Abstract #3556
Multi-modal pattern recognition: an application to schizophrenia.
Orla M Doyle 1 , Brandon Whitcher 2,3 , Steven C.R. Williams 1 , Mitul A Mehta 1 , and Stephen M Lawrie 4
1
Dept of Neuroimaging, IoPPN, King's College
London, London, United Kingdom,
2
Clinical
& Translational Imaging, Pfizer, Cambridge, MA, United
States,
3
Dept
of Mathematics, Imperial College London, London, United
Kingdom,
4
Division
of Psychiatry, University of Edinburgh, Edinburgh,
United Kingdom
This is the first time that structural brain data, rCBF
and MRS data have been jointly assessed for
discriminating schizophrenia from controls. Twenty-four
patients diagnosed with schizophrenia and 24 age- and
gender-matched controls were included. An increase in
discriminative power was not observed on combining
modalities. rCBF was the most highly weighted modality.
Predictive probabilities (the probability of belonging
to the SCZ group) were not correlated with the level of
antipsychotic medication. These results imply that
perfusion imaging is a highly sensitive marker for
schizophrenia. Future work should assess the specificity
via differential diagnosis of psychiatric disorders.
This abstract and the presentation materials are available to members only;
a login is required.
Join Here
