Atrial fibrilation is one of the most frequent arrhythmia in peripheral , specially cerebrovascular embolization, So, AFR is a strong predictore as a cause of hypercoagulable state. High voltage QRS, if general as a sign of left ventricular hyperthrophy maybe secondary to hypertension as a risk factor for CVA, or resticted to precordial leads as a sign of low LVEF or cardiomyopathy.
You might look into an association between P-wave morphology and stroke disease during Atrial fibrillation (AF). Moreover, based on this you could retool your question into e.g., : Are there any prediction system allowing me to use ECG and other biosignals as PPG e.g., to predict the occurence of stroke? Here is an interesting starter
thank you for your response. I might then suggest to maybe first check this article: Deep Neural Networks Can Predict New-Onset Atrial Fibrillation From the 12-Lead ECG and Help Identify Those at Risk of Atrial Fibrillation–Related Stroke; as more than 25% of all strokes are deemed a result of AF, and ≈20% of strokes caused by AF occur in individuals not previously diagnosed with AF. That might be interesting article to read presenting comparison of patients at high risk for new-onset atrial fibrillation and no history of AF predicting future AF and stroke using deep learning model.
other useful article might be: Electrocardiographic abnormalities in acute cerebrovascular events in patients with/without cardiovascular disease.
this is review of 361 patients with results of the most common ECG abnormalities associated with stroke being T-wave abnormalities, prolonged QTc interval and arrhythmias.
You might need to read more and then separate patients with/without cardiovascular disease