Brainstem gliomas (BSGs) are a cancerous glioma tumor that occur in the brainstem. Diffuse intrinsic pontine gliomas (DIPG) account for 80% of BSGs in children and 45~50% in adults. Nearly 80% of pediatric DIPGs are induced by reprogramming the histone H3 K27 methylation and gene expression. Therefore ,3 K27M mutation can be used as a qualified biomarker for diagnosis and therapy selection for DIPG patients. In addition, the detection and precise localization of brainstem glioma tissue is crucial for diagnosis, surgical planning, postoperative analysis, chemo/ radiotherapy planning in neurosurgery and radiomics research.
We present a novel cascaded deep convolutional neural network (CNN) to address two challenging tasks, automatic segmentation of brainstem gliomas and prediction of genotype (H3 K27M) mutation status) simultaneously. Our novel segmentation task contains two feature-fusion modules: the Gaussian-pyramid multiscale input features-fusion technique and the brainstem-region feature enhancement. The aim is to resolve very difficult problems in brainstem glioma segmentation. Our prediction model combines CNN features and support-vector-machine (SVM) classifier to automatically predict genotypes without region-of-interest (ROI) labelled-MR images and is learned jointly with the segmentation task. Firstly, Gaussian-pyramid multiscale input feature fusion is added to our glioma-segmentation task to solve the problems of size variety and weak brainstem-gliomas boundaries. Secondly, the two feature-fusion modules provide both local and global contexts to retain higher-frequency details for sharper tumor boundaries, handling the problem of the large variation of tumor shape and volume resolution. Results and Conclusion: Experiments demonstrate that our multi-task CNN method achieves not only a good tumor segmentation result with a high Dice similarity coefficient (DSC) of 77.03 %, but also a competitive genotype prediction result with an average accuracy of 94.85% upon 5-fold cross-validation.