Related Experiment Video
Updated: Jul 29, 2025
Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies
Published on: December 15, 2023
Application of Pseudo-Three-Dimensional Residual Network to Classify the Stages of Moyamoya Disease
1School of Health Science & Engineering, University of Shanghai for Science & Technology, Shanghai 200093, China.
Abstract:
It is essential to assess the condition of moyamoya disease (MMD) patients accurately and promptly to prevent MMD from endangering their lives. A Pseudo-Three-Dimensional Residual Network (P3D ResNet) was proposed to process spatial and temporal information, which was implemented in the identification of MMD stages. Digital Subtraction Angiography (DSA) sequences were split into mild, moderate and severe stages in accordance with the progression of MMD, and divided into a training set, a verification set, and a test set with a ratio of 6:2:2 after data enhancement. The features of the DSA images were processed using decoupled three-dimensional (3D) convolution. To increase the receptive field and preserve the features of the vessels, decoupled 3D dilated convolutions that are equivalent to two-dimensional dilated convolutions, plus one-dimensional dilated convolution, were utilized in the spatial and temporal domains, respectively. Then, they were coupled in serial, parallel, and serial-parallel modes to form P3D modules based on the structure of the residual unit. The three kinds of module were placed in a proper sequence to create the complete P3D ResNet. The experimental results demonstrate that the accuracy of P3D ResNet can reach 95.78% with appropriate parameter quantities, making it easy to implement in a clinical setting.
Insights
A new Pseudo-Three-Dimensional Residual Network (P3D ResNet) accurately identifies moyamoya disease (MMD) stages using Digital Subtraction Angiography (DSA) scans. This AI approach achieves 95.78% accuracy, aiding prompt clinical assessment and patient care.
Area of Science:
- Medical Imaging Analysis
- Artificial Intelligence in Medicine
- Neurology
Background:
- Accurate and timely assessment of moyamoya disease (MMD) is crucial for preventing life-threatening complications.
- Digital Subtraction Angiography (DSA) is a key imaging modality for evaluating MMD progression.
Purpose of the Study:
- To develop and validate a novel deep learning model for automated MMD staging using DSA sequences.
- To enhance the accuracy and efficiency of MMD diagnosis in clinical settings.
Main Methods:
- A Pseudo-Three-Dimensional Residual Network (P3D ResNet) was designed to process spatial and temporal information from DSA sequences.
- Decoupled 3D dilated convolutions were employed in spatial and temporal domains to enhance feature extraction.
- DSA data was augmented and split into training, verification, and test sets (6:2:2 ratio) for model development and evaluation.
Main Results:
- The P3D ResNet model achieved a high accuracy of 95.78% in identifying MMD stages (mild, moderate, severe).
- The model demonstrated efficient processing of spatial and temporal features within DSA images.
- The architecture effectively increased the receptive field while preserving vessel features.
Conclusions:
- The proposed P3D ResNet offers a highly accurate and clinically feasible method for MMD staging.
- This AI-driven approach can significantly aid clinicians in the prompt and precise assessment of moyamoya disease.
- The P3D ResNet shows promise for improving patient outcomes through early and accurate diagnosis.
