Multi-scale segmentation in GBM treatment using diffusion tensor imaging.
Haji Hosseini Khani, Mohammad Reza
Computers in biology and medicine
MetadataShow full item record
Rahmat, R., Saednia, K., Haji Hosseini Khani, M. R., Rahmati, M., Jena, R., & Price, S. (2020). Multi-scale segmentation in GBM treatment using diffusion tensor imaging.. Computers in biology and medicine, 123 103815. https://doi.org/10.1016/j.compbiomed.2020.103815
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. Di usion tensor magnetic resonance imaging is a recently developed technique (DTI) that is exquisitely sensitive to the ordered di ffusion 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 tumour in ltration and patient survival. However, tensor decomposition of DTI data is not commonly used for neurosurgery or radiotherapy treatment planning due to di culties in segmenting the resultant image maps. For this reason, automated techniques for segmentation of tensor decomposition maps would have signi cant clinical utility. In this paper, we modi ed 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 de ned ground truths for better understanding of the contribution of each image sequence to the segmentation performance. The high accuracy and e ciency of our proposed model demonstrates the potential of utilizing di usion MR images for target definition in precision radiation treatment planning and surgery in routine clinical practice.
Humans, Glioblastoma, Brain Neoplasms, Magnetic Resonance Imaging, Image Processing, Computer-Assisted, Adult, Diffusion Tensor Imaging, Neural Networks, Computer
CRUK Project grant - PRaM-GBM study (C9216/A19732) NIHR Clinician Scientist Fellowship (project reference NIHR/CS/009/011) and an NIHR Career Development Fellowship (project reference CDF-2018-11-ST2-003) for SJP.
Cancer Research UK (19732)
Department of Health (via National Institute for Health Research (NIHR)) (CDF-2018-11-ST2-003)
NIHR Academy (NIHRDH-CDF-2018-11-ST2-003)
National Institute for Health Research (NIHRDH-NIHR/CS/009/011)
External DOI: https://doi.org/10.1016/j.compbiomed.2020.103815
This record's URL: https://www.repository.cam.ac.uk/handle/1810/305271
Attribution-NonCommercial-NoDerivatives 4.0 International
Licence URL: https://creativecommons.org/licenses/by-nc-nd/4.0/