Identifying key factors for predicting O6-Methylguanine-DNA methyltransferase status in adult patients with diffuse glioma: a multimodal analysis of demographics, radiomics, and MRI by variable Vision Transformer.
Takuma UsuzakiKengo TakahashiRyusei InamoriYohei MorishitaTakashi ShizukuishiHidenobu TakagiMami IshikuroTaku ObaraKei TakasePublished in: Neuroradiology (2024)
The vViT is a competent deep learning model in predicting MGMT status. The ED radiomic features of the T2-weighted image demonstrated the most dominant contribution.
Keyphrases
- deep learning
- contrast enhanced
- magnetic resonance imaging
- emergency department
- magnetic resonance
- artificial intelligence
- convolutional neural network
- circulating tumor
- lymph node metastasis
- single molecule
- pain management
- low grade
- cell free
- machine learning
- diffusion weighted imaging
- squamous cell carcinoma
- network analysis
- circulating tumor cells