Background and Purpose— The availability of and expertise to interpret advanced neuroimaging recommended in the guideline-based endovascular stroke therapy (EST) evaluation are limited. Here, we develop and validate an automated machine learning-based method that evaluates for large vessel occlusion (LVO) and ischemic core volume in patients using a widely available modality, computed tomography angiogram (CTA). Methods— From our prospectively maintained stroke registry and electronic medical record, we identified patients with acute ischemic stroke and stroke mimics with contemporaneous CTA and computed tomography perfusion (CTP) with RAPID (IschemaView) post-processing as a part of the emergent stroke workup. A novel convolutional neural network named DeepSymNet was created and trained to identify LVO as well as infarct core from CTA source images, against CTP-RAPID definitions. Model performance was measured using 10-fold cross validation and receiver-operative curve area under the curve (AUC) statistics. Results— Among the 297 included patients, 224 (75%) had acute ischemic stroke of which 179 (60%) had LVO. Mean CTP-RAPID ischemic core volume was 23±42 mL. LVO locations included internal carotid artery (13%), M1 (44%), and M2 (21%). The DeepSymNet algorithm autonomously learned to identify the intracerebral vasculature on CTA and detected LVO with AUC 0.88. The method was also able to determine infarct core as defined by CTP-RAPID from the CTA source images with AUC 0.88 and 0.90 (ischemic core ≤30 mL and ≤50 mL). These findings were maintained in patients presenting in early (0–6 hours) and late (6–24 hours) time windows (AUCs 0.90 and 0.91, ischemic core ≤50 mL). DeepSymNet probabilities from CTA images corresponded with CTP-RAPID ischemic core volumes as a continuous variable with r =0.7 (Pearson correlation, P <0.001). Conclusions— These results demonstrate that the information needed to perform the neuroimaging evaluation for endovascular therapy with comparable accuracy to advanced imaging modalities may be present in CTA, and the ability of machine learning to automate the analysis.

No abstract

INTRODUCTION: Patients with acute ischemic stroke (AIS) and neurologic deficits are often unable to provide consent and excluded from emergency research participation. Experiences with exception from informed consent (EFIC) to facilitate research on potentially life-saving emergency interventions are limited. Here, we describe our multifaceted approach to EFIC approval for an ongoing randomized clinical trial that compares sedation versus general anesthesia (SEGA) approaches for endovascular thrombectomy during AIS. METHODS: We published a university clinical trial website with EFIC information. We initiated a social media campaign on Facebook within a 50 mile radius of Texas Medical Center. Advertisements were linked to our website, and a press release was issued with information about the trial. In-person community consultations were performed, and voluntary survey information was collected. RESULTS: A total of 193 individuals (65% female, age 46.7 ± 16.6 years) participated in seven focus group community consultations. Of the 144 (75%) that completed surveys, 88.7% agreed that they would be willing to have themselves or family enrolled in this trial under EFIC. Facebook advertisements had 134,481 (52% females; 60% ≥45 years old) views followed by 1,630 clicks to learn more. The website had 1130 views (56% regional and 44% national) with an average of 3.85 min spent. Our Institutional Review Board received zero e-mails requesting additional information or to optout. CONCLUSION: Our social media campaign and community consultation methods provide a significant outreach to potential stroke patients. We hope that our experience will inform and help future efforts for trials seeking EFIC.

No abstract

BACKGROUND: Medically refractory idiopathic intracranial hypertension (IIH) is frequently treated with venous sinus stenosis stenting with high success rates. Patient selection has been driven almost exclusively by identification of supraphysiological venous pressure gradients across stenotic regions based on theoretical assessment of likelihood of response. OBJECTIVE: To explore the possibility of benefit in low venous pressure gradient patients. METHODS: Using a single-center, prospectively maintained registry of patients with IIH undergoing venous stenting, we defined treatment groups by gradient pressures of ≤4, 5 to 8, and >8 mmHg based on the most frequently previously published thresholds for stenting. Baseline demographics, clinical, and neuro-ophthalmological outcomes (including optical coherence tomography and Humphrey visual fields) were compared. RESULTS: Among 53 patients, the mean age was 32 years and 70% female with a mean body mass index was 36 kg/m 2 . Baseline characteristics were similar between groups. The mean change in lumbar puncture opening pressure at 6 months poststenting was similar between the 3 groups (≤4, 5-8, and >8 mmHg; 13.4, 12.9, and 12.4 cmH 2 O, P = .47). Papilledema improvement was observed across groups at 6 months (100, 93, and 86, P = .7) as were all clinical symptoms. The mean changes in optical coherence tomography retinal nerve fiber layer (À30, À54, and À104, P = .5) and mean deviation in Humphrey visual fields (60, 64, and 67, P = .5) at 6 weeks were not significantly different. CONCLUSION: Patients with IH with low venous pressure gradient venous sinus stenosis seem to benefit equally from venous stenting compared with their higher gradient counterparts. Re-evaluation of our restrictive criteria for this potentially vision sparing intervention is warranted. Future prospective confirmatory studies are needed.