AUTHOR=Yuan Yifan , Yu Yang , Chang Jun , Chu Ying-Hua , Yu Wenwen , Hsu Yi-Cheng , Patrick Liebig Alexander , Liu Mianxin , Yue Qi 

TITLE=Convolutional neural network to predict IDH mutation status in glioma from chemical exchange saturation transfer imaging at 7 Tesla

JOURNAL=Frontiers in Oncology

VOLUME=Volume 13 - 2023

YEAR=2023

URL=https://www.frontiersin.org/journals/oncology/articles/10.3389/fonc.2023.1134626

DOI=10.3389/fonc.2023.1134626

ISSN=2234-943X

ABSTRACT=Background
Noninvasive prediction of isocitrate dehydrogenase (IDH) mutation status in glioma guides surgical strategies and individualized management. We explored the capability on preoperatively identifying IDH status of combining a convolutional neural network (CNN) and a novel imaging modality, ultra-high field 7.0 Tesla chemical exchange saturation transfer (CEST) imaging.
Method
We enrolled 84 glioma patients of different tumor grades in this retrospective study. Amide proton transfer CEST and structural MR imaging at 7T were performed preoperatively. And the tumor regions are manually segmented, leading to the “annotation” maps that offers the location and shape information of the tumors. The tumor region slices in CEST and T1 were further cropped out as samples and combined with the annotation maps, which were inputted to a 2D-CNN model for IDH predictions. Further comparison analysis to radiomics-based prediction methods was performed to demonstrate the crucial role of CNN for predicting IDH based on CEST and T1 images.
Results
A five-fold cross-validation was performed on the 84 patients and 4090 slices. We observed a model based on only CEST achieved accuracy of 74.01±1.15%, and the AUC of 0.8022±0.0147. When using T1 image only, the prediction performances dropped to accuracy of 72.52±1.12% and AUC of 0.7904±0.0214, which indicates minor superiority of CEST over T1.  However, when we combined CEST and T1 together with annotation maps, the performances of the CNN model were further boosted to accuracy of 82.94±1.23% and AUC of 0.8868±0.0055, suggesting the importance of a joint analysis. Finally, using same inputs, the CNN-based predictions achieved significantly improved performances above those from radiomics-based predictions (logistic regression and support vector machine) by 10% to 20% in all metrics. 
Conclusion
7T CEST and structural MRI jointly offers improved the sensitivity and specificity of preoperative non-invasive imaging for the diagnosis of IDH mutation status. As the first study of CNN model on imaging acquired at ultra-high field MR, our results could demonstrate the potential of combining ultra-high field CEST and CNN for facilitating decision-making in clinical practice. However, due to the limited cases and B1 inhomogeneities, the accuracy of this model will be improved in our further study.