On brain atlas choice and automatic segmentation methods: a comparison of MAPER & FreeSurfer using three atlas databases

Fighting the scanner effect in brain MRI segmentation with a progressive level-of-detail network trained on multi-site data

Many clinical and research studies of the human brain require accurate structural MRI segmentation. While traditional atlas-based methods can be applied to volumes from any acquisition site, recent deep learning algorithms ensure high accuracy only when tested on data from the same sites exploited in training (i.e., internal data). Performance degradation experienced on external data (i.e., unseen volumes from unseen sites) is due to the inter-site variability in intensity distributions, and to unique artefacts caused by different MR scanner models and acquisition parameters. To mitigate this site-dependency, often referred to as the scanner effect, we propose LOD-Brain, a 3D convolutional neural network with progressive levels-of-detail (LOD), able to segment brain data from any site. Coarser network levels are responsible for learning a robust anatomical prior helpful in identifying brain structures and their locations, while finer levels refine the model to handle site-specific intensity distributions and anatomical variations. We ensure robustness across sites by training the model on an unprecedentedly rich dataset aggregating data from open repositories: almost 27,000 T1w volumes from around 160 acquisition sites, at 1.5 - 3T, from a population spanning from 8 to 90 years old. Extensive tests demonstrate that LOD-Brain produces state-of-the-art results, with no significant difference in performance between internal and external sites, and robust to challenging anatomical variations. Its portability paves the way for large-scale applications across different healthcare institutions, patient populations, and imaging technology manufacturers. Code, model, and demo are available on the project website.

Deconstructing pathological tau by biological process in early stages of Alzheimer disease: a method for quantifying tau spatial spread in neuroimaging

BACKGROUND

Neuroimaging studies often quantify tau burden in standardized brain regions to assess Alzheimer disease (AD) progression. However, this method ignores another key biological process in which tau spreads to additional brain regions. We have developed a metric for calculating the extent tau pathology has spread throughout the brain and evaluate the relationship between this metric and tau burden across early stages of AD.

METHODS

445 cross-sectional participants (aged ≥ 50) who had MRI, amyloid PET, tau PET, and clinical testing were separated into disease-stage groups based on amyloid positivity and cognitive status (older cognitively normal control, preclinical AD, and symptomatic AD). Tau burden and tau spatial spread were calculated for all participants.

FINDINGS

We found both tau metrics significantly elevated across increasing disease stages (p < 0.0001) and as a function of increasing amyloid burden for participants with preclinical (p < 0.0001, p = 0.0056) and symptomatic (p = 0.010, p = 0.0021) AD. An interaction was found between tau burden and tau spatial spread when predicting amyloid burden (p = 0.00013). Analyses of slope between tau metrics demonstrated more spread than burden in preclinical AD (β = 0.59), but then tau burden elevated relative to spread (β = 0.42) once participants had symptomatic AD, when the tau metrics became highly correlated (R = 0.83).

INTERPRETATION

Tau burden and tau spatial spread are both strong biomarkers for early AD but provide unique information, particularly at the preclinical stage. Tau spatial spread may demonstrate earlier changes than tau burden which could have broad impact in clinical trial design.

FUNDING

This research was supported by the Knight Alzheimer Disease Research Center (Knight ADRC, NIH grants P30AG066444, P01AG026276, P01AG003991), Dominantly Inherited Alzheimer Network (DIAN, NIH grants U01AG042791, U19AG03243808, R01AG052550-01A1, R01AG05255003), and the Barnes-Jewish Hospital Foundation Willman Scholar Fund.

Increased iron in the substantia nigra pars compacta identifies patients with early Parkinson'sdisease: A 3T and 7T MRI study

Degeneration in the substantia nigra (SN) pars compacta (SNc) underlies motor symptoms in Parkinson's disease (PD). Currently, there are no neuroimaging biomarkers that are sufficiently sensitive, specific, reproducible, and accessible for routine diagnosis or staging of PD. Although iron is essential for cellular processes, it also mediates neurodegeneration. MRI can localize and quantify brain iron using magnetic susceptibility, which could potentially provide biomarkers of PD. We measured iron in the SNc, SN pars reticulata (SNr), total SN, and ventral tegmental area (VTA), using quantitative susceptibility mapping (QSM) and R2* relaxometry, in PD patients and age-matched healthy controls (HCs). PD patients, diagnosed within five years of participation and HCs were scanned at 3T (22 PD and 23 HCs) and 7T (17 PD and 21 HCs) MRI. Midbrain nuclei were segmented using a probabilistic subcortical atlas. QSM and R2* values were measured in midbrain subregions. For each measure, groups were contrasted, with Age and Sex as covariates, and receiver operating characteristic (ROC) curve analyses were performed with repeated k-fold cross-validation to test the potential of our measures to classify PD patients and HCs. Statistical differences of area under the curves (AUCs) were compared using the Hanley-MacNeil method (QSM versus R2*; 3T versus 7T MRI). PD patients had higher QSM values in the SNc at both 3T (padj = 0.001) and 7T (padj = 0.01), but not in SNr, total SN, or VTA, at either field strength. No significant group differences were revealed using R2* in any midbrain region at 3T, though increased R2* values in SNc at 7T MRI were marginally significant in PDs compared to HCs (padj = 0.052). ROC curve analyses showed that SNc iron measured with QSM, distinguished early PD patients from HCs at the single-subject level with good diagnostic accuracy, using 3T (mean AUC = 0.83, 95 % CI = 0.82-0.84) and 7T (mean AUC = 0.80, 95 % CI = 0.79-0.81) MRI. Mean AUCs reported here are from averages of tests in the hold-out fold of cross-validated samples. The Hanley-MacNeil method demonstrated that QSM outperforms R2* in discriminating PD patients from HCs at 3T, but not 7T. There were no significant differences between 3T and 7T in diagnostic accuracy of QSM values in SNc. This study highlights the importance of segmenting midbrain subregions, performed here using a standardized atlas, and demonstrates high accuracy of SNc iron measured with QSM at 3T MRI in identifying early PD patients. QSM measures of SNc show potential for inclusion in neuroimaging diagnostic biomarkers of early PD. An MRI diagnostic biomarker of PD would represent a significant clinical advance.

