Development and Validation of the Relational Tissue Altered (RTA) Index: Applied Artificial Intelligence for the Assessment of Structural Impact from Laser Vision Correction
Bibliographic record
Abstract
A retrospective case–control study was carried out to develop a predictive computational model to objectively and accurately represent the impact of laser vision correction (LVC) on the corneal structure. This study involved data from 3278 eyes (1690 patients) that remained stable after refractive surgery and 105 eyes (66 patients) that developed postoperative ectasia. An artificial intelligence-based machine learning approach was used to create a predictive model for the impact of corneal refractive surgery. The development process was based on the practice of knowledge discovery in databases (KDD) and addressed each step, including data selection, preprocessing, data transformation, data mining, and model evaluation. To evaluate the predictive model output, we analyzed the receiver operating characteristic (ROC) curves to determine the area under the curve (AUC) and the optimal cutoff points, as well as sensitivity and specificity. The minimal pachymetry was superior to central (apex) pachymetry for all calculations. The Relational Tissue Altered (RTA) showed the highest AUC performance, with area under the curve (AUC) values of 0.913. These AUC values were significantly higher (according to the DeLong test) than those obtained with Residual Stromal Bed (RSB) values of 0.832 and 0.825 (apex), and Percent Tissue Altered (PTA) values of 0.805 (minimum (min)) and 0.800 (apex). RTA demonstrated a sensitivity of 76.0% (95% confidence interval (CI) 68.8–83.2%) and a specificity of 89.2% (95% CI 86.2–92.2%). PTA min showed a sensitivity of 66.0% (95% CI 58.0–74.0%) and a specificity of 85.4% (95% CI 82.0–88.8%). The Relational Tissue Altered (RTA) index provides an objective and data-driven measure of the structural impact induced by laser vision correction (LVC) on the cornea. Compared with traditional parameters such as Residual Stromal Bed (RSB) and Percent Tissue Altered (PTA), which were not originally designed to quantify biomechanical disruption, RTA demonstrated superior performance in characterizing surgical impact. Although not developed as a standalone ectasia predictor, RTA offers unique value when integrated with preoperative assessments of intrinsic corneal susceptibility, including topometric, tomographic, and biomechanical metrics. This synergistic approach holds promise for enhancing risk stratification, refining surgical planning, and advancing the safety and personalization of refractive surgery. Laser vision correction (LVC) procedures are elective procedures commonly performed to reduce dependence on glasses or contact lenses. Although generally safe and effective, a rare but serious complication known as corneal ectasia can occur. This happens when the cornea becomes biomechanically unstable, weakened beyond its intrinsic ability to maintain shape, leading to potentially significant visual loss. For risk assessment, surgeons use measurements such as Residual Stromal Bed (RSB) and Percent Tissue Altered (PTA) to estimate structural stress that the surgery places on the cornea. However, these were not specifically designed to quantify the biomechanical impact of the procedure. This study analyzed data from 3278 eyes that were stable after surgery and 105 that developed ectasia. Using artificial intelligence (AI), we created a new Relational Tissue Altered (RTA) index to provide a more accurate, objective measurement of the surgical impact on the cornea. RTA significantly outperformed both RSB and PTA in identifying eyes that underwent high-impact procedures. Although RTA alone does not predict the risk of ectasia, it provides a precise and objective measure of structural impact induced by surgery. Its actual clinical value lies in quantifying the extent the cornea has been altered. The strategic integration of RTA with advanced parameters that characterize corneal geometry and biomechanics, reflecting its intrinsic biomechanical resilience or inherent susceptibility to destabilization, holds promise for generating a more comprehensive and individualized risk profile. This approach provides a more complete and personalized risk profile, empowering surgeons to make informed and safer clinical decisions, enhancing safety and long-term outcomes.
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 distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 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 teacher head, 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".