Radiomics-Based AI Model to Assist Clinicians in Intracranial Hemorrhage Diagnosis: External Validation Study
Bibliographic record
Abstract
Background: Early identification of the etiology of spontaneous intracerebral hemorrhage (ICH) could significantly contribute to planning a suitable treatment strategy. A notable radiomics-based artificial intelligence (AI) model for classifying causes of spontaneous ICH from brain computed tomography scans has been previously proposed. Objective: This study aimed to externally validate and assess the utility of this AI model. Methods: This study used 69 computed tomography scans from a separate cohort to evaluate the AI model's performance in classifying nontraumatic ICHs into primary, tumorous, and vascular malformation related. We also assessed the accuracy, sensitivity, specificity, and positive predictive value of clinicians, radiologists, and trainees in identifying the ICH causes before and after using the model's assistance. The performances were statistically analyzed by specialty and expertise levels. Results: The AI model achieved an overall accuracy of 0.65 in classifying the 3 causes of ICH. The model's assistance improved overall diagnostic performance, narrowing the gap between nonradiology and radiology groups, as well as between trainees and experts. The accuracy increased from 0.68 to 0.72, from 0.72 to 0.76, from 0.69 to 0.74, and from 0.72 to 0.75 for nonradiologists, radiologists, trainees, and specialists, respectively. With the model's support, radiology professionals demonstrated the highest accuracy, highlighting the model's potential to enhance diagnostic consistency across different levels. Conclusions: When applied to an external dataset, the accuracy of the AI model in categorizing spontaneous ICHs based on radiomics decreased. However, using the model as an assistant substantially improved the performance of all reader groups, including trainees and radiology and nonradiology specialists.
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 | 0.011 | 0.023 |
| Meta-epidemiology (narrow) | 0.002 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.002 | 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".