Incremental contribution of Corvis ST dynamic biomechanical parameters to interpretable machine learning prediction of clinician-selected refractive procedures: a retrospective observational study
Aug 2026· BMC Ophthalmology· Vol 26· 0 citations· 34 references
Medicine
TL;DR
Structured preoperative variables can partially reproduce single-center clinician-selected refractive procedure patterns but do not establish optimal surgical recommendation or external generalizability.
Abstract
This study aimed to evaluate whether structured preoperative variables can reproduce clinician-selected refractive procedure patterns and to assess the incremental contribution of Corvis ST dynamic biomechanical parameters beyond conventional refractive, tomographic, pachymetric, and risk-related features. We conducted a retrospective observational study of 395 patients (763 eyes) who underwent refractive surgery at Chongqing Aier Eye Hospital between October 2023 and November 2024. The outcome label was the procedure actually selected and performed by clinicians, including ICL, SMILE, LASIK, and SURFACE. Forty-eight structured preoperative features were analyzed, including demographic, refractive, visual, ocular-surface, tomographic, pachymetric, anterior-segment, Pentacam-derived deviation/risk, and Corvis ST-derived variables. Data were split at the patient level into a training cohort and an internal held-out test cohort. Multiple machine learning models were developed using patient-grouped cross-validation, and the final model was evaluated using accuracy, balanced accuracy, macro-F1 score, class-wise metrics, calibration analysis, patient-level clustered bootstrap resampling, and one-eye-per-patient sensitivity analysis. Structured feature-set ablation and SHAP analysis were performed to evaluate feature-domain contributions and model behavior. The two-stage XGBoost-based model, which first separated ICL from corneal laser procedures and then classified laser procedures into SMILE, SURFACE, and LASIK, achieved the most favorable overall performance. Its estimated total validation accuracy was 83.36% ± 2.59%. On the internal held-out test cohort, the model achieved an end-to-end accuracy of 78.43%, balanced accuracy of 79.17%, macro-F1 score of 77.96%, and macro AUC of 93.77%. Ablation analysis showed that most predictive gain was derived from tomographic and pachymetric variables, whereas pure Corvis ST dynamic biomechanical parameters provided modest complementary information beyond the full non-pure-Corvis feature set. SHAP analysis indicated that refractive parameters were the dominant contributors, while tomographic, pachymetric, risk-related, and Corvis ST-derived variables contributed to procedure-specific model behavior. Structured preoperative variables can partially reproduce single-center clinician-selected refractive procedure patterns. Corvis ST-derived dynamic biomechanical parameters suggested modest complementary predictive information beyond conventional refractive, tomographic, pachymetric, and risk-related features. These findings support internally validated prediction of clinician-selected procedure patterns but do not establish optimal surgical recommendation or external generalizability.
ICLGuru demonstrated useful postoperative vault predictions with performance comparable to existing machine learning and UBM-based nomograms, though it appeared to overestimate longer term vault in the authors' cohort, and may be valuable in identifying eyes at risk for excessive vault.
Sanjana Molleti, Ethan J. Lindberg, Hanna Pawlowski et al.· Clinical Ophthalmology· 0 citations
OBJECTIVE
Normal-tension glaucoma (NTG) is characterized by progressive optic nerve damage despite intraocular pressure remaining consistently within the normal range. Predicting disease progression in patients with confirmed NTG remains challenging. This study aimed to develop and validate an interpretable machine learning model integrating genetic risk scores and corneal biomechanical parameters to predict progression risk in patients with NTG, identify independent predictors, and quantify the contribution of individual features to support precision clinical management.
METHODS
A total of 342 patients with NTG were consecutively enrolled at a tertiary hospital and randomly allocated to training set (n = 238) and validation set (n = 104) at a ratio of 7:3. Baseline characteristics and six core indicators were collected. Candidate predictors were selected through univariate analysis and least absolute shrinkage and selection operator (LASSO) regression with 10-fold cross-validation and the λ-1se criterion. Independent predictors were subsequently identified using multivariable logistic regression. Three machine learning models-random forest (RF), support vector machine (SVM), and logistic regression (LR)-were developed. Model performance was evaluated using the area under the receiver operating characteristic curve (AUC), calibration curves, and decision curve analysis (DCA). SHapley Additive exPlanations (SHAP) analysis was performed to interpret feature contributions.
RESULTS
Univariate analysis revealed significant differences in all six indicators between the progression and non-progression groups (p < .05). Multivariable logistic regression further confirmed that all six indicators were independently associated with NTG progression (p < .05). The RF model demonstrated the best predictive performance, with an AUC of 0.760 (95% confidence interval (CI) : 0.678-0.842) in the training set and 0.747 (95% CI: 0.623-0.871) in the validation set. It outperformed both the SVM model (training AUC = 0.704; validation AUC = 0.694) and the LR model (training AUC = 0.742; validation AUC = 0.729). SHAP analysis ranked the features, in descending order of contribution, as mean retinal nerve fibre layer (RNFL) thickness, first applanation velocity, visual field mean deviation, relative tear GNAI1 level, polygenic risk score for NTG, and relative tear PRDX4 level. The calibration curves showed good agreement between predicted and observed probabilities, while DCA demonstrated a high clinical net benefit across a broad range of threshold probabilities.
