Deep learning-assisted diagnosis of large vessel occlusion in acute ischemic stroke based on four-dimensional computed tomography angiography.

Purpose: To develop deep learning models based on four-dimensional computed tomography angiography (4D-CTA) images for automatic detection of large vessel occlusion (LVO) in the anterior circulation that cause acute ischemic stroke. Methods: This retrospective study included 104 LVO patients and 105 non-LVO patients for deep learning models development. Another 30 LVO patients and 31 non-LVO patients formed the time-independent validation set. Four phases of 4D-CTA (arterial phase P1, arterial-venous phase P2, venous phase P3 and late venous phase P4) were arranged and combined and two input methods was used: combined input and superimposed input. Totally 26 models were constructed using a modified HRNet network. Assessment metrics included the areas under the curve (AUC), accuracy, sensitivity, specificity and F1 score. Kappa analysis was performed to assess inter-rater agreement between the best model and radiologists of different seniority. Results: The P1 + P2 model (combined input) had the best diagnostic performance. In the internal validation set, the AUC was 0.975 (95%CI: 0.878-0.999), accuracy was 0.911, sensitivity was 0.889, specificity was 0.944, and the F1 score was 0.909. In the time-independent validation set, the model demonstrated consistently high performance with an AUC of 0.942 (95%CI: 0.851-0.986), accuracy of 0.902, sensitivity of 0.867, specificity of 0.935, and an F1 score of 0.901. The best model showed strong consistency with the diagnostic efficacy of three radiologists of different seniority (k = 0.84, 0.80, 0.70, respectively). Conclusion: The deep learning model, using combined arterial and arterial-venous phase, was highly effective in detecting LVO, alerting radiologists to speed up the diagnosis.

Imaging features of cardioembolic stroke on 4-dimensional computed tomography angiography.

Background: Identifying cardioembolic stroke is important for the decision-making of endovascular treatment and anticoagulation therapy. We aimed to explore the features of cardioembolic stroke on 4-dimensional (4D) computed tomography angiography (4D-CTA) and assess whether these features can assist in classifying stroke etiology. Methods: In this retrospective study, we analyzed the images of 294 patients with acute ischemic stroke (AIS) from July 2020 to February 2022 at the First Affiliated Hospital of Chongqing Medical University, which had been consecutively collected. The data of 110 patients with occlusion of the M1/M2 segment of the middle cerebral artery (MCA) with/without intracranial internal carotid artery (ICA) occlusion were analyzed to calculate the clot burden score (CBS) and collateral score (CS), and the data of 88 patients with a clear origin and distal part were analyzed to measure clot length. Maximum intensity projection (MIP) and time MIP (tMIP) post-processing were used to assess the clot features. The Mann-Whitney U test was used to compare the clot characteristics between the 2 groups. Binary logistic regression was performed to assess the association between the image characteristics and cardioembolic stroke. Moreover, the receiver operating characteristic (ROC) curve was used to test the diagnostic efficacy of MIP/tMIP clot features in classifying cardioembolic stroke. Results: Age, high-risk factors for cerebrovascular disease, high/medium-risk sources of cardioembolic stroke, clot length, CBS, and CS were significantly different between the cardioembolic stroke group and non-cardioembolic stroke group (P<0.05). In the cardioembolic stroke group, the median MIP and tMIP clot length was 12 mm [interquartile range (IQR), 8.3-17.4 mm] and 9.3 mm (IQR, 6.8-14.3 mm), respectively. In the non-cardioembolic stroke group, the median MIP and tMIP clot length was 6.5 mm (IQR, 4.7-11.5 mm) and 5.8 mm (IQR, 3.9-10.6 mm), respectively. Binary logistic regression showed that cardioembolic stroke was significantly associated with MIP-clot length [odds ratio (OR), 1.15; 95% confidence interval (CI): 1.02-1.29; P<0.05], tMIP-clot length (OR, 1.18; 95% CI: 1.02-1.36; P<0.05), and tMIP-CBS (OR, 3.96; 95% CI: 1.08-14.58; P<0.05). The area under the ROC curve (AUC) values of MIP clot length for identifying cardioembolic stroke were 0.75 (95% CI: 0.65-0.84, P<0.05), with a cut-off value of >7.4 mm [sensitivity: 84.62% (95% CI: 69.50-94.10%); specificity: 59.18% (95% CI: 44.20-73.00%)]. The AUC value of tMIP clot length was 0.72 (95% CI: 0.61-0.81, P<0.05), with a cut-off value of >5.4 mm [sensitivity: 92.31% (95% CI: 79.10-98.40%); specificity: 48.98% (95% CI: 34.40-63.70%)]. Conclusions: Clot length and CBS were overestimated on MIP images. Among the clot characteristics, clot length could identify cardioembolic stroke.

