What question did this study set out to answer?

The study aims to create a deep learning model to predict Ki-67 expression levels in gliomas using preoperative multimodal MRI data.

June 20, 2026BMC Medical ImagingOpen Access

A transformer-based multimodal deep learning model for preoperative prediction of Ki-67 expression level in glioma

View Full Paper

Ask AI

Bookmark

Authors

Randomized trial develops a multimodal model predicting Ki-67 levels in glioma, suggesting enhanced diagnostic capabilities.

The study aims to create a deep learning model to predict Ki-67 expression levels in gliomas using preoperative multimodal MRI data.
Collected clinical and imaging data from 421 patients with grade 2-4 gliomas.
Developed three vision transformer models (CE-T1WI_ViT, CE-T2FLAIR_ViT, Combined_ViT) for prediction.
Evaluated model performance using the area under the receiver operating characteristic curve (AUC).
CE-T1WI_ViT achieved an AUC of 0.859 (95% CI, 0.792-0.926).
Combined_ViT model outperformed CNNs with an AUC of 0.922 (95% CI, 0.868-0.976).
All transformer models demonstrated high predictive accuracy for Ki-67 levels, especially Combined_ViT.

Liu et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark