Correlation between cognitive function and the association fibers in patients with Alzheimer's disease using diffusion tensor imaging

Diffusion kurtosis imaging of brain white matter alteration in patients with coronary artery disease based on the TBSS method

Object

The aim of our study was to examine the alterations in microstructure in patients with coronary artery disease (CAD) and cognitive impairment (CI) using diffusion kurtosis imaging (DKI). Additionally, we aimed to investigate the potential correlation between DKI parameters and cognitive function.

Materials and methods

A total of 28 CAD patients and 30 healthy controls (HC) were prospectively enrolled in our study. All participants underwent routine and diffusion sequences of head imaging. DKE software was utilized to generate various diffusion kurtosis imaging parameters (DKI), including kurtosis fractional anisotropy (KFA), mean kurtosis (MK), axial kurtosis (AK), radial kurtosis (RK), fractional anisotropy (FA), and mean diffusivity (MD). Nonparametric tests were conducted using tract-based spatial statistics (TBSS) to compare the parameter values between the two groups. The parameter values of the significantly different fiber tracts were extracted and correlated with the Mini-Mental State Examination (MMSE) and Montreal Cognitive Assessment (MoCA) scores.

Results

Compared to the HC group, patients with coronary artery disease exhibited significant reductions in FA values in the bilateral Superior corona radiata, bilateral Anterior corona radiata, bilateral Posterior corona radiata, corpus callosum, left Posterior thalamic radiation, right Posterior limb of internal capsule, Anterior limb of internal capsule, and Cerebral peduncle, as well as in the left Superior longitudinal fasciculus. Additionally, KFA values decreased in the bilateral Anterior corona radiata, bilateral Anterior limb of internal capsule, and Genu of the corpus callosum. The MK values decreased in the right Posterior corona radiata, Retrolenticular part of the internal capsule, Posterior thalamic radiation (including optic radiation), Superior longitudinal fasciculus, and left Posterior thalamic radiation (including optic radiation). Moreover, the RK values decreased in the bilateral Retrolenticular part of the internal capsule, right Posterior thalamic radiation (including optic radiation), and Superior longitudinal fasciculus, as well as in the left Superior longitudinal fasciculus and Posterior thalamic radiation (including optic radiation) (p < 0.01, TFCE corrected), while no significant differences were observed in other parameter values (p > 0.01, TFCE corrected). The FA values of the right posterior limb of the internal capsule (r = 0.610, p = 0.001) and the right cerebral peduncle (r = 0.622, p < 0.001) were positively correlated with MMSE scores. Additionally, a significant correlation between kurtosis and diffusion coefficient parameters (FA and KFA) was observed.

Conclusion

CAD patients showed radial shrinkage and complexity of brain white matter microstructure. Whole-brain white matter analysis based on TBSS DKI can objectively reflect the characteristics of white matter damage in CAD patients, providing a basis for the auxiliary diagnosis of CAD with CI.

Analysis of Brain Structural Connectivity Networks and White Matter Integrity in Patients With Mild Cognitive Impairment

White matter integrity and structural connectivity may be altered in mild cognitive impairment (MCI), and these changes may closely reflect decline in specific cognitive domains. Multi-shell diffusion data in healthy control (HC, n = 31) and mild cognitive impairment (MCI, n = 19) cohorts were downloaded from the ADNI3 database. The data were analyzed using an advanced approach to assess both white matter microstructural integrity and structural connectivity. Compared with HC, lower intracellular compartment (IC) and higher isotropic (ISO) values were found in MCI. Additionally, significant correlations were found between IC and Montreal Cognitive Assessment (MoCA) scores in the MCI cohort. Network analysis detected structural connectivity differences between the two groups, with lower connectivity in MCI. Additionally, significant differences between HC and MCI were observed for global network efficiency. Our results demonstrate the potential of advanced diffusion MRI biomarkers for understanding brain changes in MCI.

Association between microstructural white matter abnormalities and cognitive functioning in patients with type 2 diabetes mellitus: a diffusion tensor imaging study

Background

Diffusion tensor imaging (DTI) technique is important for exploring more sensitive imaging-based biomarkers in prevention and early treatment of cognitive dysfunction induced by type 2 diabetes mellitus (DM).

Objectives

To predict early cognitive dysfunction and detection of microstructural white matter changes in patients with type 2 DM by diffusion tensor imaging.

