Beware of white matter hyperintensities causing systematic errors in FreeSurfer gray matter segmentations!

Volumetric estimates of subcortical and cortical structures, extracted from T1-weighted MRIs, are widely used in many clinical and research applications. Here, we investigate the impact of the presence of white matter hyperintensities (WMHs) on FreeSurfer gray matter (GM) structure volumes and its possible bias on functional relationships. T1-weighted images from 1,077 participants (4,321 timepoints) from the Alzheimer's Disease Neuroimaging Initiative were processed with FreeSurfer version 6.0.0. WMHs were segmented using a previously validated algorithm on either T2-weighted or Fluid-attenuated inversion recovery images. Mixed-effects models were used to assess the relationships between overlapping WMHs and GM structure volumes and overall WMH burden, as well as to investigate whether such overlaps impact associations with age, diagnosis, and cognitive performance. Participants with higher WMH volumes had higher overlaps with GM volumes of bilateral caudate, cerebral cortex, putamen, thalamus, pallidum, and accumbens areas (p < .0001). When not corrected for WMHs, caudate volumes increased with age (p < .0001) and were not different between cognitively healthy individuals and age-matched probable Alzheimer's disease patients. After correcting for WMHs, caudate volumes decreased with age (p < .0001), and Alzheimer's disease patients had lower caudate volumes than cognitively healthy individuals (p < .01). Uncorrected caudate volume was not associated with ADAS13 scores, whereas corrected lower caudate volumes were significantly associated with poorer cognitive performance (p < .0001). Presence of WMHs leads to systematic inaccuracies in GM segmentations, particularly for the caudate, which can also change clinical associations. While specifically measured for the Freesurfer toolkit, this problem likely affects other algorithms.

Patch-wise brain age longitudinal reliability.

We recently introduced a patch-wise technique to estimate brain age from anatomical T1-weighted magnetic resonance imaging (T1w MRI) data. Here, we sought to assess its longitudinal reliability by leveraging a unique dataset of 99 longitudinal MRI scans from a single, cognitively healthy volunteer acquired over a period of 17 years (aged 29-46 years) at multiple sites. We built a robust patch-wise brain age estimation framework on the basis of 100 cognitively healthy individuals from the MindBoggle dataset (aged 19-61 years) using the Desikan-Killiany-Tourville atlas, then applied the model to the volunteer dataset. The results show a high prediction accuracy on the independent test set (R2 = .94, mean absolute error of 0.63 years) and no statistically significant difference between manufacturers, suggesting that the patch-wise technique has high reliability and can be used for longitudinal multi-centric studies.

Multivariate consistency of resting-state fMRI connectivity maps acquired on a single individual over 2.5 years, 13 sites and 3 vendors.

Studies using resting-state functional magnetic resonance imaging (rsfMRI) are increasingly collecting data at multiple sites in order to speed up recruitment or increase sample size. The main objective of this study was to assess the long-term consistency of rsfMRI connectivity maps derived at multiple sites and vendors using the Canadian Dementia Imaging Protocol (CDIP, www.cdip-pcid.ca). Nine to 10 min of functional BOLD images were acquired from an adult cognitively healthy volunteer scanned repeatedly at 13 Canadian sites on three scanner makes (General Electric, Philips and Siemens) over the course of 2.5 years. The consistency (spatial Pearson's correlation) of rsfMRI connectivity maps for seven canonical networks ranged from 0.3 to 0.8, with a negligible effect of time, but significant site and vendor effects. We noted systematic differences in data quality (i.e. head motion, number of useable time frames, temporal signal-to-noise ratio) across vendors, which may also confound some of these results, and could not be disentangled in this sample. We also pooled the long-term longitudinal data with a single-site, short-term (1 month) data sample acquired on 26 subjects (10 scans per subject), called HNU1. Using randomly selected pairs of scans from each subject, we quantified the ability of a data-driven unsupervised cluster analysis to match two scans of the same subjects. In this "fingerprinting" experiment, we found that scans from the Canadian subject (Csub) could be matched with high accuracy intra-site (>95% for some networks), but that the accuracy decreased substantially for scans drawn from different sites and vendors, even falling outside of the range of accuracies observed in HNU1. Overall, our results demonstrate good multivariate stability of rsfMRI measures over several years, but substantial impact of scanning site and vendors. How detrimental these effects are will depend on the application, yet our results demonstrate that new methods for harmonizing multisite analysis represent an important area for future work.