Comparison of Three Algorithms for Predicting Infarct Volume in Patients with Acute Ischemic Stroke by CT Perfusion Software: Bayesian, CSVD, and OSVD.

This study aimed to compare the performance of the Bayesian probabilistic method, circular Singular Value Decomposition (cSVD), and oscillation index Singular Value Decomposition (oSVD) algorithms in Olea Sphere for predicting infarct volume in patients with acute ischemic stroke (AIS). Eighty-seven patients suffering from AIS with large vessel occlusion were divided into improvement and progression groups. The improvement group included patients with successful recanalization (TICI 2b-3) after thrombectomy or whose clinical symptoms improved after thrombolysis. The progression group consisted of patients whose clinical symptoms did not improve or even got worse. The infarct core volume from the Olea Sphere software was used as the predicted infarct volume (PIV) in the improvement group, whereas the hypoperfusion volume was used as the PIV in the progression group. We defined predicted difference (PD) as PIV minus final infarct volume (FIV) measured at follow-up imaging. Differences among the three algorithms were assessed by the Friedman test. Spearman correlation analysis was used to verify the correlation between PIV and FIV. In addition, we performed a subgroup analysis of the progression group based on collateral circulation status. The median [interquartile range (IQR)] of the PD and Spearman correlation coefficients (SCCs) between PIV and FIV for the improvement group (n = 22) were: Bayesian = [6.99 (-14.72, 18.99), 0.500]; oSVD = [-12.74 (-41.06, -3.46), 0.423]; cSVD = [-15.38 (-38.92, -4.68), 0.586]. For the progression group (n = 65), the median (IQR) of PD and SCCs were: Bayesian = [1.00 (-34.07, 49.37), 0.748]; oSVD = [-0.17 (-53.42, 29.73), 0.712]; cSVD = [66.55 (7.94, 106.32), 0.674]. The Bayesian algorithm in the Olea Sphere software predicted infarct volumes with better accuracy and stability than the other two algorithms in both the progression and improvement groups.

Comparison of two computed tomography perfusion post-processing software to assess infarct volume in patients with acute ischemic stroke.

Objectives: We used two automated software commonly employed in clinical practice-Olea Sphere (Olea) and Shukun-PerfusionGo (PerfusionGo)-to compare the diagnostic utility and volumetric agreement of computed tomography perfusion (CTP)-predicted final infarct volume (FIV) with true FIV in patients with anterior-circulation acute ischemic stroke (AIS). Methods: In all, 122 patients with anterior-circulation AIS who met the inclusion and exclusion criteria were retrospectively enrolled and divided into two groups: intervention group (n = 52) and conservative group (n = 70), according to recanalization of blood vessels and clinical outcome (NIHSS) after different treatments. Patients in both groups underwent one-stop 4D-CT angiography (CTA)/CTP, and the raw CTP data were processed on a workstation using Olea and PerfusionGo post-processing software, to calculate and obtain the ischemic core (IC) and hypoperfusion (IC plus penumbra) volumes, hypoperfusion in the conservative group and IC in the intervention group were used to define the predicted FIV. The ITK-SNAP software was used to manually outline and measure true FIV on the follow-up non-enhanced CT or MRI-DWI images. Intraclass correlation coefficients (ICC), Bland-Altman, and Kappa analysis were used to compare the differences in IC and penumbra volumes calculated by the Olea and PerfusionGo software to investigate the relationship between their predicted FIV and true FIV. Results: The IC and penumbra difference between Olea and PerfusionGo within the same group (p < 0.001) was statistically significant. Olea obtained larger IC and smaller penumbra than PerfusionGo. Both software partially overestimated the infarct volume, but Olea significantly overestimated it by a larger percentage. ICC analysis showed that Olea performed better than PerfusionGo (intervention-Olea: ICC 0.633, 95%CI 0.439-0.771; intervention-PerfusionGo: ICC 0.526, 95%CI 0.299-0.696; conservative-Olea: ICC 0.623, 95%CI 0.457-0.747; conservative-PerfusionGo: ICC 0.507, 95%CI 0.312-0.662). Olea and PerfusionGo had the same capacity in accurately diagnosing and classifying patients with infarct volume <70 ml. Conclusion: Both software had differences in the evaluation of the IC and penumbra. Olea's predicted FIV was more closely correlated with the true FIV than PerfusionGo's prediction. Accurate assessment of infarction on CTP post-processing software remains challenging. Our results may have important practice implications for the clinical use of perfusion post-processing software.

