Quantitative assessment of corneal vibrations during intraocular pressure measurement with the air-puff method in patients with keratoconus

CorNet: Autonomous feature learning in raw Corvis ST data for keratoconus diagnosis via residual CNN approach

PURPOSE

To ascertain whether the integration of raw Corvis ST data with an end-to-end CNN can enhance the diagnosis of keratoconus (KC).

METHOD

The Corvis ST is a non-contact device for in vivo measurement of corneal biomechanics. The CorNet was trained and validated on a dataset consisting of 1786 Corvis ST raw data from 1112 normal eyes and 674 KC eyes. Each raw data consists of the anterior and posterior corneal surface elevation during air-puff induced dynamic deformation. The architecture of CorNet utilizes four ResNet-inspired convolutional structures that employ 1 × 1 convolution in identity mapping. Gradient-weighted Class Activation Mapping (Grad-CAM) was adopted to visualize the attention allocation to diagnostic areas. Discriminative performance was assessed using metrics including the AUC of ROC curve, sensitivity, specificity, precision, accuracy, and F1 score.

RESULTS

CorNet demonstrated outstanding performance in distinguishing KC from normal eyes, achieving an AUC of 0.971 (sensitivity: 92.49%, specificity: 91.54%) in the validation set, outperforming the best existing Corvis ST parameters, namely the Corvis Biomechanical Index (CBI) with an AUC of 0.947, and its updated version for Chinese populations (cCBI) with an AUC of 0.963. Though the ROC curve analysis showed no significant difference between CorNet and cCBI (p = 0.295), it indicated a notable difference between CorNet and CBI (p = 0.011). The Grad-CAM visualizations highlighted the significance of corneal deformation data during the loading phase rather than the unloading phase for KC diagnosis.

CONCLUSION

This study proposed an end-to-end CNN approach utilizing raw biomechanical data by Corvis ST for KC detection, showing effectiveness comparable to or surpassing existing parameters provided by Corvis ST. The CorNet, autonomously learning comprehensive temporal and spatial features, demonstrated a promising performance for advancing KC diagnosis in ophthalmology.

Accuracy of Machine Learning Assisted Detection of Keratoconus: A Systematic Review and Meta-Analysis

(1) Background: The objective of this review was to synthesize available data on the use of machine learning to evaluate its accuracy (as determined by pooled sensitivity and specificity) in detecting keratoconus (KC), and measure reporting completeness of machine learning models in KC based on TRIPOD (the transparent reporting of multivariable prediction models for individual prognosis or diagnosis) statement. (2) Methods: Two independent reviewers searched the electronic databases for all potential articles on machine learning and KC published prior to 2021. The TRIPOD 29-item checklist was used to evaluate the adherence to reporting guidelines of the studies, and the adherence rate to each item was computed. We conducted a meta-analysis to determine the pooled sensitivity and specificity of machine learning models for detecting KC. (3) Results: Thirty-five studies were included in this review. Thirty studies evaluated machine learning models for detecting KC eyes from controls and 14 studies evaluated machine learning models for detecting early KC eyes from controls. The pooled sensitivity for detecting KC was 0.970 (95% CI 0.949–0.982), with a pooled specificity of 0.985 (95% CI 0.971–0.993), whereas the pooled sensitivity of detecting early KC was 0.882 (95% CI 0.822–0.923), with a pooled specificity of 0.947 (95% CI 0.914–0.967). Between 3% and 48% of TRIPOD items were adhered to in studies, and the average (median) adherence rate for a single TRIPOD item was 23% across all studies. (4) Conclusions: Application of machine learning model has the potential to make the diagnosis and monitoring of KC more efficient, resulting in reduced vision loss to the patients. This review provides current information on the machine learning models that have been developed for detecting KC and early KC. Presently, the machine learning models performed poorly in identifying early KC from control eyes and many of these research studies did not follow established reporting standards, thus resulting in the failure of these clinical translation of these machine learning models. We present possible approaches for future studies for improvement in studies related to both KC and early KC models to more efficiently and widely utilize machine learning models for diagnostic process.

