External Validation of Multiple Predictive Models in AIS Patients Undergoing Intravenous Thrombolysis
Bibliographic record
Abstract
Abstract Background and Purpose-ASPECTS (Alberta Stroke Program Early CT Score), ASTRAL (Acute Stroke Registry and Analysis of LausanneL), DRAGON (including intensive middle cerebral artery sign, pre-stroke modified Rankin Scale score, age, glucose, onset to treatment, NIH Stroke Scale score), THRIVE-c (Total Health Risks in Vascular Events- calculation score) and START (NIHSS Stroke Scale score, Age, pre-stroke mRS score, onset-to-treatment Time) are predictive models that have been gradually developed in recent years to predict functional outcome after acute stroke in patients treated with intravenous thrombolysis, respectively. We aimed to externally validate these scores to assess their predictive performance in this advanced stroke center in China. Methods- We examined the clinical data of 835 patients with AIS who were admitted to the emergency department for intravenous thrombolysis at the Advanced Stroke Center, First Central Hospital, Baoding, China, between January 2016 and May 2022, and scored the patients using the ASPECTS, ASTRAL, DRAGON, THRIVE-c, and START scales. The 3-month modified Rankin Scale scores were observed for each score point, and patients with scores 3 to 6 were defined as having a poor prognosis and compared with the proportions predicted based on risk scores. The ROC curve was used to analyze the predictive value of each score for poor prognosis at 3 months. The total area under the ROC curve showed that it was the C value, and the C value was compared with the predictive value of the five scores; The Hosmer-Lemeshow (H-L) goodness-of-fit [χ2 (P)] test was applied to evaluate the fit of each model to the actual results; two indicators, the calibration curve and the Brier score, were used to evaluate the calibration of the models. Multivariate logistic regression coefficients for the variables in the five scores were also compared with the original derivation cohort. Results-Finally, 728 patients were included, and 318 (43.68%) had a poor prognosis. roc curve analysis, ASPECTS, ASTRAL, DRAGON, THRIVE-c, and START scores corresponded to C values of 0.851, 0.825, 0.854, 0.809, and 0819 in the overall patients, respectively, and in the pre-circulation 0.853, 0.813, 0.833, 0.804, 0.807, and 0.848, 0.862, 0.909, 0.811, 0.857 in the posterior cycle, respectively (all P > 0.05).Hosmer-Lemeshow goodness-of-fit tests for ASPECTS, ASTRAL, DRAGON, THRIVE-c, and START scores with P values of P < 0.001, 0.000365, 0.8245, P < 0.001, P < 0.001, and P < 0.001, respectively, in the pre-loop, P < 0.001, 0.005187, 0.4182, P < 0.001, P < 0.001, and P < 0.001, respectively, in the post-loop, P < 0.0008213, 0.3502, and 0.7645, P < 0.001, P < 0.001. Brier scores, 0.2406, 0.0264, 0.1691, 0.2938, 0.2266 for ASPECTS, ASTRAL, THRIVE-c, DRAGON, START models, respectively. Conclusions-All five score prediction models, ASPECTS, ASTRAL, DRAGON, THRIVE-c, and START, predicted the 3-month adverse prognostic risk in AIS patients undergoing intravenous thrombolysis in both anterior circulation and posterior circulation lesions, but the DRAGON score had the highest predictive diagnostic value in the posterior circulation. the DRAGON score had the highest predictive models predicted prognosis in good agreement with the actual probabilities, and the calibration of the remaining four prediction models was less than optimal.
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 | 0.027 | 0.051 |
| Meta-epidemiology (narrow) | 0.002 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.005 | 0.003 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.001 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.001 |
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".