Implementation of emergent MRI for wake-up stroke: a single-center experience.

PURPOSE: Recent updates in national guidelines for management of acute ischemic stroke in patients of unknown time of symptom onset ("wake-up" strokes) incorporate, for the first time, use of emergent MRI. In this retrospective case series, we analyze our experience at a Comprehensive Stroke Center implementing a new workflow including MRI in this clinical setting. This study also describes "DWI-FLAIR" mismatch, a critical concept for the interpretation of emergent brain MRIs performed for wake-up strokes. METHODS: Over a 14-month period, all brain MRIs for wake-up stroke were identified. The imaging was analyzed by two board-certified, fellowship-trained neuroradiologists, and a diagnosis of DWI-FLAIR mismatch was made by consensus. Process metrics assessed included interval between last known well time and brain imaging, interval between CT and MRI, and interval between brain MRI and interpretation. RESULTS: Sixteen patients with a history of "wake-up stroke" were identified. Thirteen of the 16 patients (81.3%) were found to have a DWI-FLAIR mismatch, suggesting infarct < 4.5 h old. The mean time between last known well and MRI was 7.89 h with mean interval between CT and MRI of 1.83 h. Forty-six percent of patients with DWI-FLAIR mismatch received intravenous thrombolysis. CONCLUSION: In this "real world" assessment of incorporation of emergent MRI for wake-up strokes, there were several key factors to successful implementation of this new workflow, including effective and accurate description of MRI findings; close collaboration amongst stakeholders; 24/7 availability of MRI; and 24/7 onsite coverage by neurology and radiology physicians.

Thin-slice T2 MRI imaging predicts vascular pathology in hemifacial spasm: a case-control study.

Hemifacial spasm (HFS) is a condition that may severely reduce patients' quality of life. We sought to determine the sensitivity and specificity of thin-slice T2 magnetic resonance imaging (MRI) for detecting vascular compression in HFS patients. Prospective information was collected on 28 patients with HFS who presented to our center between March 2011 and March 2012 with thin-slice T2 MR imaging. The sensitivity and specificity for differentiating patients from controls were calculated. Sensitivities were 78.6% and 92.9% for the blinded radiologists and 75% for the partially blinded neurosurgeon. Specificities were 42.9% and 28.6% for the blinded radiologists and 75% for the partially blinded neurosurgeon. Magnetic resonance imaging of the facial nerve can guide clinicians in selecting patients who are good surgical candidates. Thin-slice T2 MRI should be viewed as supportive rather than diagnostic.

A Blinded, Case-Control Trial Assessing the Value of Steady State Free Precession Magnetic Resonance Imaging in the Diagnosis of Trigeminal Neuralgia.

OBJECTIVE: High-resolution magnetic resonance imaging (MRI) may be a useful and readily available adjunct in identifying trigeminal neuralgia secondary to vascular contact (TNVC). This study evaluated the reliability and predictive ability of 1.5-tesla steady state free precession (SSFP) MRI sequences for the diagnosis of symptomatic vascular contact and response to operative intervention in patients with TNVC. METHODS: We performed a blinded, case-matched control trial evaluating SSFP MRI sequences in consecutive patients with unilateral TNVC with operatively proven vascular contact of the trigeminal nerve compared with healthy control subjects matched on age, sex, and laterality of the pathologic neurovascular complex. Interrater reliability was compared between 2 blinded, expert reviewers. Predictive ability of MRI was assessed in regard to accuracy, discrimination, and clinical utility. RESULTS: Inclusion criteria were met by 44 patients (22 consecutive patients with TNVC and 22 matched control subjects). Interrater reliability ranged from fair to excellent for vessel contact (κ = 0.40), location (κ = 0.81), type (κ = 0.72), and multiplicity (κ = 0.31). Vascular contact on MRI sequences did not differ significantly between cases and controls (75% vs. 82%, P = 0.30). MRI demonstrates accurate (Brier 0.15) and good discriminatory ability for clinical response after microvascular decompression (area under the receiver operating characteristic curve 0.81, 95% confidence interval = 0.6-1.0). Decision-curve analysis demonstrated that MRI could result in a net reduction of 5 cases likely to be unsuccessful per 100 patients treated. CONCLUSIONS: These results suggest the utility of SSFP MRI lies not in the diagnosis of TNVC, but rather in stratifying the likelihood of response to microvascular decompression in patients with characteristic symptoms.