Patients and methods

A case-control study included thirty patients aged ≥ 18 years old of both sexes with type 2 DM and 30 controls. All subjects underwent to Montreal Cognitive Assessment (MoCA) “Arabic version”: to detect mild cognitive impairment (MCI) and diffusion tensor imaging study (DTI).

Results

Mild cognitive impairment is related to type 2 DM (56.7% of diabetic group), reduced fractional anisotropy (FA) values, and elevated mean diffusivity (MD) values were related to cognitive impairment evaluated through Montreal Cognitive Assessment (MoCA) in patients with type 2 DM.

Conclusion

The integrity of the white matter measured using DTI vary in MCI diabetics compared with non-MCI diabetics. Such changes have major implications on the cognitive function.

Corpus callosum integrity loss predicts cognitive impairment in Leukoaraiosis

Objective

To investigate regional white matter fibers loss in Leukoaraiosis (LA) and its relationship with cognitive impairments.

Methods

Fifty‐six participants with LA and 38 healthy controls underwent clinical evaluations and MR scans. Participants with LA were classified as cognitively normal (LA‐NC, n = 18), vascular cognitive impairment of none dementia (LA‐VCIND, n = 24), and vascular dementia (LA‐VaD, n = 14) by Mini‐Mental State Examination and Clinical Dementia Rating. Cognitive domains including visual‐spatial, naming, attention, language, abstraction, memory, and orientation were assessed. With the use of Tract‐based spatial statistics, mean fractional anisotropy (FA) of major white matter fiber tracts were compared between LA and controls and among LA groups with varying levels of cognitive impairments. Regression analyses were performed to evaluate relationships between FA values and cognitive performance.

Results

Participants showed significant FA reduction in the corpus callosum (CC), bilateral corona radiata, anterior limb of the internal capsule, external capsule, posterior thalamic radiation, and superior longitudinal fasciculus compared to controls and across LA groups. The LA‐VaD group showed consistent damage in the body and genu of CC compared to the LA‐NC and LA‐VCIND groups. A positive correlation between visual‐spatial and FA reduction in right anterior corona radiates in LA‐VCIND and body of CC in LA‐ VaD.

Interpretation

We found regional fiber loss in the CC across the cognitive spectrum in patients with LA and correlations between FA and visuospatial impairment in the anterior corona radiata in patients with LA‐VCIND and in the body of CC in patients with LA‐VaD.

White Matter Changes on Diffusion Tensor Imaging in the FINGER Randomized Controlled Trial

Background:

Early pathological changes in white matter microstructure can be studied using the diffusion tensor imaging (DTI). It is not only important to study these subtle pathological changes leading to cognitive decline, but also to ascertain how an intervention would impact the white matter microstructure and cognition in persons at-risk of dementia.

Objectives:

To study the impact of a multidomain lifestyle intervention on white matter and cognitive changes during the 2-year Finnish Geriatric Intervention Study to prevent Cognitive Impairment and Disability (FINGER), a randomized controlled trial in at-risk older individuals (age 60–77 years) from the general population.

Methods:

This exploratory study consisted of a subsample of 60 FINGER participants. Participants were randomized to either a multidomain intervention (diet, exercise, cognitive training, and vascular risk management, n = 34) or control group (general health advice, n = 26). All underwent baseline and 2-year brain DTI. Changes in fractional anisotropy (FA), diffusivity along domain (F1) and non-domain (F2) diffusion orientations, mean diffusivity (MD), axial diffusivity (AxD), radial diffusivity (RD), and their correlations with cognitive changes during the 2-year multidomain intervention were analyzed.

Results:

FA decreased, and cognition improved more in the intervention group compared to the control group (p < 0.05), with no significant intergroup differences for changes in F1, F2, MD, AxD, or RD. The cognitive changes were significantly positively related to FA change, and negatively related to RD change in the control group, but not in the intervention group.

Conclusion:

The 2-year multidomain FINGER intervention may modulate white matter microstructural alterations.

