Development of a new tool to correlate stroke outcome with infarct topography: A proof-of-concept study Thanh G. Phan a,b,c , Jian Chen a , Geoffrey Donnan b,c , Velandai Srikanth a , Amanda Wood a , David C. Reutens a,d, ⁎ a Southern Clinical School, Monash University, Clayton, Australia b National Stroke Research Institute, Heidelberg, Australia c The University of Melbourne, Austin and Repatriation Medical Centre, Heidelberg, Australia d The Centre for Advanced Imaging, The University of Queensland, Queensland 4072, St. Lucia, Australia abstract article info Article history: Received 4 September 2008 Revised 24 July 2009 Accepted 29 July 2009 Available online 4 August 2009 Keywords: Stroke outcome Prediction Digital Atlas Improving the ability to assess potential stroke deficit may aid the selection of patients most likely to benefit from acute stroke therapies. Methods based only on ‘at risk’ volumes or initial neurological condition do predict eventual outcome, but not perfectly. Given the close relationship between anatomy and function in the brain, we performed a proof-of-concept study to examine how well stroke outcome correlated with infarct location and extent. A prospective study of 60 patients with ischemic stroke (38 in the training set and 22 in the validation set), using an implementation of partial least squares with penalized logistic regression (PLS-PLR), was performed. The method yielded a model relating location of infarction (on a voxel- by-voxel basis) and neurological deficits. The area under the receiver operating characteristics curve (AUC) method was used to assess the accuracy of the method for predicting outcome. In the validation phase, this model indicated the presence of neglect (AUC 0.89), aphasia (AUC 0.79), right-arm motor deficit (0.94), and right-leg motor deficit (AUC 0.94) but less accurately indicated left-arm motor deficit (0.52) and left-leg motor deficit (0.69). The model indicated no to mild disability (Rankin ≤ 2) versus moderate to severe disability (Rankin N 2) with AUC 0.78. In this proof-of-concept study, we have demonstrated that stroke outcome correlates well with infarct location raising the possibility of accurate prediction of neurological deficit in the individual stroke patient using only information on infarct location and multivariate regression methods. © 2009 Elsevier Inc. All rights reserved. Introduction Stroke remains the second leading cause of disability and death worldwide (The National Institute of Neurological Disorders and Stroke rt-PA Stroke Study Group, 1995; Bonita, 1992). Inpatient hospitalization, rehabilitation and nursing home care also contribute significantly to the economic burden of stroke care (Dewey et al., 2001). Stroke clinicians and rehabilitation specialists are often faced with making difficult decisions regarding long-term prognosis and potential rate of recovery for patients, their suitability for rehabilitation, and consequently, optimal allocation of resources. These decisions are often made early after stroke, usually within the first ten days and can have a significant impact on the overall outcome for an individual patient and their carers. Such decisions are based upon a number of clinical factors in a relatively arbitrary way and at present, the ability to estimate eventual neurological deficit and disability in individual patients is limited. Studies using the summed National Institute of Health Stroke Scale (NIHSS) to predict disability have reported different NIHSS thresholds for poor outcome making it difficult for the clinician to be confident about basing prognosis on this measure. With some studies, other variables such as infarct volume were included in the predictive model so that the independent predictive value of the NIHSS could not be assessed. Some studies have reported poor outcomes in patients with NIHSS exceeding 13–15 (Adams et al., 1999; Baird et al., 2001; Muir et al., 1996; Schlegel et al., 2003). One reason why the summed NIHSS does not provide consistent prediction is that the various neurological domains of the NIHSS are treated as if they have equal weighting for predicting disability. When used in this fashion, the summed NIHSS is weighted greatly in favor of the motor deficit. Yet, in addition to motor deficit, higher cortical neurological deficits also contribute to stroke disability (Gresham et al., 1979). A multivariate model which used only clinical features such as limb weakness, continence, aphasia and prestroke disability has been shown to predict long-term disability in only 49% of the patients (Tilling et al., 2001). There is a good biological basis for assuming that greater infarct- related damage leads to poorer outcome, although this relationship is NeuroImage 49 (2010) 127–133 ⁎ Corresponding author. The Centre for Advanced Imaging, The University of Queensland, Queensland 4072, St. Lucia, Australia. Fax: +617 3346 6301. E-mail address: d.reutens@uq.edu.au (D.C. Reutens). 1053-8119/$ – see front matter © 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.neuroimage.2009.07.067 Contents lists available at ScienceDirect NeuroImage journal homepage: www.elsevier.com/locate/ynimg