigate cell and matrix patterning. Tissue was fixed and labeled in situ with Alexa Fluor 488 phalloidin (for F-actin), and imaged using multiphoton fluorescence and second harmonic generation (SHG) imaging (for collagen). Immediately following LK, cell death occurred in the corneal stroma directly beneath the injury. At 7 and 21 days after LK, analysis of fluorescence (F-actin) and SHG results (collagen) indicated that fibroblasts were co-aligned with the collagen lamellae within this region. In contrast, stromal cells accumulating on top of the stromal wound bed were randomly arranged, contained more prominent stress fibers, and expressed alpha smooth muscle actin (α-SMA) and fibronectin. At 60 days, cells and matrix in this region had become co-aligned into lamellar-like structures; cells were elongated but did not express stress fibers. Corneal haze measured using in vivo confocal microscopy peaked at 21 days after LK, and was significantly reduced by 60 days. Cell morphology and patterning observed in vivo was similar to that observed in situ. Our results suggest that the topography and alignment of the collagen lamellae direct fibroblast patterning during repopulation of the native stroma after LK injury in the rabbit. In contrast, stromal cells accumulating on top of the stromal wound bed initially align randomly and produce a fibrotic ECM. Remarkably, over time, these cells appear to remodel the ECM to produce a lamellar structure that is similar to the native corneal stroma.

Abnormal Anterior Corneal Morphology in Diabetes Observed Using In Vivo Laser-scanning Confocal Microscopy

PURPOSE

To assess if diabetes alters corneal epithelial, anterior stromal and subbasal nerve plexus morphology and to determine the associations between these and other clinical variables.

METHODS

A cohort of 78 participants with diabetes (39 with Type 1 and 39 with Type 2 diabetes) and 29 age-matched healthy controls underwent laser-scanning confocal microscopy of the central cornea. Intermediate cell density (ICD), basal cell density (BCD), anterior stromal cell density (ASCD), corneal nerve fiber density (CNFD) and nerve fiber length (CNFL) were quantified.

RESULTS

Compared with controls, participants with diabetes showed reduced ICD (6097 ± 669 vs. 5548 ± 669 no/mm², P<.01), BCD (8925 ± 1196 vs. 7842 ± 1040 cell/mm², P<.01), CNFD (23.4 ± 9.1 vs. 17.5 ± 9.7 no/mm², P<.01) and CNFL (21.0 ± 4.0 vs. 17.4 ± 4.9 mm/mm², P<.01), with no difference in ASCD (785 ± 262 vs. 733 ± 278 cell/mm², respectively, P=.40). None of these structural parameters were associated with type of diabetes (P>.06). Multiple regression analysis showed that ICD and BCD were inversely associated with the diabetes duration and diastolic blood pressure (P<.05) and positively associated with CNFD (P<.01). CNFD and CNFL were inversely associated with HbA_1c (P<.01), while ASCD was inversely associated with age (P<.01).

CONCLUSIONS

Corneal epithelial cells and subbasal nerve fibers are reduced in patients with diabetes; however, anterior stromal cells show no difference. Furthermore, abnormalities in corneal epithelial cells and nerves are interrelated and correlated with modifiable risk factors.

Two-Dimensional Plane for Multi-Scale Quantification of Corneal Subbasal Nerve Tortuosity

PURPOSE

To assess the performance of a novel system for automated tortuosity estimation and interpretation.

METHODS

A supervised strategy (driven by observers' grading) was employed to automatically identify the combination of tortuosity measures (i.e., tortuosity representation) leading to the best agreement with the observers. We investigated 18 tortuosity measures including curvature and density of inflection points, computed at multiple spatial scales. To leverage tortuosity interpretation, we propose the tortuosity plane (TP) onto which each image is mapped. Experiments were carried out on 140 images of subbasal nerve plexus of the central cornea, covering four levels of tortuosity. Three experienced observers graded each image independently.

RESULTS

The best tortuosity representation was the combination of mean curvature at spatial scales 2 and 5. These tortuosity measures were the axes of the proposed TP (interpretation). The system for tortuosity estimation revealed strong agreement with the observers on a global and per-level basis. The agreement with each observer (Spearman's correlation) was statistically significant (αs = 0.05, P < 0.0001) and higher than that of at least one of the other observers in two out of three cases (ρOUR = 0.7594 versus ρObs3 = 0.7225; ρOUR = 0.8880 versus ρObs1 = 0.8017, ρObs3 = 0.7315). Based on paired-sample t-tests, these improvements were significant (P < 0.001).

CONCLUSIONS

Our automated system stratifies images by four tortuosity levels (discrete scale) matching or exceeding the accuracy of experienced observers. Of importance, the TP allows the assessment of tortuosity on a two-dimensional continuous scale, thus leading to a finer discrimination among images.

Comparison of Standard Versus Wide-Field Composite Images of the Corneal Subbasal Layer by In Vivo Confocal Microscopy

PURPOSE

To evaluate whether the densities of corneal subbasal nerves and epithelial immune dendritiform cells (DCs) are comparable between a set of three representative standard images of in vivo confocal microscopy (IVCM) and the wide-field mapped composite IVCM images.

METHODS

This prospective, cross-sectional, and masked study included 110 eyes of 58 patients seen in a neurology clinic who underwent laser-scanning IVCM (Heidelberg Retina Tomograph 3) of the central cornea. Densities of subbasal corneal nerves and DCs were compared between the average of three representative standard images and the wide-field mapped composite images, which were reconstructed by automated mapping.

RESULTS

There were no statistically significant differences between the average of three representative standard images (0.16 mm2 each) and the wide-field composite images (1.29 ± 0.64 mm2) in terms of mean subbasal nerve density (17.10 ± 6.10 vs. 17.17 ± 5.60 mm/mm2, respectively, P = 0.87) and mean subbasal DC density (53.2 ± 67.8 vs. 49.0 ± 54.3 cells/mm2, respectively, P = 0.43). However, there were notable differences in subbasal nerve and DC densities between these two methods in eyes with very low nerve density or very high DC density.

CONCLUSIONS

There are no significant differences in the mean subbasal nerve and DC densities between the average values of three representative standard IVCM images and wide-field mapped composite images. Therefore, these standard images can be used in clinical studies to accurately measure cellular structures in the subbasal layer.