The Montreal Cognitive Assessment (MoCA) and Brain Structure

Abstract. MoCA is a short cognitive screening tool. We examined the relationship of MoCA performance to white matter integrity, gray matter volume, and surface-based measurements at normal aging in a study in which older and younger cognitively unaffected subjects participated. The sample was split according to MoCA performance, and the data were analyzed using a general linear model (Age × MoCA). We found effects in the expected direction for all methods. The main effects on age and performance as well as interactions occurred for regions associated with aging, pathological and nonpathological. Older low-performing subjects showed structural deficits compared to older high-performing subjects. Therefore, the global index of cognitive status reflects relevant features of the brain structure.

Damage to the Frontal Aslant Tract Accounts for Visuo-Constructive Deficits in Alzheimer's Disease

The frontal aslant tract (FAT) has been described as a bundle connecting the Broca's area to the supplementary motor area (SMA) and the pre-SMA in both hemispheres. The functional properties of this tract and its role in degenerative dementia, such as Alzheimer's disease (AD), still need to be fully clarified. The aim of this study was to explore the microstructural integrity of the FAT in patients with AD and its potential relationship with cognitive functioning. Twenty-three patients with AD and 25 healthy subjects (HS) were enrolled. All subjects underwent cognitive and MRI examination. MRI, including diffusion sequences, was used for probabilistic tractography analysis. We reconstructed individual FATs bilaterally and assessed their microstructural integrity using fractional anisotropy (FA), computed as both mean tract value and voxel-wise using SPM-8. Mean FA values were then used to test for correlations with cognitive measures. Mean tract FA and voxel-wise analyses revealed that patients with AD, compared to HS, had decreased FA in the FAT bilaterally. In addition, positive associations were found between FA in the FATs and patients' performance at tests for constructional praxis and visuospatial logical reasoning. The present results reveal a bilateral damage of FAT in AD patients. The association between FATs' microscopic abnormalities and constructive abilities fits well with the knowledge of a functional involvement of SMA and pre-SMA in movement sequences when executing constructive praxis tasks. The FAT is an associative bundle critically involved in the network sub-serving constructional praxis in patients with AD.

Correlating quantitative tractography at 3T MRI and cognitive tests in healthy older adults

This study used diffusion tensor imaging tractography at 3 T MRI to relate cognitive function to white matter tracts in the brain. Brain T2 fluid attenuated inversion recovery-weighted and diffusion tensor 3 T MRI scans were acquired in thirty-three healthy participants without mild cognitive impairment or dementia. They completed a battery of neuropsychological tests including the Montreal Cognitive Assessment, Stroop test, Trail Making Test B, Wechsler Memory Scale-III Longest span forward, Wechsler Memory Scale-III Longest span backward, Mattis Dementia Rating Scale, California Verbal Learning Test Version II Long Delay Free Recall, and Letter Number Sequencing. Tractography was generated by the Fiber Assignment by Continuous Tracking method. The corpus callosum, cingulum, long association fibers, corticospinal/bulbar tracts, thalamic projection fibers, superior cerebellar peduncle, middle cerebellar peduncle and inferior cerebellar peduncle were manually segmented. The fractional anisotropy (FA) and mean diffusivity (MD) of these tracts were quantified. We studied the association between cognitive test scores and the MD and FA of tracts while controlling for age and total white matter hyperintensities volume. Worse scores on the Stroop test was associated with decreased FA of the corpus callosum, corticospinal/bulbar tract, and thalamic projection tracts. Scores on the other cognitive tests were not associated with either the FA or MD of measured tracts. In healthy persons the Stroop test appears to be a better predictor of the microstructural integrity of white matter tracts measured by DTI tractography than other cognitive tests.

Age-related changes of white matter association tracts in normal children throughout adulthood: a diffusion tensor tractography study

PURPOSE

The aim of this study is to study the age, gender and lateral asymmetry-related white matter changes of long association tracts throughout late childhood and adolescence into adulthood using diffusion tensor tractography (DTT).

METHODS

DTT was performed in 44 healthy subjects aged 7-45 years. Fractional anisotropy (FA), radial diffusivity (RD), axial diffusivity (AD), Trace, density and volume were calculated for long association tracts, namely the inferior fronto-occipital fasciculus (IFOF), inferior longitudinal fasciculus (ILF), uncinate fasciculus, superior longitudinal fasciculus (SLF) and its arcuate fibres. FA and diffusivity indices were correlated as function of age using Pearson correlation test. Comparison between males and females, and comparison between both hemispheres among all participants were also performed. A p value less than .01 was considered significant.

