Matthew J. Clarkson

Head size, age and gender adjustment in MRI studies: a necessary nuisance?

Imaging studies of cerebral volumes often adjust for factors such as age that may confound between-subject comparisons. However the use of nuisance covariates in imaging studies is inconsistent, which can make interpreting results across studies difficult. Using magnetic resonance images of 78 healthy controls we assessed the effects of age, gender, head size and scanner upgrade on region of interest (ROI) volumetry, cortical thickness and voxel-based morphometric (VBM) measures. We found numerous significant associations between these variables and volumetric measures: cerebral volumes and cortical thicknesses decreased with increasing age, men had larger volumes and smaller thicknesses than women, and increasing head size was associated with larger volumes. The relationships between most ROIs and head size volumes were non-linear. With age, gender, head size and upgrade in one model we found that volumes and thicknesses decreased with increasing age, women had larger volumes than men (VBM, whole-brain and white matter volumes), increasing head size was associated with larger volumes but not cortical thickness, and scanner upgrade had an effect on thickness and some volume measures. The effects of gender on cortical thickness when adjusting for head size, age and upgrade showed some non-significant effect (women>men), whereas the independent effect of head size showed little pattern. We conclude that age and head size should be considered in ROI volume studies, age, gender and upgrade should be considered for cortical thickness studies and all variables require consideration for VBM analyses. Division of all volumes by head size is unlikely to be adequate owing to their non-proportional relationship.

Automated cross-sectional and longitudinal hippocampal volume measurement in mild cognitive impairment and Alzheimer's disease

Volume and change in volume of the hippocampus are both important markers of Alzheimers disease (AD). Delineation of the structure on MRI is time-consuming and therefore reliable automated methods are required. We describe an improvement (multiple-atlas propagation and segmentation (MAPS)) to our template library-based segmentation technique. The improved technique uses non-linear registration of the best-matched templates from our manually segmented library to generate multiple segmentations and combines them using the simultaneous truth and performance level estimation (STAPLE) algorithm. Change in volume over 12months (MAPS-HBSI) was measured by applying the boundary shift integral using MAPS regions. Methods were developed and validated against manual measures using subsets from Alzheimers Disease Neuroimaging Initiative (ADNI). The best method was applied to 682 ADNI subjects, at baseline and 12-month follow-up, enabling assessment of volumes and atrophy rates in control, mild cognitive impairment (MCI) and AD groups, and within MCI subgroups classified by subsequent clinical outcome. We compared our measures with those generated by Surgical Navigation Technologies (SNT) available from ADNI. The accuracy of our volumes was one of the highest reported (mean(SD) Jaccard Index 0.80(0.04) (N=30)). Both MAPS baseline volume and MAPS-HBSI atrophy rate distinguished between control, MCI and AD groups. Comparing MCI subgroups (reverters, stable and converters): volumes were lower and rates higher in converters compared with stable and reverter groups (p< or =0.03). MAPS-HBSI required the lowest sample sizes (78 subjects) for a hypothetical trial. In conclusion, the MAPS and MAPS-HBSI methods give accurate and reliable volumes and atrophy rates across the clinical spectrum from healthy aging to AD.

Clinical and neuroanatomical signatures of tissue pathology in frontotemporal lobar degeneration

Relating clinical symptoms to neuroanatomical profiles of brain damage and ultimately to tissue pathology is a key challenge in the field of neurodegenerative disease and particularly relevant to the heterogeneous disorders that comprise the frontotemporal lobar degeneration spectrum. Here we present a retrospective analysis of clinical, neuropsychological and neuroimaging (volumetric and voxel-based morphometric) features in a pathologically ascertained cohort of 95 cases of frontotemporal lobar degeneration classified according to contemporary neuropathological criteria. Forty-eight cases (51%) had TDP-43 pathology, 42 (44%) had tau pathology and five (5%) had fused-in-sarcoma pathology. Certain relatively specific clinicopathological associations were identified. Semantic dementia was predominantly associated with TDP-43 type C pathology; frontotemporal dementia and motoneuron disease with TDP-43 type B pathology; young-onset behavioural variant frontotemporal dementia with FUS pathology; and the progressive supranuclear palsy syndrome with progressive supranuclear palsy pathology. Progressive non-fluent aphasia was most commonly associated with tau pathology. However, the most common clinical syndrome (behavioural variant frontotemporal dementia) was pathologically heterogeneous; while pathologically proven Picks disease and corticobasal degeneration were clinically heterogeneous, and TDP-43 type A pathology was associated with similar clinical features in cases with and without progranulin mutations. Volumetric magnetic resonance imaging, voxel-based morphometry and cluster analyses of the pathological groups here suggested a neuroanatomical framework underpinning this clinical and pathological diversity. Frontotemporal lobar degeneration-associated pathologies segregated based on their cerebral atrophy profiles, according to the following scheme: asymmetric, relatively localized (predominantly temporal lobe) atrophy (TDP-43 type C); relatively symmetric, relatively localized (predominantly temporal lobe) atrophy (microtubule-associated protein tau mutations); strongly asymmetric, distributed atrophy (Picks disease); relatively symmetric, predominantly extratemporal atrophy (corticobasal degeneration, fused-in-sarcoma pathology). TDP-43 type A pathology was associated with substantial individual variation; however, within this group progranulin mutations were associated with strongly asymmetric, distributed hemispheric atrophy. We interpret the findings in terms of emerging network models of neurodegenerative disease: the neuroanatomical specificity of particular frontotemporal lobar degeneration pathologies may depend on an interaction of disease-specific and network-specific factors.

