Prediction of Symptomatic Intracranial Hemorrhage Before Mechanical Thrombectomy Using Machine Learning in Patients with Anterior Circulation Large Vessel Occlusion
Bibliographic record
Abstract
BACKGROUND: Symptomatic intracranial hemorrhage (sICH) after mechanical thrombectomy (MT) is associated with worse outcomes. We sought to develop and internally validate a machine learning (ML) model to predict sICH prior to MT in patients with anterior circulation large vessel occlusion. METHODS: Consecutive adults who underwent MT for internal carotid artery/M1/M2 occlusions at a single institution were reviewed. The data was split into 80% training and 20% hold-out test sets. 9 ML models were screened. The top performing ML model was compared to logistic regression and previously described clinical prediction models. SHapley Additive exPlanations were used to identify the most predictive features in the ML model. RESULTS: A total of 497 patients met inclusion criteria. The top performing ML model was extreme gradient boosting. The area under the receiver operating characteristics curve for the ML model on the test set was 0.79 (95% confidence interval [CI] 0.67-0.89), which was significantly higher (P < 0.001) than the logistic regression model (0.54 [95% CI 0.33-0.76]). The ML model also performed significantly better than the TAG = TICI-ASPECTS-glucose score (0.69 [95% CI 0.55-0.85], P < 0.001), Systolic blood pressure-Time-Blood glucose-ASPECTS score (0.45 [95% CI 0.30-0.60], P < 0.001), and ChatGPT 4.0 (0.60 [95% CI 0.48-0.68], P < 0.001). Based on SHapley Additive exPlanations values the most predictive features of sICH in the ML model were lower Alberta Stroke Program Early CT score, lower collateral score, and higher presenting National Institutes of Health Stroke Scale. CONCLUSIONS: An ML model accurately predicted sICH prior to MT. It performed better than a standard statistical model and previously described clinical prediction models.
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 | 0.003 | 0.010 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.000 | 0.001 |
| Research integrity | 0.000 | 0.001 |
| 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".