Related Experiment Video
Updated: Nov 4, 2025
Identification of Disease-related Spatial Covariance Patterns using Neuroimaging Data
Published on: June 26, 2013
Learning spatiotemporal features of DSA using 3D CNN and BiConvGRU for ischemic moyamoya disease detection
1Department of Electronic Engineering, Fudan University, Shanghai, China.
Background:
Moyamoya disease (MMD) is a serious intracranial cerebrovascular disease. Cerebral hemorrhage caused by MMD will bring life risk to patients. Therefore, MMD detection is of great significance in the prevention of cerebral hemorrhage. In order to improve the accuracy of digital subtraction angiography (DSA) in the diagnosis of ischemic MMD, in this paper, a deep network architecture combined with 3D convolutional neural network (3D CNN) and bidirectional convolutional gated recurrent unit (BiConvGRU) is proposed to learn the spatiotemporal features for ischemic MMD detection.
Methods:
Firstly, 2D convolutional neural network (2D CNN) is utilized to extract spatial features for each frame of DSA. Secondly, the long-term spatiotemporal features of DSA sequence are extracted by BiConvGRU. Thirdly, the short-term spatiotemporal features of DSA are further extracted by 3D convolutional neural network (3D CNN). In addition, different features are extracted when gray images and optical flow images pass through the network, and multiple features are extracted by features fusion. Finally, the fused features are utilized to classify.
Results:
The proposed method was quantitatively evaluated on a data sets of 630 cases. The experimental results showed a detection accuracy of 0.9788, sensitivity and specificity were 0.9780 and 0.9796, respectively, and area under curve (AUC) was 0.9856. Compared with other methods, we can get the highest accuracy and AUC.
Conclusions:
The experimental results show that the proposed method is stable and reliable for ischemic MMD detection, which provides an option for doctors to accurately diagnose ischemic MMD.
Insights
This study introduces a novel deep learning model for detecting Moyamoya disease (MMD) using digital subtraction angiography (DSA). The method accurately identifies ischemic MMD, aiding in preventing life-threatening cerebral hemorrhage.
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Neurology
Background:
- Moyamoya disease (MMD) is a critical intracranial cerebrovascular condition.
- Cerebral hemorrhage in MMD patients poses significant life risks, underscoring the importance of early detection.
- Accurate diagnosis of ischemic MMD is crucial for preventing severe complications.
Purpose of the Study:
- To enhance the accuracy of digital subtraction angiography (DSA) for diagnosing ischemic Moyamoya disease.
- To develop a deep network architecture for effective spatiotemporal feature extraction in MMD detection.
Main Methods:
- A hybrid deep network combining 3D Convolutional Neural Network (3D CNN) and Bidirectional Convolutional Gated Recurrent Unit (BiConvGRU) was employed.
- Spatial features were extracted from DSA frames using 2D CNN, while BiConvGRU captured long-term spatiotemporal features and 3D CNN captured short-term features.
- Feature fusion from gray and optical flow images was performed for improved classification.
Main Results:
- The proposed method achieved a detection accuracy of 0.9788 on a dataset of 630 cases.
- High sensitivity (0.9780) and specificity (0.9796) were reported, with an Area Under the Curve (AUC) of 0.9856.
- The method demonstrated superior accuracy and AUC compared to existing approaches.
Conclusions:
- The developed deep learning model is stable and reliable for detecting ischemic MMD.
- This approach offers a valuable tool for clinicians in accurately diagnosing ischemic Moyamoya disease.
