UNCORRECTED PROOF 1 Cortical thickness estimation in longitudinal stroke studies: A 2 comparison of 3 measurement methods 3 Qi Li a, * , Heath Pardoe a,b,c , Renee Lichter a,b , Emilio Werden a,b , Audrey Raffelt a , 4 Toby Cumming a,b , Amy Brodtmann a,b 5 a The Florey Institute of Neuroscience and Mental Health, Melbourne, Australia 6 b University of Melbourne, Melbourne, Australia 7 c Comprehensive Epilepsy Center, New York University, New York, USA Q3 abstract 8 article info 9 Article history: 10 Received 9 February 2014 11 Received in revised form 17 August 2014 12 Accepted 21 August 2014 13 Available online xxxx 14 Keywords: 15 Cortical thickness 16 Laplacian 17 DiRecT 18 FreeSurfer 19 Comparison 20 Stroke 21 Surface-based 22 Voxel-based 23 There is considerable controversy about the causes of cognitive decline after stroke, with evidence for both the 24 absence and coexistence of Alzheimer pathology. A reduction in cortical thickness has been shown to be an im- 25 portant biomarker for the progression of many neurodegenerative diseases, including Alzheimer3s disease (AD). 26 However, brain volume changes following stroke are not well described. Cortical thickness estimation presents 27 an ideal way to detect regional and global post-stroke brain atrophy. In this study, we imaged a group of patients 28 in the first month after stroke and at 3 months. We compared three methods of estimating cortical thickness on 29 unmasked images: one surface-based (FreeSurfer) and two voxel-based methods (a Laplacian method and a reg- 30 istration method, DiRecT). We used three benchmarks for our analyses: accuracy of segmentation (especially 31 peri-lesional performance), reproducibility, and biological validity. We found important differences between 32 these methods in cortical thickness values and performance in high curvature areas and peri-lesional regions, 33 but similar reproducibility metrics. FreeSurfer had less reliance on manual boundary correction than the other 34 two methods, while reproducibility washighest in the Laplacian method. A discussion of the caveats for each 35 method and recommendations for use in a stroke population is included. We conclude that both surface- and 36 voxel-based methods are valid for estimating cortical thickness in stroke populations. 37 © 2014 Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license 38 (http://creativecommons.org/licenses/by-nc-nd/3.0/). 39 40 41 42 43 1. Introduction 44 Cortical thickness is a key biomarker in the diagnosis and prognosti- 45 cation of neurodegenerative disease, used extensively in diseases such 46 as Alzheimer3s disease (AD) and the frontotemporal dementias (FTD) 47 (Boccardi et al., 2011; Duering et al., 2012; Hartikainen et al., 2012; 48 Richards et al., 2009). Cortical thinning in critical brain regions has 49 been shown to correlate with disease severity and progression in neuro- 50 degenerative disease. For example, reduction in gray matter volume in 51 the hippocampi and orbitofrontal cortices is associated with conversion 52 from mild cognitive impairment to dementia, and correlated with a sub- 53 sequent diagnosis of AD (Hartikainen et al., 2012; Lerch and Evans, 54 2005). These methods are now being utilized after stroke, as interest 55 in imaging correlates with post-stroke cognitive decline increases 56 (Brodtmann et al., 2012). 57 Cognitive impairment is common after stroke, with at least one in 58 three patients developing dementia (Pendlebury and Rothwell, 2009). 59 Pathological studies have revealed that many patients with stroke 60 may have associated AD pathology, including β-amyloid plaques and 61 neurofibrillary tangles (Hesse et al., 2000). Cortical thickness changes 62 have been demonstrated after stroke, both increments and decrements. 63 We have demonstrated significant increases which may represent 64 compensatory mechanisms, mostly in the contralesional hemisphere 65 (Brodtmann et al., 2012). In contrast, decreases in ipsilesional hemi- 66 sphere have been found with motor recovery, and cortical thinning 67 overlying subcortical strokes has also been demonstrated (Duering 68 et al., 2012). 69 1.1. The challenge of cortical thickness estimation in stroke 70 Few researchers have examined changes in cortical thickness over 71 time in stroke populations. Most acute stroke trials do not include a 72 high resolution isotropic T1 scan (An et al., 2011; Lee et al., 2010; 73 Rohrer et al., 2009), often considered the workhorse of brain volume es- 74 timation. In addition, cortical thickness estimation in stroke patients 75 brings a unique set of potential difficulties. By definition, all individuals 76 with stroke will have a destructive brain lesion, sometimes large, which 77 can distort surrounding brain structures acutely, secondary to edema, NeuroImage: Clinical xxx (2014) xxx–xxx * Corresponding author at: The Florey Institute of Neuroscience and Mental Health, Austin Campus, 245 Burgundy Street, Heidelberg, Melbourne 3084, Australia. E-mail address: qi.li@florey.edu.au (Q. Li). YNICL-00338; No. of pages: 10; 4C: http://dx.doi.org/10.1016/j.nicl.2014.08.017 2213-1582/© 2014 Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/). Contents lists available at ScienceDirect NeuroImage: Clinical journal homepage: www.elsevier.com/locate/ynicl Please cite this article as: Li, Q., et al., Cortical thickness estimation in longitudinal stroke studies: A comparison of 3 measurement methods, NeuroImage: Clinical (2014), http://dx.doi.org/10.1016/j.nicl.2014.08.017