A novel patch-based procedure for estimating brain age across adulthood.

Aging is associated with structural alterations in many regions of the brain. Monitoring these changes contributes to increasing our understanding of the brain's morphological alterations across its lifespan, and could allow the identification of departures from canonical trajectories. Here, we introduce a novel and unique patch-based grading procedure for estimating a synthetic estimate of cortical aging in cognitively intact individuals. The cortical age metric is computed based on image similarity between an unknown (test) cortical label and known (training) cortical labels using machine learning algorithms. The proposed method was trained on a dataset of 100 cognitively intact individuals aged 19-61 years, within the 31 bilateral cortical labels of the Desikan-Killiany-Tourville parcellation, then tested on an independent test set of 78 cognitively intact individuals spanning a similar age range. The proposed patch-based framework yielded a R2 = 0.94, as well as a mean absolute error of 1.66 years, which compared favorably to the literature. These experimental results demonstrate that the proposed patch-based grading framework is a reliable and robust method to estimate brain age from image data, even with a limited training size.

The Canadian Dementia Imaging Protocol: Harmonizing National Cohorts.

BACKGROUND: Harmonized protocols to collect imaging data must be devised, employed, and maintained in multicentric studies to reduce interscanner variability in subsequent analyses. PURPOSE: To present a standardized protocol for multicentric research on dementia linked to neurodegeneration in aging, harmonized on all three major vendor platforms. The protocol includes a common procedure for qualification, quality control, and quality assurance and feasibility in large-scale studies. STUDY TYPE: Prospective. SUBJECTS: The study involved a geometric phantom, a single individual volunteer, and 143 cognitively healthy, mild cognitively impaired, and Alzheimer's disease participants in a large-scale, multicentric study. FIELD STRENGTH/SEQUENCES: MRI was perform with 3T scanners (GE, Philips, Siemens) and included 3D T1 w, PD/T2 w, T2* , T2 w-FLAIR, diffusion, and BOLD resting state acquisitions. ASSESSMENT: Measures included signal- and contrast-to-noise ratios (SNR and CNR, respectively), total brain volumes, and total scan time. STATISTICAL TESTS: SNR, CNR, and scan time were compared between scanner vendors using analysis of variance (ANOVA) and Tukey tests, while brain volumes were tested using linear mixed models. RESULTS: Geometric phantom T1 w SNR was significantly (P < 0.001) higher in Philips (mean: 71.4) than Siemens (29.5), while no significant difference was observed between vendors for T2 w (32.0 and 37.2, respectively, P = 0.243). Single individual volunteer T1 w CNR was higher in subcortical regions for Siemens (P < 0.001), while Philips had higher cortical CNR (P = 0.044). No significant difference in brain volumes was observed between vendors (P = 0.310/0.582/0.055). The average scan time was 41.0 minutes (SD: 2.8) and was not significantly different between sites (P = 0.071) and cognitive groups (P = 0.853). DATA CONCLUSION: The harmonized Canadian Dementia Imaging Protocol suits the needs of studies that need to ensure quality MRI data acquisition for the measurement of brain changes across adulthood, due to aging, neurodegeneration, and other etiologies. A detailed description, exam cards, and operators' manual are freely available at the following site: www.cdip-pcid.ca. LEVEL OF EVIDENCE: 2 Technical Efficacy: Stage 2 J. Magn. Reson. Imaging 2019;49:456-465.

Validation of an Magnetic Resonance Imaging Acquisition and Review Protocol for Alzheimer's Disease and Related Disorders.

