Dataset Information

Using Deep Learning to Automate Goldmann Applanation Tonometry Readings.

ABSTRACT:

Purpose

To develop an objective and automated method for measuring intraocular pressure using deep learning and fixed-force Goldmann applanation tonometry (GAT) techniques.

Design

Prospective cross-sectional study.

Participants

Patients from an academic glaucoma practice.

Methods

Intraocular pressure was estimated by analyzing videos recorded using a standard slit-lamp microscope and fixed-force GAT. Video frames were labeled to identify the outline of the reference tonometer and the applanation mires. A deep learning model was trained to localize and segment the tonometer and mires. Intraocular pressure values were calculated from the deep learning-predicted tonometer and mire diameters using the Imbert-Fick formula. A separate test set was collected prospectively in which standard and automated GAT measurements were collected in random order by 2 independent masked observers to assess the deep learning model as well as interobserver variability.

Main outcome measures

Intraocular pressure measurements between standard and automated methods were compared.

Results

Two hundred sixty-three eyes of 135 patients were included in the training and validation videos. For the test set, 50 eyes from 25 participants were included. Each eye was measured by 2 observers, resulting in 100 videos. Within the test set, the mean difference between automated and standard GAT results was -0.9 mmHg (95% limits of agreement [LoA], -5.4 to 3.6 mmHg). Mean difference between the 2 observers using standard GAT was 0.09 mmHg (LoA,-3.8 to 4.0 mmHg). Mean difference between the 2 observers using automated GAT videos was -0.3 mmHg (LoA, -4.1 to 3.5 mmHg). The coefficients of repeatability for automated and standard GAT were 3.8 and 3.9 mmHg, respectively. The bias for even-numbered measurements was reduced when using automated GAT.

Conclusions

Preliminary measurements using deep learning to automate GAT demonstrate results comparable with those of standard GAT. Automated GAT has the potential to improve on our current GAT measurement standards significantly by reducing bias and improving repeatability. In addition, ocular pulse amplitudes could be observed using this technique.

SUBMITTER: Spaide T

PROVIDER: S-EPMC7606368 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:ImportanceAccurate determination of intraocular pressure (IOP) is crucial for the diagnosis and management of glaucoma. Objective clinical evaluation of the correction equations for Goldmann applanation tonometry (GAT) is lacking.ObjectivesTo investigate the difference between corrected and conventional GAT and Pascal dynamic contour tonometry (DCT) measurements, as well as the correlation between discordant IOP values and stage of glaucoma.Design, setting, and participantsThis prospective cross-sectional case series was conducted at the Department of Ophthalmology, University Hospital Zurich, and Talacker Eye Center between July 1, 2011, and May 31, 2016, among 112 white patients with glaucoma.InterventionsIntraocular pressure measurements were performed with GAT and DCT in a randomized order. Goldmann applanation tonometry measurements were modified with 5 correction equations.Main outcomes and measuresThe primary end point was degree of concordance between corrected or uncorrected GAT and DCT measurements. The secondary end point was association between discordant IOP measurements and the stage of glaucoma, as assessed by the Glaucoma Severity Score.ResultsAmong the 112 patients (67 women and 45 men; mean [SD] age, 66.3 [13.1] years), 63 of the eyes in the study (56.3%) were left eyes and 85 patients (75.9%) were taking ocular antihypertensive medications. Mean (SD) IOP was 20.3 (4.5) mm Hg (95% CI, 19.4-21.1) as measured by DCT and 17.0 [4.1] mm Hg (95% CI, 16.3-17.8) as measured by GAT. The mean (SD) discordance between DCT and GAT measurements was -3.3 (2.0) mm Hg (95% CI, 2.9-3.6). The 5 corrected GAT values ranged from -2.7 to -5.4 mm Hg compared with DCT. The mean (SD) result of the Dresdner correction formula (17.6 [4.1] mm Hg) was closer to the DCT measurement than the original GAT measurement. The mean (SD) Glaucoma Severity Score was 4.7 (3.4) (95% CI, 4.1-5.4). The uncorrected discordance IOPDCT - IOPGAT showed a positive correlation with the Glaucoma Severity Score (rs = 0.33; P < .001) and a negative correlation with central corneal thickness (rs = -0.22; P = .02).Conclusions and relevanceIn comparison with DCT measurements, these data suggest that GAT values are significantly discordant in eyes with thin corneas and advanced glaucoma. Application of GAT-based correction formulas involves a possible risk of creating an even greater number of unpredictable measurement errors. Hence, we advise with caution, especially pertaining to eyes with thin corneas, to not place reliance on GAT readings, and abandon any correction formula.Trial registrationclinicaltrials.gov Identifier: NCT01474070.