Corneal Biomechanics Computational Analysis for Keratoconus Diagnosis

For machine learning techniques to be used in early keratoconus diagnosis, researchers aimed to find and model representations of corneal biomechanical characteristics from exam images generated by the Corvis ST. Image segments were used to identify and convert anterior data into vectors for representation and representation of apparent posterior surfaces, apparent pachymetry, and the composition of apparent anterior data in images. Chained (batch images) and simplified with wavelet, the vectors were also arranged as 2D histograms for deep learning use in a neural network. An interval of 0.7843 to 1 and a significance level of 0.0157 were used in the scoring, with the classifications getting points for being as sensitive as they could be while also being as precise as they could be. In order to train and validate the used data from examination bases in Europe and Iraq, in grades I to IV, researchers looked at data from 686 healthy eyes and 406 keratoconus-afflicted eyes. With a score of 0.8247, sensitivity of 89.49%, and specificity of 92.09%, the European database found that apparent pachymetry from batch images applied with level 4 wavelet and processed quickly had the highest accuracy. This is a 2D histogram of apparent pachymetry with a score of 0.8361, which indicates that it is 88.58 percent sensitive and 94.389% specific. According to the findings, keratoconus can be diagnosed using biomechanical models.

A new method for detecting the outer corneal contour in images from an ultra-fast Scheimpflug camera

Background

The Corvis^® ST tonometer is an innovative device which, by combining a classic non-contact tonometer with an ultra-fast Scheimpflug camera, provides a number of parameters allowing for the assessment of corneal biomechanics. The acquired biomechanical parameters improve medical diagnosis of selected eye diseases. One of the key elements in biomechanical measurements is the correct corneal contour detection, which is the basis for further calculations. The presented study deals with the problem of outer corneal edge detection based on a series of images from the afore-mentioned device. Corneal contour detection is the first and extremely important stage in the acquisition and analysis of corneal dynamic parameters.

Result

A total of 15,400 images from the Corvis^® ST tonometer acquired from 110 patients undergoing routine ophthalmologic examinations were analysed. A method of outer corneal edge detection on the basis of a series of images from the Corvis^® ST was proposed. The method was compared with known and commonly used edge detectors: Sobel, Roberts, and Canny operators, as well as others, known from the literature. The analysis was carried out in MATLAB^® version 9.0.0.341360 (R2016a) with the Image Processing Toolbox (version 9.4) and the Neural Network Toolbox (version 9.0). The method presented in this paper provided the smallest values of the mean error (0.16%), stability (standard deviation 0.19%) and resistance to noise, characteristic for Corvis^® ST tonometry tests, compared to the methods known from the literature. The errors were 5.78 ± 9.19%, 3.43 ± 6.21%, and 1.26 ± 3.11% for the Roberts, Sobel, and Canny methods, respectively.

Conclusions

The proposed new method for detecting the outer corneal contour increases the accuracy of intraocular pressure measurements. It can be used to analyse dynamic parameters of the cornea.

Optical measurement of the corneal oscillation for the determination of the intraocular pressure

Motivation

Glaucoma is currently the most common irreversible cause of blindness worldwide. A significant risk factor is an individually increased intraocular pressure (IOP). A precise measurement method is needed to determine the IOP in order to support the diagnosis of the disease and to monitor the outcome of the IOP reduction as a medical intervention. A handheld device is under development with which the patient can perform self-measurements outside the clinical environment.

Method

For the measurement principle of the self-tonometer the eye is acoustically excited to oscillate, which is analyzed and attributed to the present IOP. In order to detect the corneal oscillation, an optical sensor is required which meets the demands of a compact, battery driven self-tonometer. A combination of an infrared diode and a phototransistor provides a high-resolution measurement of the corneal oscillation in the range of 10 μm–150 μm, which is compared to a reference sensor in the context of this study. By means of an angular arrangement of the emitter and the detector, the degree of reflected radiation of the cornea can be increased, allowing a measurement with a high signal-to-noise ratio.

