TY - GEN
T1 - Brain tumor cell density estimation from multi-modal MR images based on a synthetic tumor growth model
AU - Geremia, Ezequiel
AU - Menze, Bjoern H.
AU - Prastawa, Marcel
AU - Weber, M. A.
AU - Criminisi, Antonio
AU - Ayache, Nicholas
PY - 2013
Y1 - 2013
N2 - This paper proposes to employ a detailed tumor growth model to synthesize labelled images which can then be used to train an efficient data-driven machine learning tumor predictor. Our MR image synthesis step generates images with both healthy tissues as well as various tumoral tissue types. Subsequently, a discriminative algorithm based on random regression forests is trained on the simulated ground truth to predict the continuous latent tumor cell density, and the discrete tissue class associated with each voxel. The presented method makes use of a large synthetic dataset of 740 simulated cases for training and evaluation. A quantitative evaluation on 14 real clinical cases diagnosed with low-grade gliomas demonstrates tissue class accuracy comparable with state of the art, with added benefit in terms of computational efficiency and the ability to estimate tumor cell density as a latent variable underlying the multimodal image observations. The idea of synthesizing training data to train data-driven learning algorithms can be extended to other applications where expert annotation is lacking or expensive.
AB - This paper proposes to employ a detailed tumor growth model to synthesize labelled images which can then be used to train an efficient data-driven machine learning tumor predictor. Our MR image synthesis step generates images with both healthy tissues as well as various tumoral tissue types. Subsequently, a discriminative algorithm based on random regression forests is trained on the simulated ground truth to predict the continuous latent tumor cell density, and the discrete tissue class associated with each voxel. The presented method makes use of a large synthetic dataset of 740 simulated cases for training and evaluation. A quantitative evaluation on 14 real clinical cases diagnosed with low-grade gliomas demonstrates tissue class accuracy comparable with state of the art, with added benefit in terms of computational efficiency and the ability to estimate tumor cell density as a latent variable underlying the multimodal image observations. The idea of synthesizing training data to train data-driven learning algorithms can be extended to other applications where expert annotation is lacking or expensive.
UR - http://www.scopus.com/inward/record.url?scp=84875200185&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-36620-8_27
DO - 10.1007/978-3-642-36620-8_27
M3 - Conference contribution
AN - SCOPUS:84875200185
SN - 9783642366192
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 273
EP - 282
BT - Medical Computer Vision
T2 - 2nd MICCAI Workshop on Medical Computer Vision, MICCAI-MCV 2012, Held in Conjunction with the 15th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2012
Y2 - 5 October 2012 through 5 October 2012
ER -