Editor’s Choice

Convolutional neural networks are able to learn discriminating features automatically, and these features provide added value for grading gliomas.

Abstract

BACKGROUND AND PURPOSE

Figure 1 from Gutta et al — Representative segmentation result from one glioblastoma patient. Top row: Coronal; Middle row: Sagittal; Bottom row: Axial. T1, T1c, T2, and FLAIR are shown in the first 4 column, after being resampled to 1mm, registered, and skull-stripped. The rightmost column corresponds to the segmentation result overlapped on the FLAIR image. Segmentation was performed using cascaded convolutional networks by Wang et al. [21]. In the segmentation image, green corresponds to edema, yellow corresponds to enhancing, and red corresponds to non-enhancing regions.

Accurate determination of glioma grade leads to improved treatment planning. The criterion standard for glioma grading is invasive tissue sampling. Recently, radiomic features have shown excellent potential in glioma-grade prediction. These features may not fully exploit the underlying information in MR images. The objective of this study was to investigate the performance of features learned by a convolutional neural network compared with standard radiomic features for grade prediction.

MATERIALS AND METHODS

A total of 237 patients with gliomas were included in this study. All images were resampled, registered, skull-stripped, and segmented to extract the tumors. The learned features from the trained convolutional neural network were used for grade prediction. The performance of the proposed method was compared with standard machine learning approaches, support vector machine, random forests, and gradient boosting trained with radiomic features.

RESULTS

The experimental results demonstrate that using learned features extracted from the convolutional neural network achieves an average accuracy of 87%, outperforming the methods considering radiomic features alone. The top-performing machine learning model is gradient boosting with an average accuracy of 64%. Thus, there is a 23% improvement in accuracy, and it is an efficient technique for grade prediction.

CONCLUSIONS

Convolutional neural networks are able to learn discriminating features automatically, and these features provide added value for grading gliomas. The proposed framework may provide substantial improvement in glioma-grade prediction; however, further validation is needed.

Read this article: https://bit.ly/2NGDiQg

Improved Glioma Grading Using Deep Convolutional Neural Networks

AJNR Blog

The Official Blog of the American Journal of Neuroradiology