CONCLUSION
A model integrating corneal biomechanical parameters, genetic risk scores, and tear molecular biomarkers was developed to predict the risk of NTG progression and demonstrated potential clinical utility. This model may provide a quantitative reference for risk stratification and personalised management in patients with NTG.
This exploratory pilot study introduces a machine learning framework designed as a hypothesis-generating tool to handle extremely imbalanced ophthalmic data and identify potential preoperative biometric features associated with toric IOL micro-rotation, effectively identifying minority risk features and laying the groundwork for future AI-driven surgical navigation systems.
Kuo-Chi Hung, Pi-Jung Lin, T. Ho et al.· Scientific Reports· 0 citations
Objective: To assess and compare corneal tomographic and biomechanical characteristics between siblings ofpatients with keratoconus and healthy siblings of individuals without ocular pathology, aiming to detect earlysubclinical changes indicative of increased KC risk.Study Design: Prospective, observational, cross-sectional study.Place and Duration of Study: This study was conducted at the Armed Forces Institute of Ophthalmology (AFIO), Pak Emirates Military Hospital (PEMH), Rawalpindi, Pakistan, from 21st August 2024 to 21st February 2025.Methods: A total of 28 participants (14 in each group) were recruited. Group A included biologically related siblings of Keratoconus patients; Group B consisted of healthy siblings of individuals without ocular disease. Corneal tomography parameters (K1, K2, Kmax, Front Elevation Thinnest Pachymetry, Back Elevation Thinnest Pachymetry, and BAD-D) were measured using Pentacam. Biomechanical parameters, including CBI and TBI values, were assessed via Corvis® ST. Best Corrected Visual Acuity (BCVA) was recorded in logMAR. Data were analyzed using SPSS v25. t-tests and Mann-Whitney U tests were used based on normality, with P < 0.05 considered statistically significant. Effect sizes were calculated.Results: K2 and Kmax were significantly higher in KC siblings (P < 0.05). D value, Front Elevation Thinnest Pachymetry, and Back Elevation Thinnest Pachymetry also showed significant differences (P < 0.001). CBI and TBI values were significantly higher in the KC sibling group compared to normal siblings (P < 0.001). No significant differences were observed for K1 or logMAR BCVA.Conclusion: Siblings of KC patients demonstrate early tomographic and biomechanical corneal changes, despite being clinically asymptomatic. These findings support the value of advanced imaging for early detection and risk stratification in high-risk populations.
How to cite this: Machlovi SI, Fawad A, Shah AA, Sohail A, Shabbir A, Mehboob N. Corneal Tomography and BiomechanicalCharacteristics in Siblings of Patients with Keratoconus: A Cross-Sectional Study. Life and Science. 2026; 7(3): 278-284. doi:http://doi.org/10.37185/LnS.1.1.1050
Saira Ishaque Machlovi, Aisha Fawad, Aizaz Ali Shah et al.· Life Science· 0 citations
The relationship between preoperative biometrics and postoperative equivalent power appears predominantly linear, Though larger datasets may enhance machine-learning performance, though larger datasets may enhance machine-learning performance.
Amanda Pan, K. P. Kaiser, Stefan Raidl et al.· Current Eye Research· 0 citations
Combined intrastromal corneal ring segment implantation with corneal collagen cross-linking (CXL) aims to improve optical regularity and reduce the risk of ectatic progression in selected eyes with keratoconus. In this retrospective, single-center, complete-case longitudinal series, 58 eyes from 40 patients treated at OCULENS Ophthalmology Clinic between 2019 and 2022 were analyzed after same-session KeraRing implantation and conventional epithelium-off CXL, with baseline and 6-, 12-, 24-, and 36-month follow-up data. Outcomes included uncorrected distance visual acuity (UDVA), corrected distance visual acuity (CDVA), thinnest pachymetry, maximum keratometry (Kmax), the Belin/Ambrosio enhanced ectasia total deviation index (BAD-D), the C and D components of the ABCD classification, and Corvis ST-derived dynamic deformation parameters. Longitudinal continuous outcomes were analyzed as eye-level repeated-measures estimates with Holm-adjusted paired comparisons, and stage distributions were analyzed with the Friedman test. At 36 months, UDVA improved from 0.560 ± 0.151 to 0.469 ± 0.136 logMAR and CDVA from 0.350 ± 0.109 to 0.287 ± 0.092 logMAR (both p < 0.001). Kmax decreased from 56.11 ± 3.17 D to 54.87 ± 2.86 D, BAD-D improved from 5.62 ± 1.89 to 4.70 ± 1.75, and thinnest pachymetry measured 440.7 ± 21.9 µm at 36 months, corresponding to 97.0% of baseline thickness. Corvis ST-derived deformation behavior shifted toward lower deformation amplitude and higher stiffness parameter at first applanation (SP-A1), but these device-derived parameters should be interpreted as indirect deformation response metrics rather than as direct tissue stiffness measurements. Two clinically significant complications were recorded (2/58 eyes, 3.4%): one sterile corneal infiltrate requiring ring explantation and one case of corneal melting requiring ring explantation and referral for keratoplasty. The findings support sustained 36-month visual, tomographic, and Corvis ST-derived changes after combined treatment, while interpretation remains limited by the retrospective complete-case design, unavailable denominator for cases excluded because of incomplete follow-up, eye-level analysis with bilateral cases, lack of a control group, and indirect nature of Corvis ST-derived biomechanical assessment.
R. Pop, P. Nicula, C. Nicula et al.· Vision· 0 citations
