Predictors of poor outcome in embolic stroke of undetermined source

Machine learning prediction of motor function in chronic stroke patients: a systematic review and meta-analysis

Background

Recent studies have reported that machine learning (ML), with a relatively strong capacity for processing non-linear data and adaptive ability, could improve the accuracy and efficiency of prediction. The article summarizes the published studies on ML models that predict motor function 3-6 months post-stroke.

Methods

A systematic literature search was conducted in PubMed, Embase, Cochorane and Web of Science as of April 3, 2023 for studies on ML prediction of motor function in stroke patients. The quality of the literature was assessed using the Prediction model Risk Of Bias Assessment Tool (PROBAST). A random-effects model was preferred for meta-analysis using R4.2.0 because of the different variables and parameters.

Results

A total of 44 studies were included in this meta-analysis, involving 72,368 patients and 136 models. Models were categorized into subgroups according to the predicted outcome Modified Rankin Scale cut-off value and whether they were constructed based on radiomics. C-statistics, sensitivity, and specificity were calculated. The random-effects model showed that the C-statistics of all models were 0.81 (95% CI: 0.79; 0.83) in the training set and 0.82 (95% CI: 0.80; 0.85) in the validation set. According to different Modified Rankin Scale cut-off values, C-statistics of ML models predicting Modified Rankin Scale>2(used most widely) in stroke patients were 0.81 (95% CI: 0.78; 0.84) in the training set, and 0.84 (95% CI: 0.81; 0.87) in the validation set. C-statistics of radiomics-based ML models in the training set and validation set were 0.81 (95% CI: 0.78; 0.84) and 0.87 (95% CI: 0.83; 0.90), respectively.

Conclusion

ML can be used as an assessment tool for predicting the motor function in patients with 3-6 months of post-stroke. Additionally, the study found that ML models with radiomics as a predictive variable were also demonstrated to have good predictive capabilities. This systematic review provides valuable guidance for the future optimization of ML prediction systems that predict poor motor outcomes in stroke patients.

Systematic review registration

https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42022335260, identifier: CRD42022335260.

XGBoost-Based Simple Three-Item Model Accurately Predicts Outcomes of Acute Ischemic Stroke

An all-inclusive and accurate prediction of outcomes for patients with acute ischemic stroke (AIS) is crucial for clinical decision-making. This study developed extreme gradient boosting (XGBoost)-based models using three simple factors-age, fasting glucose, and National Institutes of Health Stroke Scale (NIHSS) scores-to predict the three-month functional outcomes after AIS. We retrieved the medical records of 1848 patients diagnosed with AIS and managed at a single medical center between 2016 and 2020. We developed and validated the predictions and ranked the importance of each variable. The XGBoost model achieved notable performance, with an area under the curve of 0.8595. As predicted by the model, the patients with initial NIHSS score > 5, aged over 64 years, and fasting blood glucose > 86 mg/dL were associated with unfavorable prognoses. For patients receiving endovascular therapy, fasting glucose was the most important predictor. The NIHSS score at admission was the most significant predictor for those who received other treatments. Our proposed XGBoost model showed a reliable predictive power of AIS outcomes using readily available and simple predictors and also demonstrated the validity of the model for application in patients receiving different AIS treatments, providing clinical evidence for future optimization of AIS treatment strategies.

Myosteatosis as an independent predictor of short-term mortality in successfully reperfused acute ischemic stroke

BACKGROUND AND PURPOSE

Poor clinical outcomes are still common in successfully reperfused acute ischemic stroke patients. The aim of our study was to assess the impact of sarcopenia and myosteatosis on neurological outcomes and mortality in successfully reperfused acute ischemic stroke patients.

MATERIALS AND METHODS

We included in our retrospective observational study 166 consecutive patients who underwent technically successful mechanical thrombectomy for anterior circulation acute ischemic stroke between Jan 2016 and Dec 2019. ASPECTS and collateral score were assessed on pre-operative CT/CTA. Masseter muscles area and attenuation were measured on CTA images. Clinical and radiological variables were tested in multivariate logistic models to predict the probability of death and, among survivors, of incurring poor outcome.

RESULTS

At admission, mean NIHSS was 19 (SD = 6.5), mean body mass index 25.5 (SD = 4.4) kg/m2, and mean ASPECTS 8.0 (SD = 1.9). Of all, 48.2% patients showed good collaterals, 38.5% intermediate collaterals, and 13.3% poor collaterals. Overall, 90 days mRS was ≤2 in 48.2% of the patients, 3-5 in 30.7%, and 6 in 21.1%. At multivariate logistic regression, age (OR = 1.08, p = 0.036), ASPECTS (OR = 0.59, p = 0.013), and masseter muscles attenuation (OR = 0.93, p = 0.010) were independent predictors of mortality, whereas sex (OR = 7.15, p = 0.043), age (OR = 1.05, p = 0.042), body mass index (OR = 1.35, p = 0.013), NIHSS (OR = 1.12, p = 0.012), and ASPECTS (OR = 0.64, p = 0.024) were independent predictors of poor neurological outcome (mRS 3-5).

CONCLUSION

Beyond other well-known variables, low masseter attenuation, indicating myosteatosis, represents an independent negative prognostic factor for 90 days mortality in patients successfully reperfused after anterior circulation stroke.

Characteristics and outcomes of Embolic Stroke of Undetermined Source according to stroke severity

Background and aims

Patients with embolic strokes of undetermined source (ESUS) usually present with mild symptoms. We aimed to compare the baseline characteristics between mild and severe ESUS, identify predictors for severe ESUS, and assess outcomes of patients with severe ESUS.

Methods

In the AF-ESUS (AF-ESUS) dataset, we stratified ESUS severity using the median National Institutes of Health Stroke Scale (NIHSS) score on admission as cut-off. We performed multivariable stepwise regression analyses to identify independent predictors of severe ESUS and to assess the association between ESUS severity and stroke recurrence, death, and new incident atrial fibrillation (AF) on follow-up. The 10-year cumulative probabilities of outcome incidence were estimated by the Kaplan–Meier product limit method.

Results

In 772 patients (median NIHSS: 6 (interquartile range: 3–12)), 414 (53.6%) patients had severe ESUS (i.e. NIHSS ≥6). Female sex was the only independent predictor for severe ESUS (odds ratio: 1.72 (1.27–2.33)). The rates of recurrence (3.3%/year vs. 3.4%/year, adjusted-hazard ratio: 1.09 (0.73–1.62)) and new incident AF (13.5% vs. 17.0%, adjusted odds ratio: 0.67 (0.44–1.03)) were similar between severe and mild ESUS, but mortality was higher (5.4%/year vs. 3.7%/year, adjusted-hazard ratio: 1.51 (1.05–2.16)) in severe ESUS. The 10-year cumulative probability for stroke recurrence was similar between severe and mild ESUS (38.1% (29.2–48.6) vs. 36.6% (27.8–47.0), log-rank test: 0.01, p = 0.920). The 10-year cumulative probability of death was higher in patients with severe ESUS compared with mild ESUS (40.5% (32.5–50.0) vs. 34.0% (26.0–43.6) respectively; log-rank test: 4.54, p = 0.033).

Conclusions

Women have more severe ESUS compared with men. Patients with severe ESUS have similar rates of stroke recurrence and new incident AF, but higher mortality compared with mild ESUS.