Results

By adjusting the angle of incidence between the detector and the emitter, the signal-to-noise ratio was improved by 40 dB which now allows reasonable measurements of the corneal oscillation. For low amplitudes (10 μm) the signal-to-noise ratio is 10% higher than that of the commercial reference sensor. On the basis of amplitude variations at different IOP levels, the estimated standard uncertainty amounts to <0.5 mm Hg in the physiological pressure range with the proposed measuring approach.

Conclusion

With a compact and cost-effective approach, that suits the requirements for a handheld self-tonometer, the corneal oscillation can be detected with high temporal resolution. The cross-sensitivity of the sensor concept concerning a distance variation can be reduced by adding a distance sensor. Existing systematic influences of corneal biomechanics will be integrated in the sensor concept as a consecutive step.

Corneal deformation amplitude analysis for keratoconus detection through compensation for intraocular pressure and integration with horizontal thickness profile

BACKGROUND

The Corvis ST provides measurements of intraocular pressure (IOP) and a biomechanically-corrected IOP (bIOP). IOP influences corneal deflection amplitude (DA), which may affect the diagnosis of keratoconus. Compensating for IOP in DA values may improve the detection of keratoconus.

METHODS

195 healthy eyes and 136 eyes with keratoconus were included for developing different approaches to distinguish normal and keratoconic corneas using attribute selection and discriminant function. The IOP compensation is proposed by dividing the DA by the IOP values. The first approaches include DA compensated for either IOP or bIOP and other parameters from the deformation corneal response (DCR). Another approach integrated the horizontal corneal thickness profile (HCTP). The best classifiers developed were applied in a validation database of 156 healthy eyes and 87 eyes with keratoconus. Results were compared with the current Corvis Biomechanical Index (CBI).

RESULTS

The best biomechanical approach used the DA values compensated by IOP (Approach 2) using a linear discriminant function and reached AUC 0.954, with a sensitivity of 88.2% and a specificity of 97.4%. When thickness horizontal profile data was integrated (Approach 4), the best function was the diagquadratic, resulting in an AUC of 0.960, with a sensitivity of 89.7% and a specificity of 96.4%. There was no significant difference in the results between approaches 2 and 4 with the CBI in the training and validation databases.

CONCLUSIONS

By compensating for the IOP, and with the horizontal thickness profile included or excluded, it was possible to generate a classifier based only on biomechanical information with a similar result to the CBI.

Non-contact tonometry using Corvis ST: analysis of corneal vibrations and their relation with intraocular pressure

The aim of this study was to determine characteristic frequencies of corneal vibrations occurring during air-puff intraocular pressure (IOP) measurement using the Corvis ST tonometer. Relations of frequency of the corneal vibrations with IOP were examined. Two selected vibration frequencies-frequency with maximum amplitude, and mass center of the frequency distribution area, for which the amplitude was higher than 50% (CM50)-present significant correlations with non-corrected IOP and biomechanical corrected IOP (bIOP). The highest correlation was found between the mean values of CM50 and bIOP (r=0.91). Based on the results obtained, it can be stated that the vibration frequencies of corneal peaks are closely related to the measured non-corrected and biomechanical corrected IOPs.

Novel dynamic corneal response parameters in a practice use: a critical review

Background

Non-contact tonometers based on the method using air puff and Scheimpflug’s fast camera are one of the latest devices allowing the measurement of intraocular pressure and additional biomechanical parameters of the cornea. Biomechanical features significantly affect changes in intraocular pressure values, as well as their changes, may indicate the possibility of corneal ectasia. This work presents the latest and already known biomechanical parameters available in the new offered software. The authors focused on their practical application and the diagnostic credibility indicated in the literature.