RESULTS

The majority of the examined tracts (SLF and IFOF of both hemispheres, and the arcuate fasciculus, uncinate fasciculus, and ILF of the left hemisphere) followed a common pattern of metric changes with age. This pattern was characterized by significant FA increase accompanied by reduction in RD, Trace without significant AD changes. The right arcuate fasciculus showed similar pattern but without significant FA changes. The right uncinate and right ILF fasciculus demonstrated significant reduction in RD, Trace and AD, with and without significant FA increase, respectively. Left hemispheric dominance regarding the FA and diffusivity indices was demonstrated in uncinate fasciculus with no significant gender-related differences.

CONCLUSION

Significant microstructural tract-specific maturation processes continue throughout late childhood into adulthood. These processes may represent stages in a cascade of age-related maturation in white matter microstructure.

Insulin Resistance as a Link between Amyloid-Beta and Tau Pathologies in Alzheimer's Disease

Current hypotheses and theories regarding the pathogenesis of Alzheimer’s disease (AD) heavily implicate brain insulin resistance (IR) as a key factor. Despite the many well-validated metrics for systemic IR, the absence of biomarkers for brain-specific IR represents a translational gap that has hindered its study in living humans. In our lab, we have been working to develop biomarkers that reflect the common mechanisms of brain IR and AD that may be used to follow their engagement by experimental treatments. We present two promising biomarkers for brain IR in AD: insulin cascade mediators probed in extracellular vesicles (EVs) enriched for neuronal origin, and two-dimensional magnetic resonance spectroscopy (MRS) measures of brain glucose. As further evidence for a fundamental link between brain IR and AD, we provide a novel analysis demonstrating the close spatial correlation between brain expression of genes implicated in IR (using Allen Human Brain Atlas data) and tau and beta-amyloid pathologies. We proceed to propose the bold hypotheses that baseline differences in the metabolic reliance on glycolysis, and the expression of glucose transporters (GLUT) and insulin signaling genes determine the vulnerability of different brain regions to Tau and/or Amyloid beta (Aβ) pathology, and that IR is a critical link between these two pathologies that define AD. Lastly, we provide an overview of ongoing clinical trials that target IR as an angle to treat AD, and suggest how biomarkers may be used to evaluate treatment efficacy and target engagement.

Insulin resistance in Alzheimer's disease

The links between systemic insulin resistance (IR), brain-specific IR, and Alzheimer's disease (AD) have been an extremely productive area of current research. This review will cover the fundamentals and pathways leading to IR, its connection to AD via cellular mechanisms, the most prominent methods and models used to examine it, an introduction to the role of extracellular vesicles (EVs) as a source of biomarkers for IR and AD, and an overview of modern clinical studies on the subject. To provide additional context, we also present a novel analysis of the spatial correlation of gene expression in the brain with the aid of Allen Human Brain Atlas data. Ultimately, examining the relation between IR and AD can be seen as a means of advancing the understanding of both disease states, with IR being a promising target for therapeutic strategies in AD treatment. In conclusion, we highlight the therapeutic potential of targeting brain IR in AD and the main strategies to pursue this goal.

White matter abnormalities in long-term anabolic-androgenic steroid users: A pilot study

Recent studies of long-term anabolic-androgenic steroid (AAS) users reported amygdala structural and functional connectivity abnormalities. We assessed white matter microstructure in the inferior-fronto-occipital fasciculus (IFOF), a major associative bundle of the amygdala network. Diffusion weighted images acquired from 9 male long-term AAS users and 8 matched controls aged 36-51 years old were processed using a standardized pipeline (Tract-Based Spatial Statistics). Group differences were examined using linear regression with adjustment for age and current testosterone level. Compared to nonusers, AAS users exhibited significantly higher fractional anisotropy (FA) in the IFOF. Users showed markedly greater FA than nonusers on the left IFOF but only a modest, nonsignificant difference on the right IFOF. Moreover, FA was positively associated with lifetime cumulative AAS dose. Our results suggest that long-term AAS use alters IFOF white matter organization and integrity, which in turn might affect amygdala-related processes such as reward system function. Accordingly, further studies are needed to replicate findings in larger subject groups to determine the functional significance of the FA abnormality.