PURPOSE: Magnetic resonance imaging (MRI) of the brain allows for the identification of structural lesions typical of Alzheimer's disease (AD), the main cause of dementia. However, to have a clinical impact, it is imperative that acquisition and reporting of this MRI-based evidence be standardized, ensuring the highest possible reliability and reproducibility. Our objective was to validate a systematic radiological MRI acquisition and review process in the context of AD. METHODS: We included 100 individuals with a suspicion of dementia due to AD for whom MRI were acquired using our proposed protocol of clinically achievable acquisitions and used a unified reading grid to gather semi-quantitative evidence guiding diagnostic. MRIs were read by 3 raters with different experience levels. Interrater reliability was measured using Cohen's kappa statistic. RESULTS: Interrater reliability average for lesions occupying space, hemorrhage, or ischemia, was respectively 0.754, 0.715, and 0.501. Average reliability of white matter hyperintensity burden (Fazekas), global cortical atrophy, and temporal lobe atrophy (Scheltens) scales was 0.687, 0.473, and 0.621 (right)/0.599 (left), respectively. The kappas for regional cortical atrophy (frontal, parietal, occipital, temporal, and posterior cingulum) varied from 0.281-0.678. The average MRI reading time varied between 1.43-5.22 minutes. CONCLUSIONS: The presence of space occupying lesions, hemorrhagic or ischemic phenomena, and radiological scales have a good interrater reproducibility in MRI. Coupled with standardized acquisitions, such a protocol should be used when evaluating possible dementias, especially those due to probable AD.

Freesurfer cortical normative data for adults using Desikan-Killiany-Tourville and ex vivo protocols.

We recently built normative data for FreeSurfer morphometric estimates of cortical regions using its default atlas parcellation (Desikan-Killiany or DK) according to individual and scanner characteristics. We aimed to produced similar normative values for Desikan-Killianny-Tourville (DKT) and ex vivo-based labeling protocols, as well as examine the differences between these three atlases. Surfaces, thicknesses, and volumes of cortical regions were produced using cross-sectional magnetic resonance scans from the same 2713 healthy individuals aged 18-94 years as used in the reported DK norms. Models predicting regional cortical estimates of each hemisphere were produced using age, sex, estimated intracranial volume (eTIV), scanner manufacturer and magnetic field strength (MFS) as predictors. The DKT and DK models generally included the same predictors and produced similar R2. Comparison between DK, DKT, ex vivo atlases normative cortical measures showed that the three protocols generally produced similar normative values.

Normative morphometric data for cerebral cortical areas over the lifetime of the adult human brain.

Proper normative data of anatomical measurements of cortical regions, allowing to quantify brain abnormalities, are lacking. We developed norms for regional cortical surface areas, thicknesses, and volumes based on cross-sectional MRI scans from 2713 healthy individuals aged 18 to 94 years using 23 samples provided by 21 independent research groups. The segmentation was conducted using FreeSurfer, a widely used and freely available automated segmentation software. Models predicting regional cortical estimates of each hemisphere were produced using age, sex, estimated total intracranial volume (eTIV), scanner manufacturer, magnetic field strength, and interactions as predictors. The explained variance for the left/right cortex was 76%/76% for surface area, 43%/42% for thickness, and 80%/80% for volume. The mean explained variance for all regions was 41% for surface areas, 27% for thicknesses, and 46% for volumes. Age, sex and eTIV predicted most of the explained variance for surface areas and volumes while age was the main predictors for thicknesses. Scanner characteristics generally predicted a limited amount of variance, but this effect was stronger for thicknesses than surface areas and volumes. For new individuals, estimates of their expected surface area, thickness and volume based on their characteristics and the scanner characteristics can be obtained using the derived formulas, as well as Z score effect sizes denoting the extent of the deviation from the normative sample. Models predicting normative values were validated in independent samples of healthy adults, showing satisfactory validation R2. Deviations from the normative sample were measured in individuals with mild Alzheimer's disease and schizophrenia and expected patterns of deviations were observed.

