SUMMARY: Artificial intelligence technology is a rapidly expanding field with many applications in acute stroke imaging, including ischemic and hemorrhage subtypes. Early identification of acute stroke is critical for initiating prompt intervention to reduce morbidity and mortality. Artificial intelligence can help with various aspects of the stroke treatment paradigm, including infarct or hemorrhage detection, segmentation, classification, large vessel occlusion detection, Alberta Stroke Program Early CT Score grading, and prognostication. In particular, emerging artificial intelligence techniques such as convolutional neural networks show promise in performing these imaging-based tasks efficiently and accurately. The purpose of this review is twofold: first, to describe AI methods and available public and commercial platforms in stroke imaging, and second, to summarize the literature of current artificial intelligence–driven applications for acute stroke triage, surveillance, and prediction.

BACKGROUND AND PURPOSE: Convolutional neural networks are a powerful technology for image recognition. This study evaluates a convolutional neural network optimized for the detection and quantification of intraparenchymal, epidural/subdural, and subarachnoid hemorrhages on noncontrast CT. MATERIALS AND METHODS: This study was performed in 2 phases. First, a training cohort of all NCCTs acquired at a single institution between January 1, 2017, and July 31, 2017, was used to develop and cross-validate a custom hybrid 3D/2D mask ROI-based convolutional neural network architecture for hemorrhage evaluation. Second, the trained network was applied prospectively to all NCCTs ordered from the emergency department between February 1, 2018, and February 28, 2018, in an automated inference pipeline. Hemorrhage-detection accuracy, area under the curve, sensitivity, specificity, positive predictive value, and negative predictive value were assessed for full and balanced datasets and were further stratified by hemorrhage type and size. Quantification was assessed by the Dice score coefficient and the Pearson correlation. RESULTS: A 10,159-examination training cohort (512,598 images; 901/8.1% hemorrhages) and an 862-examination test cohort (23,668 images; 82/12% hemorrhages) were used in this study. Accuracy, area under the curve, sensitivity, specificity, positive predictive value, and negative-predictive value for hemorrhage detection were 0.975, 0.983, 0.971, 0.975, 0.793, and 0.997 on training cohort cross-validation and 0.970, 0.981, 0.951, 0.973, 0.829, and 0.993 for the prospective test set. Dice scores for intraparenchymal hemorrhage, epidural/subdural hemorrhage, and SAH were 0.931, 0.863, and 0.772, respectively. CONCLUSIONS: A customized deep learning tool is accurate in the detection and quantification of hemorrhage on NCCT. Demonstrated high performance on prospective NCCTs ordered from the emergency department suggests the clinical viability of the proposed deep learning tool.

Intracerebral haemorrhage (ICH) is the most devastating and disabling type of stroke. Uncontrolled hypertension (HTN) is the most common cause of spontaneous ICH. Recent advances in neuroimaging, organised stroke care, dedicated Neuro-ICUs, medical and surgical management have improved the management of ICH. Early airway protection, control of malignant HTN, urgent reversal of coagulopathy and surgical intervention may increase the chance of survival for patients with severe ICH. Intensive lowering of systolic blood pressure to 30 cbic centmetre) or symptomatic perihaematoma oedema, it may be beneficial to keep serum sodium level at 140–150 mEq/L for 7–10 days to minimise oedema expansion and mass effect. Mannitol and HTS can be used emergently for worsening cerebral oedema, elevated intracranial pressure (ICP) or pending herniation. HTS should be administered via central line as continuous infusion (3%) or bolus (23.4%). Ventriculostomy is indicated for patients with severe intraventricular haemorrhage, hydrocephalus or elevated ICP. Patients with large cerebellar or temporal ICH may benefit from emergent haematoma evacuation. It is important to start intermittent pneumatic compression devices at the time of admission and subcutaneous unfractionated heparin in stable patients within 48 hours of admission for prophylaxis of venous thromboembolism. There is no benefit for seizure prophylaxis or aggressive management of fever or hyperglycaemia. Early aggressive comprehensive care may improve survival and functional recovery.

Intracranial atherosclerotic stenosis (ICAS) is one of the most common causes of stroke worldwide, accounting for 8% to 10% of strokes in North America and 30% to 50% of strokes in Asia.^1-5 The current recommended strategy for stroke prevention in patients with ICAS is standard medical therapy.⁶ In recent years, standard medical therapy has substantially reduced stroke recurrence in patients with ICAS.^4,7,8 However, a subset of patients are at a high risk for stroke recurrence despite medical treatment.^9,10 For instance, patients with border zone infarcts and impaired collaterals in the Stenting and Aggressive Medical Management for Preventing Recurrent Stroke in Intracranial Stenosis (SAMMPRIS) trial had a 37% 1-year stroke recurrence rate.¹¹

Background Prior studies evaluating the Wingspan stent for treatment of symptomatic intracranial atherosclerotic disease have included patients with a spectrum of both on-label and off-label indications for the stent. The WEAVE trial assessed 152 patients stented with the Wingspan stent strictly by its current on-label indication and found a 2.6% periprocedural stroke and death rate. Objective This WOVEN study assesses the 1-year follow-up from this cohort. Methods Twelve of the original 24 sites enrolling patients in the WEAVE trial performed follow-up chart review and imaging analysis up to 1 year after stenting. Assessment of delayed stroke and death was made in 129 patients, as well as vascular imaging follow-up to assess for in-stent re-stenosis. Results In the 1-year follow-up period, seven patients had a stroke (six minor, one major). Subsequent to the periprocedural period, no deaths were recorded in the cohort. Including the four patients who had periprocedural events in the WEAVE study, there were 11 strokes or deaths of the 129 patients (8.5%) at the 1-year follow-up. Conclusions The WOVEN study provides the 1-year follow-up on a cohort of 129 patients who were stented according to the current on-label use. It provides a more homogeneous patient group for analysis than prior studies, and demonstrates a relatively low 8.5% 1-year stroke and death rate in stented patients.

Antiplatelet therapy is one of the mainstays for secondary stroke prevention. This narrative review aimed to highlight the current evidence and recommendations of antiplatelet therapy for stroke prevention. We conducted advanced literature search for antiplatelet therapy. Landmark studies and randomised controlled trials evaluating antiplatelet therapy for secondary stroke prevention are reviewed. Results from Cochrane systematic review, pooled data analysis and meta-analysis are discussed. Single-antiplatelet therapy (SAPT) with aspirin, aspirin/extended-release dipyridamole or clopidogrel reduces the risk of recurrent ischaemic stroke in patients with non-cardioembolic ischaemic stroke or transient ischaemic attack (TIA). Dual-antiplatelet therapy (DAPT) with aspirin and clopidogrel or ticagrelor for 21–30 days is more effective than SAPT in patients with minor acute noncardioembolic ischaemic stroke or high-risk TIA. Prolonged use of DAPT is associated with higher risk of haemorrhage without reduction in stroke recurrence than SAPT. Compared with placebo, aspirin reduces the relative risk of recurrent stroke by approximately 22%. Aspirin/dipyridamole and cilostazol are superior to aspirin but associated with significant side effects. Cilostazol or ticagrelor might be more effective than aspirin or clopidogrel in patients with intracranial stenosis. SAPT is indicated for secondary stroke prevention in patients with non-cardioembolic ischaemic stroke or TIA. DAPT with aspirin and clopidogrel or ticagrelor for 21–30 days followed by SAPT is recommended for patients with minor acute noncardioembolic stroke or high-risk TIA. Selection of appropriate antiplatelet therapy should also be based on compliance, drug tolerance or resistance.