Predictive modelling of glioblastoma growth: Simulating tumour propagation using diffusion MRI
Poster No:
1280
Submission Type:
Abstract Submission
Authors:
Aitor Alberdi Escudero1, Liam Butler2, Andrew Sammut3, Kenneth Scerri4, Claude Bajada2
Institutions:
1University of Malta, Msida, NA, 2Department of Physiology and Biochemistry, Faculty of Medicine and Surgery, University of Malta, Msida, Malta, 3Department of Electronic Systems Engineering, Faculty of Engineering, University of Malta, Msida, Malta, 4Department of Systems & Control Engineering, Faculty of Engineering, University of Malta, Msida, Malta
First Author:
Co-Author(s):
Liam Butler
Department of Physiology and Biochemistry, Faculty of Medicine and Surgery, University of Malta
Msida, Malta
Department of Physiology and Biochemistry, Faculty of Medicine and Surgery, University of Malta
Msida, Malta
Andrew Sammut
Department of Electronic Systems Engineering, Faculty of Engineering, University of Malta
Msida, Malta
Department of Electronic Systems Engineering, Faculty of Engineering, University of Malta
Msida, Malta
Kenneth Scerri
Department of Systems & Control Engineering, Faculty of Engineering, University of Malta
Msida, Malta
Department of Systems & Control Engineering, Faculty of Engineering, University of Malta
Msida, Malta
Claude Bajada
Department of Physiology and Biochemistry, Faculty of Medicine and Surgery, University of Malta
Msida, Malta
Department of Physiology and Biochemistry, Faculty of Medicine and Surgery, University of Malta
Msida, Malta
Introduction:
Glioblastomas are aggressive brain tumours that infiltrate neighbouring tissue and predominantly affect the cerebral hemispheres in adults (1,2). Glioblastoma, the most common invasive central nervous system tumour, has an incidence of 3.21 per 100,000 individuals and a poor prognosis due to challenges in tumour localisation, infiltration, and recurrent growth (3). Traditional growth models often assume linear or exponential expansion, and can fail to account for the brain's structural heterogeneity and white matter pathways (4,5). Diffusion MRI, which maps these pathways, can enable tractography-based approaches to more accurately simulate tumour dynamics. This study aims to develop preliminary tumour growth models that leverage tractography-based algorithms to simulate glioblastoma proliferation and infiltration. These models lay the groundwork for future validation with longitudinal data, aiming to improve tumour progression predictions and guide treatment strategies.
Methods:
The Lumiere dataset (glioblastoma patients) and Human Connectome Project (HCP) data (healthy controls) were used (6,7). For this proof-of-concept, one case and one control were analysed. Tumour masks, pre-segmented using DeepBraTumIA, were registered to the space of an HCP participant. To improve alignment, a synthetic image standardising intensity and contrast was generated from the patient image. Registration first employed FLIRT (FMRIB's Linear Image Registration Tool) using linear alignment for initial registration approximation and FNIRT (FMRIB's Nonlinear Image Registration Tool) for nonlinear deformation and ensure more robust registration, with resulting transformations applied to the tumour mask.
HCP diffusion MRI data was processed to generate diffusion (rather than fiber) orientation distribution functions (dODFs).
A tumour mask was used to generate random seeds which then generates the streamlines using MRtrix's iFOD2 algorithm to simulate glioblastoma growth along white matter tracts (8,9). Tractography parameters were adjusted to model growth patterns: a 30° step angle for restricted, elongated growth and 180° for spherical, isotropic growth. Seed number and streamline length were set based on literature (e.g. 5,10), enabling simulation of distinct tumour growth phases, including proliferation and diffusion.
HCP diffusion MRI data was processed to generate diffusion (rather than fiber) orientation distribution functions (dODFs).
A tumour mask was used to generate random seeds which then generates the streamlines using MRtrix's iFOD2 algorithm to simulate glioblastoma growth along white matter tracts (8,9). Tractography parameters were adjusted to model growth patterns: a 30° step angle for restricted, elongated growth and 180° for spherical, isotropic growth. Seed number and streamline length were set based on literature (e.g. 5,10), enabling simulation of distinct tumour growth phases, including proliferation and diffusion.