Discussion

An overview of available literature indicates the importance of new dynamic corneal parameters. The latest parameters developed on the basis of biomechanics analysis of corneal deformation process, available in non-contact tonometers using Scheimpflug’s fast camera, are used in the evaluation of laser refractive surgery procedures, e.g. LASIK procedure. In addition, the assessment of changes in biomechanically corrected intraocular pressure confirms its independence from changes in the corneal biomechanics which may allow an intraocular pressure real assessment. The newly developed Corvis Biomechanical Index combined with the corneal tomography and topography assessment is an important aid in the classification of patients with keratoconus.

Conclusion

New parameters characterising corneal deformation, including Corvis Biomechanical Index and biomechanical compensated intraocular pressure, significantly extend the diagnostic capabilities of this device and may be helpful in assessing corneal diseases of the eye. Nevertheless, further research is needed to confirm their diagnostic pertinence.

In vivo measurement of regional corneal tangent modulus

Currently available clinical devices are unable to measure corneal biomechanics other than at the central region. Corneal stiffness (S), thickness, and radius of curvature was measured at the central cornea (primary fixation) and 3 mm from the temporal limbus (primary and nasal fixations). The corneal tangent modulus (E) of 25 healthy subjects was calculated from these data. After confirming normality, repeated measures analysis of variance (RMANOVA) revealed significant difference in S (F(2, 48) = 21.36, p < 0.001) at different corneal regions and direction of fixations. E also varied significantly at different corneal regions and direction of fixations (RMANOVA: F(2, 48) = 23.06, p < 0.001). A higher S and a lower E were observed at the temporal region compared with the corneal centre. Nasal fixation further increased S and E values compared with primary fixation. Due to the specific arrangement of corneal collagen fibrils, heterogeneity of corneal biomechanical properties is expected. In future clinical practice, localized corneal biomechanical alternation and measurement might assist corneal disease detection and post-surgery management. In addition, practitioners should be aware of the fixation effect on corneal biomechanical measurement.

Corneal Biomechanics in Ectatic Diseases: Refractive Surgery Implications

BACKGROUND

Ectasia development occurs due to a chronic corneal biomechanical decompensation or weakness, resulting in stromal thinning and corneal protrusion. This leads to corneal steepening, increase in astigmatism, and irregularity. In corneal refractive surgery, the detection of mild forms of ectasia pre-operatively is essential to avoid post-operative progressive ectasia, which also depends on the impact of the procedure on the cornea.

METHOD

The advent of 3D tomography is proven as a significant advancement to further characterize corneal shape beyond front surface topography, which is still relevant. While screening tests for ectasia had been limited to corneal shape (geometry) assessment, clinical biomechanical assessment has been possible since the introduction of the Ocular Response Analyzer (Reichert Ophthalmic Instruments, Buffalo, USA) in 2005 and the Corvis ST (Oculus Optikgeräte GmbH, Wetzlar, Germany) in 2010. Direct clinical biomechanical evaluation is recognized as paramount, especially in detection of mild ectatic cases and characterization of the susceptibility for ectasia progression for any cornea.

CONCLUSIONS

The purpose of this review is to describe the current state of clinical evaluation of corneal biomechanics, focusing on the most recent advances of commercially available instruments and also on future developments, such as Brillouin microscopy.

Quantitative Assessment of the Impact of Blood Pulsation on Intraocular Pressure Measurement Results in Healthy Subjects

