You are in: Home » Research at the Unit » Methods Research and Infrastructure
CBU Methods Research
The Methods Research and Infrastructure Group (Methods Group) coordinates research which develops and implements methods in support of the scientific research programmes at the CBU. The expertise of the group is in four broad areas: the Physics of Magnetic Resonance Imaging; Pre-processing of MR data; Analysis and Modelling of MR data; Statistical modelling.
Current Research Activity
MRI Physics, Data Processing and Analysis
I develop new methods for fMRI analysis, and with Stefan Hetzer and Marta Correia, new techniques for acquisition. For more information, take a look at the work of my lab. Or, see our tools for neuroimaging analysis at Cambridge Neuroimaging and the automatic analysis wiki
MEG Data Analysis
We develop and evaluate methods for MEG analysis, in particular for source estimation, statistics and multi-variate analysis techniques. For more information please visit our MEG Analysis Wiki Pages, or contact
MRI Data Processing and Analysis
I joined the Methods Group in April 2008 as a Visiting Student. The focus of my work will be developing models, methods and tools towards structural segmentations of the cortex based on white matter tractography derived from MR diffusion imaging.
Statistics
|
|
Cluster analysis of neural coding patterns in a population of frontal neurons |
|
|
Shared and individually differentiated parsing of dance sequences |
 |
Event related covariates and semi-parametric modulation |
- I apply, explain and make available, for general use, a broad range of statistical techniques primarily using R, SPSS and EXCEL for addressing problems which crop up in neurological research at the CBSU (See the statistical resources below). In particular I have taught technical aspects of SPSS, namely macros and the output management system, at various workshops. Recent statistical applications include the creation of power calculators for determining sample size for grant proposal submissions, using structural equation models to assess factor structures in studies involving emotional and attentional data and computation of memory test norms. I have also spoken at the SPSS annual users meeting in York, which I now organise, three times on analyses of MRC data using respectively cosinor analysis of attentional data, survival analysis of cognitive data and repeated measures in SPSS in the presence of a varying covariate (various). I have also contributed a primer chapter on statistical aspects of survival analysis covered in the above talk in A handbook of research methods for clinical and health psychology, edited by Miles, J. and Gilbert, P. (2005) OUP.
- Upcoming work involves using generalised estimating equations to compare rates of psychiatric disorders in cancer patients and their siblings adjusting for effects of IQ and more work on aspects of power analyses including the obtaining of confidence intervals for effect sizes.