Utilization of healthcare services among elderly with cognitive impairment no dementia and influence of depression and anxiety: a longitudinal study.

OBJECTIVES: Little objective and nationally representative data are available concerning the influence of cognitive impairment no dementia (CIND) on utilization of healthcare services. The main objective was to compare the use of healthcare services over three years, between elders with current or incident CIND and those without CIND. A second objective was to evaluate the effect of depression and anxiety. METHODS: Cross-sectional and longitudinal data from a population-based survey of 2265 older adults living in Quebec (Canada) were used. CIND was identified using normative data for the Mini-Mental State Examination and was linked with medical records from public health insurance plan. Multinomial logistic regressions adjusted for relevant socio-demographic, social network and health-related confounders were conducted for each service. Interaction between CIND and depression/anxiety was also examined. MAIN RESULTS: Current CIND was a predictor of longer anxiolytic/sedative/hypnotic medication use. Incident CIND led to longer hospital stay. Depression raised the likelihood of frequenting geriatricians, psychiatrists or neurologists and emergency department, but lessened the likelihood of visiting general practitioners. The addition of the psychiatric conditions to the incident CIND did not increase the likelihood of consuming antidepressants, while the incident CIND cases without psychiatric conditions increased this likelihood. DISCUSSION: Compared to older adults without CIND, older adults with CIND have a distinct utilization of healthcare services. Multiple evaluations over many years may help to better understand the utilization of healthcare services in individuals with CIND. In the meantime, evaluations of these conditions at key moments could allow a more efficient use of health resources.

Normative data for subcortical regional volumes over the lifetime of the adult human brain.

Normative data for volumetric estimates of brain structures are necessary to adequately assess brain volume alterations in individuals with suspected neurological or psychiatric conditions. Although many studies have described age and sex effects in healthy individuals for brain morphometry assessed via magnetic resonance imaging, proper normative values allowing to quantify potential brain abnormalities are needed. We developed norms for volumetric estimates of subcortical brain regions based on cross-sectional magnetic resonance scans from 2790 healthy individuals aged 18 to 94years using 23 samples provided by 21 independent research groups. The segmentation was conducted using FreeSurfer, a widely used and freely available automated segmentation software. Models predicting subcortical regional volumes of each hemisphere were produced including age, sex, estimated total intracranial volume (eTIV), scanner manufacturer, magnetic field strength, and interactions as predictors. The mean explained variance by the models was 48%. For most regions, age, sex and eTIV predicted most of the explained variance while manufacturer, magnetic field strength and interactions predicted a limited amount. Estimates of the expected volumes of an individual based on its characteristics and the scanner characteristics can be obtained using derived formulas. For a new individual, significance test for volume abnormality, effect size and estimated percentage of the normative population with a smaller volume can be obtained. Normative values were validated in independent samples of healthy adults and in adults with Alzheimer's disease and schizophrenia.

When the Wedding March becomes sad: Semantic memory impairment for music in the semantic variant of primary progressive aphasia.

Music can induce particular emotions and activate semantic knowledge. In the semantic variant of primary progressive aphasia (svPPA), semantic memory is impaired as a result of anterior temporal lobe (ATL) atrophy. Semantics is responsible for the encoding and retrieval of factual knowledge about music, including associative and emotional attributes. In the present study, we report the performance of two individuals with svPPA in three experiments. NG with bilateral ATL atrophy and ND with atrophy largely restricted to the left ATL. Experiment 1 assessed the recognition of musical excerpts and both patients were unimpaired. Experiment 2 studied the emotions conveyed by music and only NG showed impaired performance. Experiment 3 tested the association of semantic concepts to musical excerpts and both patients were impaired. These results suggest that the right ATL seems essential for the recognition of emotions conveyed by music and that the left ATL is involved in binding music to semantics. They are in line with the notion that the ATLs are devoted to the binding of different modality-specific properties and suggest that they are also differentially involved in the processing of factual and emotional knowledge associated with music.