Background. Blood pulsation affects the results obtained using various medical devices in many different ways. Method. The paper proves the effect of blood pulsation on intraocular pressure measurements. Six measurements for each of the 10 healthy subjects were performed in various phases of blood pulsation. A total of 8400 corneal deformation images were recorded. The results of intraocular pressure measurements were related to the results of heartbeat phases measured with a pulse oximeter placed on the index finger of the subject's left hand. Results. The correlation between the heartbeat phase measured with a pulse oximeter and intraocular pressure is 0.69 ± 0.26 (p < 0.05). The phase shift calculated for the maximum correlation is equal to 60 ± 40° (p < 0.05). When the moment of measuring intraocular pressure with an air-puff tonometer is not synchronized, the changes in IOP for the analysed group of subjects can vary in the range of ±2.31 mmHg (p < 0.3). Conclusions. Blood pulsation has a statistically significant effect on the results of intraocular pressure measurement. For this reason, in modern ophthalmic devices, the measurement should be synchronized with the heartbeat phases. The paper proposes an additional method for synchronizing the time of pressure measurement with the blood pulsation phase.

Measurement of corneal tangent modulus using ultrasound indentation

Biomechanical properties are potential information for the diagnosis of corneal pathologies. An ultrasound indentation probe consisting of a load cell and a miniature ultrasound transducer as indenter was developed to detect the force-indentation relationship of the cornea. The key idea was to utilize the ultrasound transducer to compress the cornea and to ultrasonically measure the corneal deformation with the eyeball overall displacement compensated. Twelve corneal silicone phantoms were fabricated with different stiffness for the validation of measurement with reference to an extension test. In addition, fifteen fresh porcine eyes were measured by the developed system in vitro. The tangent moduli of the corneal phantoms calculated using the ultrasound indentation data agreed well with the results from the tensile test of the corresponding phantom strips (R(2)=0.96). The mean tangent moduli of the porcine corneas measured by the proposed method were 0.089±0.026MPa at intraocular pressure (IOP) of 15mmHg and 0.220±0.053MPa at IOP of 30mmHg, respectively. The coefficient of variation (CV) and intraclass correlation coefficient (ICC) of tangent modulus were 14.4% and 0.765 at 15mmHg, and 8.6% and 0.870 at 30mmHg, respectively. The preliminary study showed that ultrasound indentation could be applied to the measurement of corneal tangent modulus with good repeatability and improved measurement accuracy compared to conventional surface displacement-based measurement method. The ultrasound indentation can be a potential tool for the corneal biomechanical properties measurement in vivo.

Characterization of Ocular Biomechanics in Pellucid Marginal Degeneration

PURPOSE

This study sought to investigate the diagnostic capacity of corneal biomechanical response parameters in a group of patients with pellucid marginal degeneration (PMD) using the Ocular Response Analyzer (ORA) and Corvis ST devices.

METHODS

In this prospective clinical study, we used the Corvis ST and ORA devices to investigate the ocular biomechanics of patients with PMD. Eighty-one eyes were included, and 2 study groups were formed: the PMD group (the study group, n = 29) and the control group (n = 52). We focused on 13 biomechanical parameters. Statistical analysis was performed using SPSS. Biomechanical parameters for the 2 groups were compared using analysis of covariance.

RESULTS

The ORA results demonstrated that the Keratoconus Match Index was significantly lower in the PMD group than in the control group (0.031 ± 0.37 vs. 0.79 ± 0.33; P = 0.001). The 2 groups did not significantly differ with respect to intraocular pressure- and central corneal thickness-adjusted values for corneal hysteresis or corneal resistance factor. Regarding the Corvis parameters, differences between the control and PMD groups were detected for CorWmax amp (control 1.01 ± 0.01, PMD 1.06 ± 0.01; P = 0.020) and CorA2 t (control 21.78 ± 0.03, PMD 21.66 ± 0.04; P = 0.0003).

CONCLUSIONS

We identified 2 Corvis parameters that could be used to characterize PMD and differentiate PMD corneas from normal corneas. These parameters support the hypothesis that there is significantly less deformation of the central cornea in PMD corneas than in healthy corneas. However, because useful "first-line" diagnostic devices for diagnosing PMD (such as Pentacam and the ORA) exist, the Corvis ST serves as an additional diagnostic tool that can also be used for long-term monitoring after diagnosis confirmation.