Neurological impressions on the organization of language networks in the human brain

BACKGROUND

More than 95% of right-handed individuals, as well as almost 80% of left-handed individuals, have left hemisphere dominance for language. The perisylvian networks of the dominant hemisphere tend to be the most important language systems in human brains, usually connected by bidirectional fibres originated from the superior longitudinal fascicle/arcuate fascicle system and potentially modifiable by learning. Neuroplasticity mechanisms take place to preserve neural functions after brain injuries. Language is dependent on a hierarchical interlinkage of serial and parallel processing areas in distinct brain regions considered to be elementary processing units. Whereas aphasic syndromes typically result from injuries to the dominant hemisphere, the extent of the distribution of language functions seems to be variable for each individual.

METHOD

Review of the literature Results: Several theories try to explain the organization of language networks in the human brain from a point of view that involves either modular or distributed processing or sometimes both. The most important evidence for each approach is discussed under the light of modern theories of organization of neural networks.

CONCLUSIONS

Understanding the connectivity patterns of language networks may provide deeper insights into language functions, supporting evidence-based rehabilitation strategies that focus on the enhancement of language organization for patients with aphasic syndromes.

In-vivo imaging of grey and white matter neuroinflammation in Alzheimer's disease: a positron emission tomography study with a novel radioligand, [18F]-FEPPA

Our primary aim was to compare neuroinflammation in cognitively intact control subjects and patients with Alzheimer's disease (AD) by using positron emission tomography (PET) with translocator protein 18 kDa (TSPO)-specific radioligand [(18)F]-FEPPA. [(18)F]-FEPPA PET scans were acquired on a high-resolution research tomograph in 21 patients with AD (47- 81 years) and 21 control subjects (49-82 years). They were analyzed by using a 2-tissue compartment model with arterial plasma input function. Differences in neuroinflammation, indexed as [(18)F]-FEPPA binding were compared, adjusting for differences in binding affinity class as determined by a single polymorphism in the TSPO gene (rs6971). In grey matter areas, [(18)F]-FEPPA was significantly higher in AD compared with healthy control subjects. Large increases were seen in the hippocampus, prefrontal, temporal, parietal and occipital cortex (average Cohen's d= 0.89). Voxel-based analyses confirmed significant clusters of neuroinflammation in the frontal, temporal and parietal cortex in patients with AD. In white matter, [(18)F]-FEPPA binding was elevated in the posterior limb of the internal capsule, and the cingulum bundle. Higher neuroinflammation in the parietal cortex (r= -0.7, P= 0.005), and posterior limb of the internal capsule (r= -0.8, P=0.001) was associated with poorer visuospatial function. In addition, a higher [(18)F]-FEPPA binding in the posterior limb of the internal capsule was associated with a greater impairment in language ability (r= -0.7, P=0.004). Elevated neuroinflammation can be detected in AD patients throughout the brain grey and white matter by using [(18)F]-FEPPA PET. Our results also suggest that neuroinflammation is associated with some cognitive deficits.

Local but not long-range microstructural differences of the ventral temporal cortex in developmental prosopagnosia

Individuals with developmental prosopagnosia (DP) experience face recognition impairments despite normal intellect and low-level vision and no history of brain damage. Prior studies using diffusion tensor imaging in small samples of subjects with DP (n=6 or n=8) offer conflicting views on the neurobiological bases for DP, with one suggesting white matter differences in two major long-range tracts running through the temporal cortex, and another suggesting white matter differences confined to fibers local to ventral temporal face-specific functional regions of interest (fROIs) in the fusiform gyrus. Here, we address these inconsistent findings using a comprehensive set of analyzes in a sample of DP subjects larger than both prior studies combined (n=16). While we found no microstructural differences in long-range tracts between DP and age-matched control participants, we found differences local to face-specific fROIs, and relationships between these microstructural measures with face recognition ability. We conclude that subtle differences in local rather than long-range tracts in the ventral temporal lobe are more likely associated with developmental prosopagnosia.

