Automatic segmentation of vestibular schwannomas from T1-weighted MRI with a deep neural network

Abstract

Background: Long-term follow-up using volumetric measurement could significantly assist in the management of vestibular schwannomas (VS). Manual segmentation of VS from MRI for treatment planning and follow-up assessment is labor-intensive and time-consuming. This study aims to develop a deep learning technique to fully automatically segment VS from MRI.

Methods: This study retrospectively analyzed MRI data of 737 patients who received gamma knife radiosurgery for VS. Treatment planning T1-weighted isotropic MR and manually contoured gross tumor volumes (GTV) were used for model development. A 3D convolutional neural network (CNN) was built on ResNet blocks. Spatial attenuation and deep supervision modules were integrated in each decoder level to enhance the training for the small tumor volume on brain MRI. The model was trained and tested on 587 and 150 patient data, respectively, from this institution (n = 495) and a publicly available dataset (n = 242). The model performance were assessed by the Dice similarity coefficient (DSC), 95% Hausdorff distance (HD95), average symmetric surface (ASSD) and relative absolute volume difference (RAVD) of the model segmentation results against the GTVs.

Results: Measured on combined testing data from two institutions, the proposed method achieved mean DSC of 0.91 ± 0.08, ASSD of 0.3 ± 0.4 mm, HD95 of 1.3 ± 1.6 mm, and RAVD of 0.09 ± 0.15. The DSCs were 0.91 ± 0.09 and 0.92 ± 0.06 on 100 testing patients of this institution and 50 of the public data, respectively.

Conclusions: A CNN model was developed for fully automated segmentation of VS on T1-Weighted isotropic MRI. The model achieved good performance compared with physician clinical delineations on a sizeable dataset from two institutions. The proposed method potentially facilitates clinical workflow of radiosurgery for VS patient management.

Keywords: Deep neural network; Image segmentation; MRI; Radiosurgery; Vestibular schwannomas.

Fig. 1

Architecture of the CNN model

Fig. 2

Box-and-whisker plots of the metrics to evaluate model performances on the testing dataset

Fig. 3

Three examples of the automatic segmentation results. Row: VS with different sizes, the last is the smallest in total dataset; Column: axial, coronal, sagittal slices of the MRI. Blue curve: ground truth tumor contours; Red curve: model segmentation results

Fig. 4

Model performances on segmentation of VS with different tumor sizes

Fig. 5

Three outliers of the automatic segmentation on the testing data. Row: (a, b) mixed cystic VSs; (c) post-surgical resection tumor. Column: axial, coronal and sagittal slices of MRI. Blue curve: ground truth tumor contours, Red curve: model segmentation results