Short STAI-Y anxiety scales: validation and normative data for elderly subjects.

OBJECTIVES: The aim of this study was to develop short forms of the STAI-Y trait and state scales and associated norms suitable for the screening of anxiety in elderly populations. METHOD: This study was based on population-based cohorts of older persons from two epidemiological French studies that each included one subscale of the STAI-Y, i.e. state and trait anxiety scales. For both scales, the most discriminative items were retained and their factorial structure was examined using principal components analysis. Internal consistency (Cronbach's alpha) was estimated and cut-offs and norms were computed. RESULTS: A 10-item STAI-Y version produced scores similar to those obtained with the full form of the STAI-Y. The factorial structure of the shortened form is comparable to that of the full scales. Results showed good internal consistency (alpha coefficients were 0.92 and 0.85 for short STAI-Y state and trait scales, respectively). Moreover, both short STAI-Y state and trait scales correctly classified 88% of the participants using a cut-off point of 23. Norms for both short trait and state anxiety scales are provided according to age, gender, educational level and depressive symptoms. CONCLUSION: Both shortened scales have similar factorial structure and internal consistency to the longer scales and classify anxious/non-anxious elderly with acceptable accuracy. The shorter form is likely to be more acceptable to elderly persons through reduction of fatigue effects.

Using virtual reality to improve the efficacy of cognitive-behavioral therapy (CBT) in the treatment of late-life anxiety: preliminary recommendations for future research.

Cognitive-behavioral therapy (CBT) using traditional exposure techniques (i.e. imaginal and in vivo) seems less effective to treat anxiety in older adults than in younger ones. This is particularly true when imaginal exposure is used to confront the older patient to inaccessible (e.g. fear of flying) or less tangible/controllable anxiety triggers (e.g. fear of illness). Indeed, imaginal exposure may become less effective as the person gets older since normal aging is characterized by the decline in cognitive functions involved in the creation of vivid/detailed mental images. One way to circumvent this difficulty is to expose the older patient to a virtual environment that does not require the ability to imagine the frightening situation. In virtuo exposure has proven to be efficient to treat anxiety in working-age people. In virtuo exposure could be employed to improve the efficacy of CBT with exposure sessions in the treatment of late-life anxiety? The current paper explores this question and suggests new research avenues.

Gray matter characteristics associated with trait anxiety in older adults are moderated by depression.

BACKGROUND: Structural gray matter characteristics of anxiety remain unclear. The aim of this study was to assess the influence of current depressive symptoms and history of depression on the gray matter characteristics of trait anxiety. METHODS: Structural magnetic resonance imaging (MRI) data from 393 individuals aged 65 years or older were used. Regions of interest (ROIs) included the amygdala, anterior cingulate cortex (ACC), insula, orbitofrontal cortex (OFC), and temporal cortex. Trait anxiety was measured by the State-Trait Anxiety Inventory (STAI). Depression and depressive symptoms were measured using DSM-IV criteria and the Center for Epidemiological Studies Depression Scale (CESD). RESULTS: After adjustments for sociodemographics and health-related variables, anxiety had a significant influence on the gray matter characteristics in all cortical ROIs. First, in participants without depression antecedents, higher trait anxiety was associated with a larger cortical thickness in all cortical ROIs. Second, in participants with a previous history of depression, higher trait anxiety was associated with a smaller cortical thickness in all cortical ROIs. CONCLUSIONS: These results suggest that anxiety is related to cortical thickness differently in healthy older adults and in older adults with psychiatric antecedents. Anxiety associated with thinner cortical areas could reflect symptoms of a specific type of depression or a vulnerability to develop depression.

Subjective sleep characteristics associated with anxiety and depression in older adults: a population-based study.

