MOVIII Demonstrator Project: Bio-Feedback Using Real-Time fMRI
Despite the enormous complexity of the human mind, fMRI techniques are able to partially observe the state of a brain in action. In this paper we describe an experimental setup for real-time fMRI in a bio-feedback loop. One of the main challenges in the project is to reach a detection speed, accuracy and spatial resolution necessary to attain sufficient bandwidth of communication to close the bio-feedback loop. To this end we have banked on our previous work on real-time filtering for fMRI and system identification, which has been tailored for use in the experiment setup. In the experiments presented the system is trained to estimate where a person in the MRI scanner is looking from signals derived from the visual cortex only. We have been able to demonstrate that the user can induce an action and perform simple tasks with her mind sensed using real-time fMRI.The technique may have several clinical applications, for instance to allow paralyzed and "locked in" people to communicate with the outside world. In the meanwhile, the need for improved fMRI performance and brain state detection poses a challenge to the signal processing community. We also expect that the setup will serve as an invaluable tool for neuro science research in general.
Anders Ynnerman , Prof.
Hans Knutsson , Prof.
Henrik Ohlsson , Doktorand
Joakim Rydell , Postdoc
Anders Brun , Postdoc
Jacob Roll , Fo-Ass
Mats Andersson , Ass. Prof.
Anders Eklund , Doktorand
- Former Staff:
- Project Description:
The fMRI group within LiU's Center for Medical Image Science and Visualization (CMIV) has developed a new technique for analyzing fMRI data (Canonical Correlation Analysis) [Friman 2003, 2004], which is widely considered the best of all presently available methods for analyzing fMRI data. Other areas of research include the development of examination paradigms for neuroscience, new data acquisition methods, and clinical use of the fMRI method [Engström 2005]. Available equipment includes a new Philips Achieva system with state-of-the-art hardware and software.
Spatio-temporal signal processing
Detection of active brain areas is a highly challenging problem. A successful method will need to efficiently model the spatio-temporal structure of the activation in order to handle the very high level of noise. I.e. models that can adapt both to the, perhaps time-varying, 3-dimensional shapes of the activated regions as well as the time courses of the activated voxels. A fundamental theoretical tool in the research will continue to be Canonical Correlation Analysis (CCA). Currently spatial basis functions are used that implicitly perform an adaptive spatial filtering of the fMRI images significantly improving detection performance. Further improvement has been possible by introducing non-linear constrains and the present project aims at going even further in this direction. By including/developing tools for reconstruction of the brains cortical surface new and highly significant local spatial priors can be included in the fMRI data analysis.
Brain process identification
A rather visionary idea is to apply techniques used in system identification for the analysis and `control' of brain activity. Here an additional problem will be the very slow sampling rate. A time span of at least one minute will most likely be required to make any statistically reasonable estimate of the current state. This will constrain what is possible to achieve but the approach has the potential to open up a whole new field in brain research. The `state of mind' could be steered towards a goal state (activation pattern) by producing a sequence of stimuli that is dependent on the estimated activation pattern sequence. A dual view is that a person can be told to try to make the stimuli produced move towards a target stimulus by
will. In the future it may in this way be possible to analyze brain function in terms of brain state transition probability matrices.
The visualization of fMRI data is challenging, in view of the large datasets obtained and the need for simultaneous display of functional and anatomical information. New software for this purpose, combining volume rendering of both functional and anatomical data, has recently been implemented within CMIV as an add-on to the existing 3D Slicer software from Harvard [Hernell 2004]. As a continuation of this project, we are currently attempting to use Mixed Reality techniques [Belcher 2003], which fuse MRI data with direct visual information in an operating theater.
One field with prompt clinical application for fMRI is preoperative planning and intra-operative guidance in neurosurgery. It has e.g. been shown that fMRI is very valuable in mapping language lateralisation in candidates for epilepsy surgery. Another potentially clinically useful approach is the study of drug effects (e.g. sedative) on brain activity, a field where CMIV's fMRI group is pioneering.
Ogawa S, Lee TM, Kay AR, Tank DW.
Brain magnetic resonance imaging with contrast dependent on blood oxygenation.
Proc Natl Acad Sci U S A 1990;87(24):9868-72.
Friman O, Borga M, Lundberg P, Knutsson H.
Detection an Detrending in fMRI Data Analysis.
NeuroImage 2004, 22, pp 645-655.
Friman O, Borga M, Lundberg P, Knutsson H.
Adaptive analysis of fMRI data.
Implementation of volume rendering in 3D Slicer for functional MRI.
Belcher D, Billinghurst M.
Using Augmented Reality for Visualizing Complex Graphs in Three Dimensions.
The Second IEEE and ACM International Symposium on Mixed and Augmented Reality, Tokyo, Japan, 2003.