Significant Slowed Cortical Venous Blood Flow in Patients with Acute Ischemic Stroke with Large Vessel Occlusion Suggests Poor Collateral Circulation and Prognosis.

RATIONALE AND OBJECTIVES: To investigate the change of cortical venous flow in acute ischemic stroke patients with large vessel occlusion (LVO-AIS) and its clinical value. MATERIALS AND METHODS: Baseline whole-brain 4D-CTA/CTP and clinical data of LVO-AIS and a control group were collected from June 2020 to October 2021. Venous inflow time (VIT), venous peak time (VPT), and venous outflow time (VOT) were analyzed on both sides of patients and normal controls. The VIT/VPT/VOT were statistically described and compared between the patient group and normal controls, then, in patients with different collateral circulation and prognoses. Next, the correlation between cortical venous drainage time and collateral circulation grading was analyzed. Finally, logistic regression analysis was used to explore the relationship between the three venous times and prognosis, and receiver operating characteristic (ROC) curves were plotted to assess the value of delayed cortical venous imaging in predicting prognosis. RESULTS: 149 LVO-AIS and 73 normal controls were collected. VIT, VPT, and VOT were significantly delayed on the affected side in the patient group compared with the healthy side (p<0.05) and the controls (p<0.05); VIT and VPT were also significantly delayed on the healthy side of patients compared with the controls (p<0.05). Delayed VIT and VPT on the affected side in the patient group were more significant in patients with poor collateral circulation (p<0.05), and VIT and VPT on the affected side in the patient group were negatively correlated with arterial collateral scores. VIT and VPT were significantly delayed in both sides of patients in the poor prognosis group compared with the good prognosis group (p<0.05). logistic regression showed that patients' affected VPT, arterial collateral scores, and NIHSS were independent predictors of poor prognosis, with an accuracy of 79.6% in predicting poor prognosis. The affected VPT and NIHSS were independent predictors of poor prognosis for patients presenting within 24 hours, with an accuracy of 79.6% in predicting poor prognosis. CONCLUSION: Cortical venous flow was significantly slowed in both sides of LVO-AIS patients. delayed ipsilateral VPT in LVO-AIS patients can be used as an imaging indicator to determine poor collateral circulation and predict poor prognosis.

Estimation of ischemic core in acute ischemic stroke with CT angiography and non-contrast CT: Attenuation changes in ASPECTS regions vs. automated ASPECTS scoring.

Purpose: Reperfusion therapies for acute ischemic stroke due to large-vessel occlusion (AIS-LVO) are highly time-dependent, and large infarction is related to poor outcomes and risk of symptomatic hemorrhage. It is of significance to investigate and optimize the screening means and selection criteria for reperfusion therapies to identify more appropriate patients with better outcomes. This study aimed to compare the performance of attenuation changes vs. automated Alberta Stroke Program Early CT Score (ASPECTS) and using CT angiography (CTA) source images vs. non-contrast CT (NCCT) in distinguishing the infarction extent of ischemic core volumes ≥ 70 ml within different time windows. Methods: A total of 73 patients with AIS-LVO who received multimodal CT were analyzed. The automated software was used to calculate ASPECTS. Attenuation change was defined as the sum of products of relative Hounsfield unit (rHU) values times weighting factors of all 10 ASPECTS regions. rHU value of each region was the HU of the ischemic side over that of the contralateral. The corresponding weighting factors were the regression coefficients derived from a multivariable linear regression model which was used to correlate regional rHU with ischemic core volumes, because each region in the ASPECTS template is weighted disproportionally in the ASPECTS system. Automated ASPECTS and attenuation changes were both calculated using CTA and NCCT, respectively. Results: Attenuation changes were correlated with ischemic core volumes within different time windows (Rho ranging from 0.439 to 0.637). In classification of the ischemic core ≥ 70 ml, the performances of attenuation changes were comparable with ASPECTS (area under the curve [AUC] ranging from 0.799 to 0.891), with DeLong's test (P = 0.079, P = 0.373); using CTA (AUC = 0.842) was not different from NCCT (AUC = 0.838). Conclusion: Attenuation changes in ASPECTS regions were correlated with ischemic core volumes. In the classification of infarction volumes, attenuation changes had a high diagnostic ability comparable with automated ASPECTS. Measurement of attenuation changes is not involved in complicated scoring algorithms. This measurement can be used as an available, rapid, reliable, and accurate means to evaluate infarction extent within different time windows. The usefulness of infarction volumes measured by attenuation changes to identify more appropriate patients for reperfusion therapies can be validated in future clinical trials.