October 2013 - January 2015
Gaussian processes have received increased attention in the machine-learning community in the last two decades, and have been widely used for various tasks, such as dimensionality reduction, classification and regression. Here, an extension of this framework was explored.
Project overview
Gaussian process (GP) models are a flexible means of performing non-parametric Bayesian regression. However, the majority of existing work using GP models in healthcare data is defined for univariate output time-series, denoted as single-task GPs (STGP). In this project, we investigated how GPs could be used to model multiple correlated univariate physiological time-series simultaneously. The resulting multi-task GP (MTGP) framework can learn the correlation within multiple signals even though they might be sampled at different frequencies and have training sets available for different intervals. We illustrate the basic properties of MTGPs using a synthetic case-study with respiratory motion data. Finally, two real-world biomedical problems are investigated from the field of patient monitoring and motion compensation in radiotherapy. The results are compared to STGPs and other standard methods in the respective fields. In both cases, MTGPs learned the correlation between physiological time-series efficiently, which leads to improved modelling accuracy.
Download
Find the details for downloading the (Matlab) source code here.
Using the code should be straightforward: the download comes with some toy datasets on which the demos (included in the link above) can be run. Alternatively, use the scripts in the “example” folder to perform the same.
Relevant publications
[PDF]
Multitask Gaussian processes for multivariate physiological time-series analysis
R Dürichen, MAF Pimentel, L Clifton, A Schweikard, DA Clifton
IEEE Transactions on Biomedical Engineering, 2015, 62 (1), 314-22
[PDF]
Multi-task Gaussian process models for biomedical applications
R Dürichen, MAF Pimentel, L Clifton, A Schweikard, DA Clifton
In Proceedings of the IEEE-EMBS International Conference on Biomedical and Health Informatics (BHI), 2014, 492-5
[PDF]
A unified approach for respiratory motion prediction and correlation with multi-task Gaussian Processes
R Dürichen, T Wissel, F Ernst, MAF Pimentel, L Clifton, DA Clifton, A Schweikard
In Proceedings of the IEEE International Workshop on Machine Learning for Signal Processing, 2014, 1-6
Acknowledgements
A few people were involved on this work, including Robert Dürichen, Lei Clifton and David Clifton. The project is also described here.