Deep learning for glioblastoma segmentation using preoperative magnetic resonance imaging identifies volumetric features associated with survival.

BACKGROUND: Measurement of volumetric features is challenging in glioblastoma. We investigate whether volumetric features derived from preoperative MRI using a convolutional neural network-assisted segmentation is correlated with survival. METHODS: Preoperative MRI of 120 patients were scored using Visually Accessible Rembrandt Images (VASARI) features. We trained and tested a multilayer, multi-scale convolutional neural network on multimodal brain tumour segmentation challenge (BRATS) data, prior to testing on our dataset. The automated labels were manually edited to generate ground truth segmentations. Network performance for our data and BRATS data was compared. Multivariable Cox regression analysis corrected for multiple testing using the false discovery rate was performed to correlate clinical and imaging variables with overall survival. RESULTS: Median Dice coefficients in our sample were (1) whole tumour 0.94 (IQR, 0.82-0.98) compared to 0.91 (IQR, 0.83-0.94 p = 0.012), (2) FLAIR region 0.84 (IQR, 0.63-0.95) compared to 0.81 (IQR, 0.69-0.8 p = 0.170), (3) contrast-enhancing region 0.91 (IQR, 0.74-0.98) compared to 0.83 (IQR, 0.78-0.89 p = 0.003) and (4) necrosis region were 0.82 (IQR, 0.47-0.97) compared to 0.67 (IQR, 0.42-0.81 p = 0.005). Contrast-enhancing region/tumour core ratio (HR 4.73 [95% CI, 1.67-13.40], corrected p = 0.017) and necrotic core/tumour core ratio (HR 8.13 [95% CI, 2.06-32.12], corrected p = 0.011) were independently associated with overall survival. CONCLUSION: Semi-automated segmentation of glioblastoma using a convolutional neural network trained on independent data is robust when applied to routine clinical data. The segmented volumes have prognostic significance.

Multi-scale segmentation in GBM treatment using diffusion tensor imaging.

Glioblastoma (GBM) is the commonest primary malignant brain tumor in adults, and despite advances in multi-modality therapy, the outlook for patients has changed little in the last 10 years. Local recurrence is the predominant pattern of treatment failure, hence improved local therapies (surgery and radiotherapy) are needed to improve patient outcomes. Currently segmentation of GBM for surgery or radiotherapy (RT) planning is labor intensive, especially for high-dimensional MR imaging methods that may provide more sensitive indicators of tumor phenotype. Automating processing and segmentation of these images will aid treatment planning. Diffusion tensor magnetic resonance imaging is a recently developed technique (DTI) that is exquisitely sensitive to the ordered diffusion of water in white matter tracts. Our group has shown that decomposition of the tensor information into the isotropic component (p - shown to represent tumor invasion) and the anisotropic component (q - shown to represent the tumor bulk) can provide valuable prognostic information regarding tumor infiltration and patient survival. However, tensor decomposition of DTI data is not commonly used for neurosurgery or radiotherapy treatment planning due to difficulties in segmenting the resultant image maps. For this reason, automated techniques for segmentation of tensor decomposition maps would have significant clinical utility. In this paper, we modified a well-established convolutional neural network architecture (CNN) for medical image segmentation and used it as an automatic multi-sequence GBM segmentation based on both DTI image maps (p and q maps) and conventional MRI sequences (T2-FLAIR and T1 weighted post contrast (T1c)). In this proof-of-concept work, we have used multiple MRI sequences, each with individually defined ground truths for better understanding of the contribution of each image sequence to the segmentation performance. The high accuracy and efficiency of our proposed model demonstrates the potential of utilizing diffusion MR images for target definition in precision radiation treatment planning and surgery in routine clinical practice.

Semi-automated construction of patient individualised clinical target volumes for radiotherapy treatment of glioblastoma utilising diffusion tensor decomposition maps.

OBJECTIVES: Glioblastoma multiforme (GBM) is a highly infiltrative primary brain tumour with an aggressive clinical course. Diffusion tensor imaging (DT-MRI or DTI) is a recently developed technique capable of visualising subclinical tumour spread into adjacent brain tissue. Tensor decomposition through p and q maps can be used for planning of treatment. Our objective was to develop a tool to automate the segmentation of DTI decomposed p and q maps in GBM patients in order to inform construction of radiotherapy target volumes. METHODS: Chan-Vese level set model is applied to segment the p map using the q map as its initial starting point. The reason of choosing this model is because of the robustness of this model on either conventional MRI or only DTI. The method was applied on a data set consisting of 50 patients having their gross tumour volume delineated on their q map and Chan-Vese level set model uses these superimposed masks to incorporate the infiltrative edges. RESULTS: The expansion of tumour boundary from q map to p map is clearly visible in all cases and the Dice coefficient (DC) showed a mean similarity of 74% across all 50 patients between the manually segmented ground truth p map and the level set automatic segmentation. CONCLUSION: Automated segmentation of the tumour infiltration boundary using DTI and tensor decomposition is possible using Chan-Vese level set methods to expand q map to p map. We have provided initial validation of this technique against manual contours performed by experienced clinicians. ADVANCES IN KNOWLEDGE: This novel automated technique to generate p maps has the potential to individualise radiation treatment volumes and act as a decision support tool for the treating oncologist.