Quantitative Assessment of Strabismus Using Cloud AI Computing: Validation Study
Bibliographic record
Abstract
Background Strabismus measurement is essential in vision assessment and screening. It typically requires skilled clinicians or specialized equipment. Photographic strabismus measurement methods have value in terms of accessibility and convenience of use. Objective This study aimed to evaluate Eyeturn Cloud, a cloud-based artificial intelligence (AI) system for measuring strabismus angles based on eye images captured with smartphone cameras under cover test conditions. Methods The Eyeturn Cloud web app uses AI models to recognize eyes, eye lid, and iris, and then to segment iris precisely. It then computes strabismus based on ellipse fitting of the iris boundary and corneal reflection. The system was evaluated in patients (without glasses) with manifest strabismus and control participants. Clinicians measured eye deviations using the prism alternate cover test and also captured pictures of their eyes under alternate cover and unilateral cover conditions. The pictures were processed by Eyeturn Cloud. Results In total, 79 (mean age 11.9, SD 6.3 years; esotropia: n=15, exotropia: n=55, orthotropia: n=9) participants were enrolled; of which, data were available for 71 participants (8/79, 10.1% processing failure). The range of prism alternate cover test strabismus magnitude was from 78 base in to 78 base out prism diopters (PDs). A strong correlation was found between Eyeturn Cloud and clinical measurements (R2=0.95; slope=0.91; P<.001). Bland-Altman analysis revealed that 95% limits of agreement between the 2 measurements were –20.2 to 14.6 PD. A repeatability test with 15 participants (4 photos each) found a 1.53 PD SD. Conclusions The cloud AI web app can compute strabismus angles reliably under alternate and unilateral cover conditions in clinical settings, and its potential for use in telehealth settings needs further evaluation.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.007 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".