Predicting post-stroke functional outcome using explainable machine learning and integrated data
Clinical Snapshot
PICO Framework
| P — Population | 506 patients aged 18-69 years with acute ischemic stroke enrolled at four stroke units in western Sweden |
| I — Intervention | Machine learning models (extreme gradient boosting, multilayer perceptron, L1- and L2-regularized logistic regression) using clinical characteristics and blood biomarkers |
| C — Comparator | Different machine learning algorithms compared against each other |
| O — Outcomes | 3-month functional outcome measured by modified Rankin Scale (unfavorable outcome defined as mRS > 2) |
Bottom Line
This Swedish cohort study developed machine learning models to predict 3-month functional outcomes in 506 acute ischemic stroke patients aged 18-69 years. All models achieved excellent discrimination (AUROC >0.90), with stroke severity (NIHSS) being the dominant predictor. Novel blood biomarkers, particularly brain-derived tau and inflammatory proteins, provided additional predictive value beyond clinical variables. While the results are promising, the study lacks external validation and is limited to younger stroke patients. The clinical utility remains uncertain without comparison to existing prognostic tools or development of a practical prediction rule. Australian clinicians should note that the required biomarkers are not routinely available and would need regulatory approval. Further validation in diverse populations and older patients is essential before clinical implementation.
Key Findings
P Value: Not reported
Effect Size: AUROC 0.900-0.906 across models
Primary Outcome: Unfavorable functional outcome (mRS > 2) at 3 months
Nnt Or Sensitivity: Sensitivity 0.655 ± 0.096, Specificity 0.920 ± 0.035 for MLP model
Confidence Interval: AUPRC 0.773 ± 0.080 for best performing model
Clinical Application
Requires blood biomarker testing (BD-tau, inflammatory proteins) which may not be routinely available Relevant to Australian stroke units; BD-tau testing not currently PBS-funded; would need TGA approval for clinical use Adults aged 18-69 years with acute ischemic stroke, particularly in similar healthcare systems
Abstract
Functional outcome after acute ischemic stroke (AIS) varies widely, and existing prognostic scores may not capture complex relationships. We evaluated a diverse set of clinical characteristics and blood biomarkers with multiple machine learning models to predict 3-month functional outcome after AIS, and used explainable artificial intelligence to identify key drivers of performance. Models were trained on 506 patients aged 18-69 years with AIS enrolled at four stroke units in western Sweden. We compared extreme gradient boosting, multilayer perceptron (MLP), and L1- and L2-regularized logistic regression. Feature importance was assessed with Shapley additive global explanations. Of the 506 patients, 105 had an unfavorable outcome (modified Rankin Scale score > 2). All models showed high area under the curve (AUROC, 0.900-0.906). The MLP achieved the highest precision-recall performance (AUPRC, 0.773 ± 0.080) and sensitivity (0.655 ± 0.096), though with lower specificity (0.920 ± 0.035). Stroke severity (NIH Stroke Scale score) was the dominant predictor across models. Among biomarkers, brain-derived tau (BD-tau) was most informative, followed by inflammation-related plasma proteins. In conclusion, machine learning accurately predicted functional outcome after AIS. BD-tau and inflammation-related proteins contributed predictive information above stroke severity, suggesting a potential for blood biomarkers to enhance individualized prognostication after AIS.
Related Research
Sleep medicine reviews
Machine and deep learning in REM sleep behavior disorder: a scoping review and analysis of reporting quality
3 Aug 2026
Neurological research
Federated deep learning model for epilepsy seizure detection using electroencephalogram (EEG) signal
2 Aug 2026
Journal of psychiatric research
Systematic review of machine learning and deep learning models for EEG-based detection of depression.
2 Aug 2026
This content is for educational purposes for healthcare professionals only and does not constitute clinical advice. Clinical decisions should be based on individual patient assessment, current guidelines, and appropriate specialist consultation. Editorial Standards · Privacy Policy · Terms of Service