Progressive logopenic/phonological aphasia: Erosion of the language network

The primary progressive aphasias (PPA) are paradigmatic disorders of language network breakdown associated with focal degeneration of the left cerebral hemisphere. Here we addressed brain correlates of PPA in a detailed neuroanatomical analysis of the third canonical syndrome of PPA, logopenic/phonological aphasia (LPA), in relation to the more widely studied clinico-anatomical syndromes of semantic dementia (SD) and progressive nonfluent aphasia (PNFA). 32 PPA patients (9 SD, 14 PNFA, 9 LPA) and 18 cognitively normal controls had volumetric brain MRI with regional volumetry, cortical thickness, grey and white matter voxel-based morphometry analyses. Five of nine patients with LPA had cerebrospinal fluid biomarkers consistent with Alzheimer (AD) pathology (AD-PPA) and 2/9 patients had progranulin (GRN) mutations (GRN-PPA). The LPA group had tissue loss in a widespread left hemisphere network. Compared with PNFA and SD, the LPA group had more extensive involvement of grey matter in posterior temporal and parietal cortices and long association white matter tracts. Overlapping but distinct networks were involved in the AD-PPA and GRN-PPA subgroups, with more anterior temporal lobe involvement in GRN-PPA. The importance of these findings is threefold: firstly, the clinico-anatomical entity of LPA has a profile of brain damage that is complementary to the network-based disorders of SD and PNFA; secondly, the core phonological processing deficit in LPA is likely to arise from temporo-parietal junction damage but disease spread occurs through the dorsal language network (and in GRN-PPA, also the ventral language network); and finally, GRN mutations provide a specific molecular substrate for language network dysfunction.

Robust atrophy rate measurement in Alzheimer's disease using multi-site serial MRI: Tissue-specific intensity normalization and parameter selection

We describe an improved method of measuring brain atrophy rates from serial MRI for multi-site imaging studies of Alzheimers disease (AD). The method (referred to as KN-BSI) improves an existing brain atrophy measurement technique-the boundary shift integral (classic-BSI), by performing tissue-specific intensity normalization and parameter selection. We applied KN-BSI to measure brain atrophy rates of 200 normal and 141 AD subjects using baseline and 1-year MRI scans downloaded from the Alzheimers Disease Neuroimaging Initiative database. Baseline and repeat images were reviewed as pairs by expert raters and given quality scores. Including all image pairs, regardless of quality score, mean KN-BSI atrophy rates were 0.09% higher (95% CI 0.03% to 0.16%, p=0.007) than classic-BSI rates in controls and 0.07% higher (-0.01% to 0.16%, p=0.07) higher in ADs. The SD of the KN-BSI rates was 22% lower (15% to 29%, p<0.001) in controls and 13% lower (6% to 20%, p=0.001) in ADs, compared to classic-BSI. Using these results, the estimated sample size (needed per treatment arm) for a hypothetical trial of a treatment for AD (80% power, 5% significance to detect a 25% reduction in atrophy rate) would be reduced from 120 to 81 (a 32% reduction, 95% CI=18% to 45%, p<0.001) when using KN-BSI instead of classic-BSI. We concluded that KN-BSI offers more robust brain atrophy measurement than classic-BSI and substantially reduces sample sizes needed in clinical trials.

A comparison of voxel and surface based cortical thickness estimation methods

Cortical thickness estimation performed in-vivo via magnetic resonance imaging is an important technique for the diagnosis and understanding of the progression of neurodegenerative diseases. Currently, two different computational paradigms exist, with methods generally classified as either surface or voxel-based. This paper provides a much needed comparison of the surface-based method FreeSurfer and two voxel-based methods using clinical data. We test the effects of computing regional statistics using two different atlases and demonstrate that this makes a significant difference to the cortical thickness results. We assess reproducibility, and show that FreeSurfer has a regional standard deviation of thickness difference on same day scans that is significantly lower than either a Laplacian or Registration based method and discuss the trade off between reproducibility and segmentation accuracy caused by bending energy constraints. We demonstrate that voxel-based methods can detect similar patterns of group-wise differences as well as FreeSurfer in typical applications such as producing group-wise maps of statistically significant thickness change, but that regional statistics can vary between methods. We use a Support Vector Machine to classify patients against controls and did not find statistically significantly different results with voxel based methods compared to FreeSurfer. Finally we assessed longitudinal performance and concluded that currently FreeSurfer provides the most plausible measure of change over time, with further work required for voxel based methods.

