TY - GEN
T1 - Bayesian personalization of brain tumor growth model
AU - Lê, Matthieu
AU - Delingette, Hervé
AU - Kalpathy-Cramer, Jayashree
AU - Gerstner, Elizabeth R.
AU - Batchelor, Tracy
AU - Unkelbach, Jan
AU - Ayache, Nicholas
N1 - Publisher Copyright:
© Springer International Publishing Switzerland 2015.
PY - 2015
Y1 - 2015
N2 - Recent work on brain tumor growth modeling for glioblastoma using reaction-diffusion equations suggests that the diffusion coefficient and the proliferation rate can be related to clinically relevant information. However, estimating these parameters is difficult due to the lack of identifiability of the parameters, the uncertainty in the tumor segmentations, and the model approximation, which cannot perfectly capture the dynamics of the tumor. Therefore, we propose a method for conducting the Bayesian personalization of the tumor growth model parameters. Our approach estimates the posterior probability of the parameters, and allows the analysis of the parameters correlations and uncertainty. Moreover, this method provides a way to compute the evidence of a model, which is a mathematically sound way of assessing the validity of different model hypotheses. Our approach is based on a highly parallelized implementation of the reaction-diffusion equation, and the Gaussian Process Hamiltonian Monte Carlo (GPHMC), a high acceptance rate Monte Carlo technique. We demonstrate our method on synthetic data, and four glioblastoma patients. This promising approach shows that the infiltration is better captured by the model compared to the speed of growth.
AB - Recent work on brain tumor growth modeling for glioblastoma using reaction-diffusion equations suggests that the diffusion coefficient and the proliferation rate can be related to clinically relevant information. However, estimating these parameters is difficult due to the lack of identifiability of the parameters, the uncertainty in the tumor segmentations, and the model approximation, which cannot perfectly capture the dynamics of the tumor. Therefore, we propose a method for conducting the Bayesian personalization of the tumor growth model parameters. Our approach estimates the posterior probability of the parameters, and allows the analysis of the parameters correlations and uncertainty. Moreover, this method provides a way to compute the evidence of a model, which is a mathematically sound way of assessing the validity of different model hypotheses. Our approach is based on a highly parallelized implementation of the reaction-diffusion equation, and the Gaussian Process Hamiltonian Monte Carlo (GPHMC), a high acceptance rate Monte Carlo technique. We demonstrate our method on synthetic data, and four glioblastoma patients. This promising approach shows that the infiltration is better captured by the model compared to the speed of growth.
UR - http://www.scopus.com/inward/record.url?scp=84951019858&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84951019858&partnerID=8YFLogxK
U2 - 10.1007/978-3-319-24571-3_51
DO - 10.1007/978-3-319-24571-3_51
M3 - Conference contribution
AN - SCOPUS:84951019858
SN - 9783319245706
SN - 9783319245706
SN - 9783319245706
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 424
EP - 432
BT - Medical Image Computing and Computer-Assisted Intervention - MICCAI 2015 - 18th International Conference, Proceedings
A2 - Hornegger, Joachim
A2 - Frangi, Alejandro F.
A2 - Wells, William M.
A2 - Frangi, Alejandro F.
A2 - Navab, Nassir
A2 - Hornegger, Joachim
A2 - Navab, Nassir
A2 - Wells, William M.
A2 - Wells, William M.
A2 - Frangi, Alejandro F.
A2 - Hornegger, Joachim
A2 - Navab, Nassir
PB - Springer-Verlag
T2 - 18th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2015
Y2 - 5 October 2015 through 9 October 2015
ER -