1
|
Ren H, Song H, Wang J, Xiong H, Long B, Gong M, Liu J, He Z, Liu L, Jiang X, Li L, Li H, Cui S, Li Y. A clinical-radiomics model based on noncontrast computed tomography to predict hemorrhagic transformation after stroke by machine learning: a multicenter study. Insights Imaging 2023; 14:52. [PMID: 36977913 PMCID: PMC10050271 DOI: 10.1186/s13244-023-01399-5] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/22/2022] [Accepted: 03/08/2023] [Indexed: 03/30/2023] Open
Abstract
OBJECTIVE To build a clinical-radiomics model based on noncontrast computed tomography images to identify the risk of hemorrhagic transformation (HT) in patients with acute ischemic stroke (AIS) following intravenous thrombolysis (IVT). MATERIALS AND METHODS A total of 517 consecutive patients with AIS were screened for inclusion. Datasets from six hospitals were randomly divided into a training cohort and an internal cohort with an 8:2 ratio. The dataset of the seventh hospital was used for an independent external verification. The best dimensionality reduction method to choose features and the best machine learning (ML) algorithm to develop a model were selected. Then, the clinical, radiomics and clinical-radiomics models were developed. Finally, the performance of the models was measured using the area under the receiver operating characteristic curve (AUC). RESULTS Of 517 from seven hospitals, 249 (48%) had HT. The best method for choosing features was recursive feature elimination, and the best ML algorithm to build models was extreme gradient boosting. In distinguishing patients with HT, the AUC of the clinical model was 0.898 (95% CI 0.873-0.921) in the internal validation cohort, and 0.911 (95% CI 0.891-0.928) in the external validation cohort; the AUC of radiomics model was 0.922 (95% CI 0.896-0.941) and 0.883 (95% CI 0.851-0.902), while the AUC of clinical-radiomics model was 0.950 (95% CI 0.925-0.967) and 0.942 (95% CI 0.927-0.958) respectively. CONCLUSION The proposed clinical-radiomics model is a dependable approach that could provide risk assessment of HT for patients who receive IVT after stroke.
Collapse
Affiliation(s)
- Huanhuan Ren
- Department of Radiology, The First Affiliated Hospital of Chongqing Medical University, No. 1 Youyi Road, Yuzhong District, Chongqing, 400016, China
- Department of Radiology, Chongqing General Hospital, Chongqing, China
| | - Haojie Song
- College of Computer and Information Science, Chongqing Normal University, No. 37, Middle University Town Road, Shapingba District, Chongqing, 400016, China
| | - Jingjie Wang
- Department of Radiology, The First Affiliated Hospital of Chongqing Medical University, No. 1 Youyi Road, Yuzhong District, Chongqing, 400016, China
| | - Hua Xiong
- Department of Radiology, Chongqing General Hospital, Chongqing, China
| | - Bangyuan Long
- Department of Radiology, Chongqing General Hospital, Chongqing, China
| | - Meilin Gong
- Department of Radiology, Chongqing General Hospital, Chongqing, China
| | - Jiayang Liu
- Department of Radiology, The First Affiliated Hospital of Chongqing Medical University, No. 1 Youyi Road, Yuzhong District, Chongqing, 400016, China
| | - Zhanping He
- Department of Radiology, Haikou Affiliated Hospital of Central South University Xiangya School of Medicine, Haikou, China
| | - Li Liu
- Department of Radiology, People's Hospital of Yubei District of Chongqing City, Chongqing, China
| | - Xili Jiang
- Department of Radiology, The Second People's Hospital of Hunan Province/Brain Hospital of Hunan Province, Changsha, China
| | - Lifeng Li
- Department of Radiology, Changsha Central Hospital (The Affiliated Changsha Central Hospital, Hengyang Medical School, University of South China), Changsha, China
| | - Hanjian Li
- Department of Radiology, The First Affiliated Hospital of Hainan Medical University, Haikou, China
| | - Shaoguo Cui
- College of Computer and Information Science, Chongqing Normal University, No. 37, Middle University Town Road, Shapingba District, Chongqing, 400016, China.