Results:
Simulations of glioblastoma growth demonstrated distinct patterns under varying tractography parameters. In the first case (Fig 1), minimal restriction (maximum angle: 180°) resulted in spherical tumour expansion, reflecting a proliferative growth phase where fibre direction minimally influenced spread. Streamlines exhibited high voxel density near the tumour origin, consistent with gaussian-like growth. In the second case (Fig 2), stricter constraints (maximum angle: 30°) directed growth along white matter fibres, simulating invasive branching. Tumour infiltration extended further into the frontal lobe, with reduced voxel density due to streamline alignment with fibre pathways. Comparison of the two approaches highlights key differences in tumour dynamics. The unrestricted growth (Figure 1) showed higher voxel density but limited spatial reach, while the constrained growth (Figure 2) demonstrated deeper infiltration into distant regions with fewer hits per voxel. Together, these approaches enable the simulation of both proliferative and invasive phases, setting a baseline for further predictive analysis for tumour progression, size, and potential metastatic regions post-surgery.
Conclusions:
This study shows that there is merit in simulating glioblastoma tumour growth using a tractography-based approach. By making use of the two approaches, it might be possible to simulate both proliferation and invasion. Future work will focus on estimating key parameters-such as dispersion angle and streamline length-necessary for realistic tumour growth modeling from patient data.
Disorders of the Nervous System:
Neurodegenerative/ Late Life (eg. Parkinson’s, Alzheimer’s)
Modeling and Analysis Methods:
Diffusion MRI Modeling and Analysis 1
Other Methods 2
Keywords:
Astrocyte
DISORDERS
White Matter
WHITE MATTER IMAGING - DTI, HARDI, DSI, ETC
Other - Glioblastoma
1|2Indicates the priority used for review
Abstract Information
By submitting your proposal, you grant permission for the Organization for Human Brain Mapping (OHBM) to distribute your work in any format, including video, audio print and electronic text through OHBM OnDemand, social media channels, the OHBM website, or other electronic publications and media.
The Open Science Special Interest Group (OSSIG) is introducing a reproducibility challenge for OHBM 2025. This new initiative aims to enhance the reproducibility of scientific results and foster collaborations between labs. Teams will consist of a “source” party and a “reproducing” party, and will be evaluated on the success of their replication, the openness of the source work, and additional deliverables. Click here for more information. Propose your OHBM abstract(s) as source work for future OHBM meetings by selecting one of the following options:
Please indicate below if your study was a "resting state" or "task-activation” study.
Healthy subjects only or patients (note that patient studies may also involve healthy subjects):
Was this research conducted in the United States?
Were any human subjects research approved by the relevant Institutional Review Board or ethics panel? NOTE: Any human subjects studies without IRB approval will be automatically rejected.
Were any animal research approved by the relevant IACUC or other animal research panel? NOTE: Any animal studies without IACUC approval will be automatically rejected.
Please indicate which methods were used in your research:
For human MRI, what field strength scanner do you use?
Which processing packages did you use for your study?
Provide references using APA citation style.
1. Melhem JM et al. Dose-dependent efficacy of bevacizumab in recurrent glioblastoma. J Neurooncol. 2023 Feb 1;161(3):633–41.
2. Melhem JM et al. Updates in IDH-Wildtype Glioblastoma. Neurotherapeutics. 2022 Oct 1;19(6):1705–23.
3. Tan AC et al. Management of glioblastoma: State of the art and future directions. CA Cancer J Clin. 2020;70(4):299–312.
4. Rutter EM et al. Mathematical analysis of glioma growth in a murine model. Sci Rep. 2017;7(1):2508.
5. Painter K et al. Mathematical modelling of glioma growth: the use of diffusion tensor imaging (DTI) data to predict the anisotropic pathways of cancer invasion. J Theor Biol. 2013;323:25–39.
6. Suter Y et al. The LUMIERE dataset: Longitudinal Glioblastoma MRI with expert RANO evaluation. Sci Data. 2022 Dec 15;9(1):768.
7. Van Essen DC et al. The WU-Minn Human Connectome Project: An overview. NeuroImage. 2013 Oct 15;80:62–79.
8. Tournier JD et al. MRtrix: diffusion tractography in crossing fiber regions. Int J Imaging Syst Technol. 2012;22(1):53–66.
9. Tournier JD et al. tckgen — MRtrix3 3.0 documentation. Available from: https://mrtrix.readthedocs.io/en/dev/reference/commands/tckgen.html
10. Berman JI et al. Probabilistic streamline q-ball tractography using the residual bootstrap. NeuroImage. 2008 Jan 1;39(1):215–22.
Funding Sources
This study is being financed by the Research Innovation and Development Trust (RIDT) of the University of Malta, from donations made by the ALIVE Charity Foundation (Project: BRIAN, Grant ID: RIDT 02/2024).
Acknowledgements
Data were provided [in part] by the Human Connectome Project, WU-Minn Consortium (Principal Investigators: David Van Essen and Kamil Ugurbil; 1U54MH091657) funded by the 16 NIH Institutes and Centers that support the NIH Blueprint for Neuroscience Research; and by the McDonnell Center for Systems Neuroscience at Washington University.