Vestibular loss and balance training cause similar changes in human cerebral white matter fractional anisotropy

Patients with bilateral vestibular loss suffer from severe balance deficits during normal everyday movements. Ballet dancers, figure skaters, or slackliners, in contrast, are extraordinarily well trained in maintaining balance for the extreme balance situations that they are exposed to. Both training and disease can lead to changes in the diffusion properties of white matter that are related to skill level or disease progression respectively. In this study, we used diffusion tensor imaging (DTI) to compare white matter diffusivity between these two study groups and their age- and sex-matched controls. We found that vestibular patients and balance-trained subjects show a reduction of fractional anisotropy in similar white matter tracts, due to a relative increase in radial diffusivity (perpendicular to the main diffusion direction). Reduced fractional anisotropy was not only found in sensory and motor areas, but in a widespread network including long-range connections, limbic and association pathways. The reduced fractional anisotropy did not correlate with any cognitive, disease-related or skill-related factors. The similarity in FA between the two study groups, together with the absence of a relationship between skill or disease factors and white matter changes, suggests a common mechanism for these white matter differences. We propose that both study groups must exert increased effort to meet their respective usual balance requirements. Since balance training has been shown to effectively reduce the symptoms of vestibular failure, the changes in white matter shown here may represent a neuronal mechanism for rehabilitation.

Intracranial injection of PEG-PEI/ROCK II-siRNA improves cognitive impairment in a mouse model of Alzheimer's disease

PURPOSE

A plenty of studies have demonstrated that the Rho/ROCK pathway is involved in the neuronal loss and inhibition of axonal regeneration observed in Alzheimer's disease (AD). Therefore, we conducted this study to evaluate whether intracranial injection of PEG-PEI/ROCK II siRNA (PPRS) would improve the cognitive impairments in a senescence-accelerated mouse (SAM) model of AD.

MATERIALS AND METHODS

Five male senescence-resistant inbred strain (SAMR1) mice and 15 male senescence-accelerated mouse prone-8 (SAMP8) strain mice were divided into the following three groups:PPRS group, PEG-PEI/ ROCK II-Scramble (PPRScr) siRNA group, and normal group (SAMR1). Total volumes of 2.3 μl of nanoparticles or saline were intracranially injected under the guidance of a stereotaxic apparatus. The injections were performed every three days and lasted for two weeks. Four weeks after injection, the Morris water maze (MWM) was used to evaluate the spatial learning and memory functions of the mice. Choline acetyltransferase (ChAT) activity was detected by immunohistochemistry.

RESULTS

Mice in the PPRS-treated group exhibited decreases in escape latencies over the three successive days of navigating the test and crossing the target quadrant during the spatial probe test more frequently than did the mice in the PPRScr-treated group. Analyses of ChAT activity revealed that greater numbers of ChAT-positive cells were present in the hippocampal regions of the PPRS-treated mice than in the PPRScr group.

CONCLUSIONS

Intracranial injection of PPRS improved the cognitive impairments of SAM mice, and this improvement may have been mediated by enhancement of ChAT activity in the hippocampus.

Robust automated detection of microstructural white matter degeneration in Alzheimer's disease using machine learning classification of multicenter DTI data

Diffusion tensor imaging (DTI) based assessment of white matter fiber tract integrity can support the diagnosis of Alzheimer’s disease (AD). The use of DTI as a biomarker, however, depends on its applicability in a multicenter setting accounting for effects of different MRI scanners. We applied multivariate machine learning (ML) to a large multicenter sample from the recently created framework of the European DTI study on Dementia (EDSD). We hypothesized that ML approaches may amend effects of multicenter acquisition. We included a sample of 137 patients with clinically probable AD (MMSE 20.6±5.3) and 143 healthy elderly controls, scanned in nine different scanners. For diagnostic classification we used the DTI indices fractional anisotropy (FA) and mean diffusivity (MD) and, for comparison, gray matter and white matter density maps from anatomical MRI. Data were classified using a Support Vector Machine (SVM) and a Naïve Bayes (NB) classifier. We used two cross-validation approaches, (i) test and training samples randomly drawn from the entire data set (pooled cross-validation) and (ii) data from each scanner as test set, and the data from the remaining scanners as training set (scanner-specific cross-validation). In the pooled cross-validation, SVM achieved an accuracy of 80% for FA and 83% for MD. Accuracies for NB were significantly lower, ranging between 68% and 75%. Removing variance components arising from scanners using principal component analysis did not significantly change the classification results for both classifiers. For the scanner-specific cross-validation, the classification accuracy was reduced for both SVM and NB. After mean correction, classification accuracy reached a level comparable to the results obtained from the pooled cross-validation. Our findings support the notion that machine learning classification allows robust classification of DTI data sets arising from multiple scanners, even if a new data set comes from a scanner that was not part of the training sample.