LoAd: A locally adaptive cortical segmentation algorithm

Thickness measurements of the cerebral cortex can aid diagnosis and provide valuable information about the temporal evolution of diseases such as Alzheimers, Huntingtons, and schizophrenia. Methods that measure the thickness of the cerebral cortex from in-vivo magnetic resonance (MR) images rely on an accurate segmentation of the MR data. However, segmenting the cortex in a robust and accurate way still poses a challenge due to the presence of noise, intensity non-uniformity, partial volume effects, the limited resolution of MRI and the highly convoluted shape of the cortical folds. Beginning with a well-established probabilistic segmentation model with anatomical tissue priors, we propose three post-processing refinements: a novel modification of the prior information to reduce segmentation bias; introduction of explicit partial volume classes; and a locally varying MRF-based model for enhancement of sulci and gyri. Experiments performed on a new digital phantom, on BrainWeb data and on data from the Alzheimers Disease Neuroimaging Initiative (ADNI) show statistically significant improvements in Dice scores and PV estimation (p<10(-3)) and also increased thickness estimation accuracy when compared to three well established techniques.

Volume and Shape Preservation of Enhancing Lesions when Applying Non-rigid Registration to a Time Series of Contrast Enhancing MR Breast Images

In this paper we show first that a non-rigid registration algorithm used to register time-series MR images of the breast, can result in significant volume changes in the region of the enhanced lesion. Since this is physically implausible, given the short duration of the MR time series acquisition, the non-rigid registration algorithm was extended to allow the incorporation of rigid regions. In this way the registration is done in two stages. The enhanced lesions are first detected using the non-rigid registration algorithm in its original form. Secondly, the region of the enhanced lesion is set to be rigid and the new algorithm is applied to integrate this rigid region into the existing registration. By definition, volume and shape will be preserved in this rigid region. Preliminary results of applying this algorithm to 15 datasets are described.

Reduced cortical thickness in the posterior cingulate gyrus is characteristic of both typical and atypical Alzheimer's disease.

Alzheimers disease (AD) and frontotemporal lobar degeneration (FTLD) can be difficult to differentiate clinically due to overlapping symptoms. Subject classification in research studies is often based on clinical rather than pathological criteria which may mean some subjects are misdiagnosed and misclassified. Recently, methods measuring cortical thickness using magnetic resonance imaging have been suggested to be effective in differentiating between clinically-defined AD and frontotemporal dementia (FTD) in addition to showing disease-related patterns of atrophy. In this study we used FreeSurfer, a freely-available and automated software tool, to measure cortical thickness in 28 pathologically-confirmed AD patients, of which 11 had a typical amnestic presentation and 17 an atypical presentation during life, 23 pathologically-confirmed FTLD subjects, and 25 healthy controls. Patients with AD pathology, irrespective of clinical diagnosis, showed reduced cortical thickness bilaterally in the medial temporal lobe, posterior cingulate gyrus, precuneus, posterior parietal lobe, and frontal pole compared with controls. We further showed that lower cortical thickness in the posterior cingulate gyrus, parietal lobe, and frontal pole is suggestive of AD pathology in patients with behavioral or language deficits. In contrast, lower cortical thickness in the anterior temporal lobe and frontal lobe is indicative of the presence of FTLD pathology in patients with a clinical presentation of FTD. Reduced cortical thickness in the posterior cingulate gyrus is characteristic of AD pathology in patients with typical and atypical clinical presentations of AD, and may assist a clinical distinction of AD pathology from FTLD pathology.

Stereo Augmented Reality in the Surgical Microscope

This paper describes the MAGI (microscope-assisted guided interventions) augmented-reality system, which allows surgeons to view virtual features segmented from preoperative radiological images accurately overlaid in stereo in the optical path of a surgical microscope. The aim of the system is to enable the surgeon to see in the correct 3-D position the structures that are beneath the physical surface. The technical challenges involved are calibration, segmentation, registration, tracking, and visualization. This paper details our solutions to these problems. As it is difficult to make reliable quantitative assessments of the accuracy of augmented-reality systems, results are presented from a numerical simulation, and these show that the system has a theoretical overlay accuracy of better than 1 mm at the focal plane of the microscope. Implementations of the system have been tested on volunteers, phantoms, and seven patients in the operating room. Observations are consistent with this accuracy prediction.