Abstract T P26: Combining Clinical and Imaging Data to Develop a Highly Predictive Model of Outcomes in the MR RESCUE Trial
Bibliographic record
Abstract
Background: Identifying patient characteristics that predict outcomes in acute ischemic stroke may assist in triaging those who are candidates for endovascular therapies. We sought to identify predictors of outcome in the overall Mechanical Retrieval and Recanalization of Stroke Clots Using Embolectomy (MR RESCUE) cohort and compare results to the previously validated Totaled Health Risks in Vascular Events (THRIVE) score. Methods: MR RESCUE randomized 118 acute ischemic stroke patients with multimodal imaging to embolectomy or standard care within 8 hours of onset. For this analysis, we investigated 17 baseline variables (e.g. age, predicted core volume, time to enrollment) and 8 intermediate variables (e.g. hemorrhagic transformation, day 7 recanalization, final infarct volume) with the potential to impact outcomes (day 90 mRS). The baseline variables were analyzed employing bivariate and multivariate methods (random forest and logistic regression). Two models were developed, one including only significant baseline variables, and the second also incorporating significant intermediate variables. Results: A multivariate model (Table) employing only baseline covariates achieved an overall accuracy (C statistic) of 85% in predicting poor outcome (day 90 mRS 3-6) compared to 80.5% for the THRIVE score. A second model (Table) adding significant intermediate variables achieved 89% accuracy in predicting day 90 mRS. Conclusions: In the MR RESCUE trial, advanced imaging variables, including predicted core volume and site of vessel occlusion, contributed to a highly accurate multivariable model of outcome. In the development phase, this model achieved higher accuracy than the THRIVE score. Future studies are needed to validate this model in an independent cohort.
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 | 0.025 | 0.025 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.002 | 0.003 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.003 | 0.001 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.002 | 0.003 |
| Insufficient payload (model declined to judge) | 0.005 | 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".