Comparison of the SeizCT Primer and Optimized Models for Predicting Positive Computed Tomography Findings in Patients With Non-Traumatic Seizures
Bibliographic record
Abstract
Background: Previous studies developed the SeizCT primer and optimized models, both demonstrating similar values for the area under the receiver operating characteristic curve (AuROC). The optimized model incorporates Glasgow Coma Scale (GCS) change from baseline instead of categorized GCS at emergency department (ED) presentation. This study aimed to validate these two models for predicting positive computed tomography (CT) findings in patients with non-traumatic seizures. Methods: A retrospective cross-sectional study was conducted among adult patients (≥ 18 years) with non-traumatic seizures who underwent CT brain imaging in the ED at Lampang Hospital. Data were collected between December 2023 and July 2024 based on parameters from the SeizCT primer and optimized models. External validation compared model performance using AuROC, calibration, decision curve analysis (DCA), and confusion matrices. Results: The validation cohort included 312 patients (210 (67.3%) male; mean age 53 years). Positive CT findings were found in 58 patients (18.6%). The SeizCT primer model had an AuROC of 0.7218 (95% confidence interval (CI): 0.6476, 0.7961), while the SeizCT optimized model achieved 0.7394 (95% CI: 0.6663, 0.8124). An equivalence test showed statistically equivalent discrimination between the two models (P < 0.015), with the optimized model demonstrating better calibration (slope: 0.582 vs. 0.701; P = 0.002; observed-to-expected ratio: 0.934 vs. 0.971; P = 0.030). Conclusions: Both models demonstrated fair to acceptable discrimination after external validation. The SeizCT optimized model is recommended, as it showed superior calibration and its incorporation of GCS change from baseline offers greater clinical applicability.
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 | 0.009 | 0.022 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.002 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 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".