OBJECTIVE: Sleep complaints are often associated with anxiety and depression, but the specific complaints related to each syndrome are poorly characterized, especially in older adults. The objective was to identify subjective sleep characteristics specific to anxiety and depression in this population. METHODS: A random sample of 2393 individuals aged 65 years or older was used. Anxiety and depression were categorized using DSM-V criteria for phobias, panic disorder, generalized anxiety disorder, unspecified anxiety disorder, major depressive episode, and depressive episode with insufficient symptoms. Subjective sleep characteristics were measured using the Pittsburgh Sleep Quality Index. Logistic regression models predicting anxiety or depression were used to determine the independent sleep characteristics associated with each syndrome adjusting for age, sex, education level, cognitive functioning, anxiolytic/sedative/hypnotic use, antidepressants use, subjective health, chronic diseases, cardiovascular conditions, and anxiety or depression (as appropriate). RESULTS: Nearly all Pittsburgh Sleep Quality Index subscales were significantly associated with anxiety, but these subscales shared variance and only sleep duration in hours, sleep disturbance score, and daytime functioning score were independently related to anxiety. Within these significant subscales, the main specific sleep complaints associated with anxiety were daytime sleepiness and sleep disturbances related to coughing/snoring, feeling cold, and bad dreams. The use of sleeping medication was the only specific sleep characteristic associated with depression. CONCLUSIONS: These results suggest that in older adults, symptoms of short sleep duration, daytime sleepiness and sleep disturbances are independently related to anxiety while the use of sleep medication is independently associated to depression.

Subjective sleep quality in non-demented older adults with and without cognitive impairment.

OBJECTIVE: Sleep problems are prevalent among older adults who are at risk of developing dementia. Until now, there have been relatively few studies investigating subjective sleep quality in these individuals. The first objective of this study was to compare seniors with cognitive impairment no dementia (CIND) and older adults without cognitive impairment (non-CIND) on several subjective sleep measures. The second objective was to verify whether sleep parameters associated with CIND differ between men and women. METHODS: The population sample consisted of 2287 French-speaking older adults from Québec (Canada) aged between 65 and 96 years. Participants were classified as CIND or non-CIND on the basis of their mini mental state examination score using sex, age, and education-stratified normative data. All participants completed the Pittsburgh Sleep Quality Index, and responses of CIND and non-CIND individuals were compared. A series of confounding variables (age, education, chronic diseases, mood disorders, anxiety disorders, and use of psychotropic drugs) were statistically controlled for. Student's t-tests were performed to compare characteristics of CIND and non-CIND individuals; data from male and female participants were analyzed separately. Moreover, the association between each sleep variable and CIND was measured by odds ratios based on logistic regression. RESULTS: On the whole, analyses revealed no significant association between subjective sleep parameters and CIND. Moreover, no difference was observed between men and women regarding subjective sleep quality. CONCLUSIONS: Overall, these results suggest that subjective measures of sleep do not allow differentiating cognitively impaired older individuals from those with normal cognition.

A Scoping Review of Alzheimers Disease Hypotheses: An Array of Uni- and Multi-Factorial Theories.

Background: There is a common agreement that Alzheimers disease (AD) is inherently complex; otherwise, a general disagreement remains on its etiological underpinning, with numerous alternative hypotheses having been proposed. Objective: To perform a scoping review of original manuscripts describing hypotheses and theories of AD published in the past decades. Results: We reviewed 131 original manuscripts that fulfilled our inclusion criteria out of more than 13,807 references extracted from open databases. Each entry was characterized as having a single or multifactorial focus and assigned to one of 15 theoretical groupings. Impact was tracked using open citation tools. Results: Three stages can be discerned in terms of hypotheses generation, with three quarter of studies proposing a hypothesis characterized as being single-focus. The most important theoretical groupings were the Amyloid group, followed by Metabolism and Mitochondrial dysfunction, then Infections and Cerebrovascular. Lately, evidence towards Genetics and especially Gut/Brain interactions came to the fore. Conclusions: When viewed together, these multi-faceted reports reinforce the notion that AD affects multiple sub-cellular, cellular, anatomical, and physiological systems at the same time but at varying degree between individuals. The challenge of providing a comprehensive view of all systems and their interactions remains, alongside ways to manage this inherent complexity.