| | - Yongmei Li
- Department of Radiology, The First Affiliated Hospital of Chongqing Medical University, No. 1 Youyi Road, Yuzhong District, Chongqing, 400016, China.
| |
Collapse
|
2
|
Wen X, Shu Z, Li Y, Hu X, Gong X. Developing a model for estimating infarction onset time based on computed tomography radiomics in patients with acute middle cerebral artery occlusion. BMC Med Imaging 2021; 21:147. [PMID: 34635087 PMCID: PMC8507216 DOI: 10.1186/s12880-021-00678-1] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/20/2021] [Accepted: 09/29/2021] [Indexed: 11/10/2022] Open
Abstract
BACKGROUND Radiomics analysis is a newly emerging quantitative image analysis technique. The aim of this study was to extract a radiomics signature from the computed tomography (CT) imaging to determine the infarction onset time in patients with acute middle cerebral artery occlusion (MCAO). METHODS A total of 123 patients with acute MCAO in the M1 segment (85 patients in the development cohort and 38 patients in the validation cohort) were enrolled in the present study. Clinicoradiological profiles, including head CT without contrast enhancement and computed tomographic angiography (CTA), were collected. The time from stroke onset (TFS) was classified into two subcategories: ≤ 4.5 h, and > 4.5 h. The middle cerebral artery (MCA) territory on CT images was segmented to extract and score the radiomics features associated with the TFS. In addition, the clinicoradiological factors related to the TFS were identified. Subsequently, a combined model of the radiomics signature and clinicoradiological factors was constructed to distinguish the TFS ≤ 4.5 h. Finally, we evaluated the overall performance of our constructed model in an external validation sample of ischemic stroke patients with acute MCAO in the M1 segment. RESULTS The area under the curve (AUC) of the radiomics signature for discriminating the TFS in the development and validation cohorts was 0.770 (95% confidence interval (CI): 0.665-0.875) and 0.792 (95% CI: 0.633-0.950), respectively. The AUC of the combined model comprised of the radiomics signature, age and ASPECTS on CT in the development and validation cohorts was 0.808 (95% CI: 0.701-0.916) and 0.833 (95% CI: 0.702-0.965), respectively. In the external validation cohort, the AUC of the radiomics signature was 0.755 (95% CI: 0.614-0.897), and the AUC of the combined model was 0.820 (95% CI: 0.712-0.928). CONCLUSIONS The CT-based radiomics signature is a valuable tool for discriminating the TFS in patients with acute MCAO in the M1 segment, which may guide the use of thrombolysis therapy in patients with indeterminate stroke onset time.
Collapse
Affiliation(s)
- Xuehua Wen
- Department of Radiology, Zhejiang Provincial People's Hospital, Affiliated People's Hospital, Hangzhou Medical College, Hangzhou, Zhejiang, China
| | - Zhenyu Shu
- Department of Radiology, Zhejiang Provincial People's Hospital, Affiliated People's Hospital, Hangzhou Medical College, Hangzhou, Zhejiang, China
| | - Yumei Li
- Department of Radiology, Zhejiang Provincial People's Hospital, Affiliated People's Hospital, Hangzhou Medical College, Hangzhou, Zhejiang, China
| | - Xingfei Hu
- Department of Radiology, Zhejiang Provincial People's Hospital, Affiliated People's Hospital, Hangzhou Medical College, Hangzhou, Zhejiang, China
| | - Xiangyang Gong
- Department of Radiology, Zhejiang Provincial People's Hospital, Affiliated People's Hospital, Hangzhou Medical College, Hangzhou, Zhejiang, China. .,Institute of Artificial Intelligence and Remote Imaging, Hangzhou Medical College, Hangzhou, Zhejiang, China.
| |
Collapse
|