Deep learning-assisted identification and quantification of aneurysmal subarachnoid hemorrhage in non-contrast CT scans: Development and external validation of Hybrid 2D/3D UNet.

Accurate stroke assessment and consequent favorable clinical outcomes rely on the early identification and quantification of aneurysmal subarachnoid hemorrhage (aSAH) in non-contrast computed tomography (NCCT) images. However, hemorrhagic lesions can be complex and difficult to distinguish manually. To solve these problems, here we propose a novel Hybrid 2D/3D UNet deep-learning framework for automatic aSAH identification and quantification in NCCT images. We evaluated 1824 consecutive patients admitted with aSAH to four hospitals in China between June 2018 and May 2022. Accuracy and precision, Dice scores and intersection over union (IoU), and interclass correlation coefficients (ICC) were calculated to assess model performance, segmentation performance, and correlations between automatic and manual segmentation, respectively. A total of 1355 patients with aSAH were enrolled: 931, 101, 179, and 144 in four datasets, of whom 326 were scanned with Siemens, 640 with Philips, and 389 with GE Medical Systems scanners. Our proposed deep-learning method accurately identified (accuracies 0.993-0.999) and segmented (Dice scores 0.550-0.897) hemorrhage in both the internal and external datasets, even combinations of hemorrhage subtypes. We further developed a convenient AI-assisted platform based on our algorithm to assist clinical workflows, whose performance was comparable to manual measurements by experienced neurosurgeons (ICCs 0.815-0.957) but with greater efficiency and reduced cost. While this tool has not yet been prospectively tested in clinical practice, our innovative hybrid network algorithm and platform can accurately identify and quantify aSAH, paving the way for fast and cheap NCCT interpretation and a reliable AI-based approach to expedite clinical decision-making for aSAH patients.

Effect of Surgical Clipping versus Endovascular Coiling on the Incidence of Delayed Cerebral Ischemia in Patients with Aneurysmal Subarachnoid Hemorrhage: A Multicenter Observational Cohort Study with Propensity Score Matching.

OBJECTIVE: The effect of surgical clipping (SC) and endovascular coiling (EC) on the incidence of delayed cerebral ischemia (DCI) in patients with aneurysmal subarachnoid hemorrhage (aSAH) has always been a controversial topic. Hence, it is necessary to reanalyze the effects of the 2 surgical methods on DCI, which determines the choice of the most favorable method for patients who are suitable for both surgical modalities. METHODS: A multicenter retrospective observational cohort study was performed to evaluate all consecutive patients with aSAH admitted to 5 medical centers in China between April 2019 and June 2021. Univariable and multivariable analyses were used to confirm risk factors of DCI after aSAH. A 1:1 propensity score matching model was generated in the EC and SC groups to reduce the influence of all confounding factors on DCI. RESULTS: A total of 412 patients were included, and 115 patients (27.9%) developed DCI. After propensity score matching for controlling demographic information, past medical history, admission clinical status, aneurysm characteristics, and inflammatory factors associated with DCI, 133 patients with SC and 133 patients with EC treatment were matched. The results of the matched cohorts indicate a significantly lower incidence of DCI when patients received EC than SC (31.9% vs. 20%; adjusted odds ratio, 1.87; 95% confidence interval, 1.08-3.29; P = 0.027). CONCLUSIONS: The study found that the patients who received SC treatment had a higher incidence of DCI than did those who received EC and suggested that ruptured intracerebral aneurysm is preferentially coiled rather than clipped if the aneurysm is suitable for both surgical modalities.