Comparative study of multi-modal MRI automatic post-processing software based on multicenter data with patients of acute ischemic stroke
10.3760/cma.j.cn112149-20240130-00055
- VernacularTitle:基于多中心急性缺血性脑卒中多模态MRI数据全自动后处理软件的对比研究
- Author:
Mingming WANG
1
;
Hongyu GAO
;
Zhenying CAI
;
Yuan RAO
;
Shuangxing HOU
;
Yu LUO
;
Qi YANG
Author Information
1. 同济大学附属上海市第四人民医院放射科，上海 200434
- Keywords:
Stroke;
Magnetic resonance imaging;
Infarct core;
Ischemic penumbra;
Automatic post-processing software
- From:
Chinese Journal of Radiology
2024;58(6):633-639
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To investigate the consistency of domestic F-STROKE, NeuBrainCARE MRI automatic post-processing software and RAPID MRI automatic post-processing software in the output of infarction core area volume, time-to-maximum volume and ischemic penumbra volume in patients with acute ischemic stroke.Methods:The research was cross-sectional. The clinical and imaging data of patients with acute ischemic stroke from January 2016 to March 2021 were retrospectively collected, including 149 cases from Shanghai Fourth People′s Hospital Affiliated to Tongji University (Center 1), 120 cases from Langfang Changzheng Hospital of Hebei Province (Center 2), and 45 cases from Wuzhou Workers Hospital (Center 3). All patients underwent diffusion weighted imaging (DWI) and dynamic magnetic sensitivity contrast-perfusion weighted imaging (DSC-PWI). RAPID, F-STROKE and NeuBrainCARE automatic post-processing software were used to perform automatic post-processing analysis of MRI images of all patients with acute ischemic stroke. The infarct core (apparent diffusion coefficient<620×10 -6 mm 2/s) volume, time-to-maximum (T max>6 s) volume and the ischemic penumbra (PWI-DWI mismatch) volume were output. The Wilcoxon test was used to analyze the difference between F-STROKE, NeuBrainCARE, and RAPID software outputs of infarct core volume, time to maximum peak volume, and ischemic penumbra volume. Bland-Altman and intraclass correlation coefficient ( ICC) were used to analyze the consistency of the infarct core volume, time-to-maximum volume and ischemic penumbra volume output by F-STROKE, NeuBrainCARE and RAPID software. Results:There were statistically significant differences in the core infarct volume between F-STROKE and RAPID software, NeuBrainCARE and RAPID software ( Z=-10.17, -5.43, both P<0.001). There were significant differences in the time-to-maximum volume between F-STROKE and RAPID software, NeuBrainCARE and RAPID software ( Z=-3.17, -5.51, both P<0.05). There was no significant difference in the ischemic penumbra volume between F-STROKE software and RAPID software ( Z=-1.43, P=0.153), and there was significant difference in the ischemic penumbra volume between NeuBrainCARE software and RAPID software ( Z=-6.45, P<0.05). Bland-Altman analysis showed that the values within the limits of agreement accounted for more than 93.31% of all point values. ICC analysis showed high agreement between F-STROKE, NeuBrainCARE, and RAPID software outputs of infarct core volume, time to maximum peak volume, and ischemic penumbra volume ( ICC>0.6). Conclusion:Domestic F-STROKE software, NeuBrainCARE software and RAPID software have good consistency in evaluating the infarct core volume, time-to-maximum volume and ischemic penumbra volume in patients with acute ischemic stroke, which is worthy of clinical promotion.