Artificial Intelligence in Neuroradiology: A Review of Current Topics and Competition Challenges

There is an expanding body of literature that describes the application of deep learning and other machine learning and artificial intelligence methods with potential relevance to neuroradiology practice. In this article, we performed a literature review to identify recent developments on the topics of artificial intelligence in neuroradiology, with particular emphasis on large datasets and large-scale algorithm assessments, such as those used in imaging AI competition challenges. Numerous applications relevant to ischemic stroke, intracranial hemorrhage, brain tumors, demyelinating disease, and neurodegenerative/neurocognitive disorders were discussed. The potential applications of these methods to spinal fractures, scoliosis grading, head and neck oncology, and vascular imaging were also reviewed. The AI applications examined perform a variety of tasks, including localization, segmentation, longitudinal monitoring, diagnostic classification, and prognostication. While research on this topic is ongoing, several applications have been cleared for clinical use and have the potential to augment the accuracy or efficiency of neuroradiologists.

Automatic and manual segmentation of the piriform cortex: Method development and validation in patients with temporal lobe epilepsy and Alzheimer's disease

The piriform cortex (PC) is located at the junction of the temporal and frontal lobes. It is involved physiologically in olfaction as well as memory and plays an important role in epilepsy. Its study at scale is held back by the absence of automatic segmentation methods on MRI. We devised a manual segmentation protocol for PC volumes, integrated those manually derived images into the Hammers Atlas Database (n = 30) and used an extensively validated method (multi-atlas propagation with enhanced registration, MAPER) for automatic PC segmentation. We applied automated PC volumetry to patients with unilateral temporal lobe epilepsy with hippocampal sclerosis (TLE; n = 174 including n = 58 controls) and to the Alzheimer's Disease Neuroimaging Initiative cohort (ADNI; n = 151, of whom with mild cognitive impairment (MCI), n = 71; Alzheimer's disease (AD), n = 33; controls, n = 47). In controls, mean PC volume was 485 mm³ on the right and 461 mm³ on the left. Automatic and manual segmentations overlapped with a Jaccard coefficient (intersection/union) of ~0.5 and a mean absolute volume difference of ~22 mm³ in healthy controls, ~0.40/ ~28 mm³ in patients with TLE, and ~ 0.34/~29 mm³ in patients with AD. In patients with TLE, PC atrophy lateralised to the side of hippocampal sclerosis (p < .001). In patients with MCI and AD, PC volumes were lower than those of controls bilaterally (p < .001). Overall, we have validated automatic PC volumetry in healthy controls and two types of pathology. The novel finding of early atrophy of PC at the stage of MCI possibly adds a novel biomarker. PC volumetry can now be applied at scale.

A quantified comparison of cortical atlases on the basis of trait morphometricity

BACKGROUND

Many different brain atlases exist that subdivide the human cortex into dozens or hundreds of regions-of-interest (ROIs). Inconsistency across studies using one or another cortical atlas may contribute to the replication crisis across the neurosciences.

METHODS

Here, we provide a quantitative comparison between seven popular cortical atlases (Yeo, Desikan-Killiany, Destrieux, Jülich-Brain, Gordon, Glasser, Schaefer) and vertex-wise measures (thickness, surface area, and volume), to determine which parcellation retains the most information in the analysis of behavioural traits (incl. age, sex, body mass index, and cognitive ability) in the UK Biobank sample (N∼40,000). We use linear mixed models to compare whole-brain morphometricity; the proportion of trait variance accounted for when using a given atlas.

RESULTS

Commonly-used atlases resulted in a considerable loss of information compared to vertex-wise representations of cortical structure. Morphometricity increased linearly as a function of the log-number of ROIs included in an atlas, indicating atlas-based analyses miss many true associations and yield limited prediction accuracy. Likelihood ratio tests revealed that low-dimensional atlases accounted for unique trait variance rather than variance common between atlases, suggesting that previous studies likely returned atlas-specific findings. Finally, we found that the commonly-used atlases yielded brain-behaviour associations on par with those obtained with random parcellations, where specific region boundaries were randomly generated.

DISCUSSION

Our findings motivate future structural neuroimaging studies to favour vertex-wise cortical representations over coarser atlases, or to consider repeating analyses across multiple atlases, should the use of low-dimensional atlases be necessary. The insights uncovered here imply that cortical atlas choices likely contribute to the lack of reproducibility in ROI-based studies.

Consistency of Graph Theoretical Measurements of Alzheimer's Disease Fiber Density Connectomes Across Multiple Parcellation Scales

Graph theoretical measures have frequently been used to study disrupted connectivity in Alzheimer's disease human brain connectomes. However, prior studies have noted that differences in graph creation methods are confounding factors that may alter the topological observations found in these measures. In this study, we conduct a novel investigation regarding the effect of parcellation scale on graph theoretical measures computed for fiber density networks derived from diffusion tensor imaging. We computed 4 network-wide graph theoretical measures of average clustering coefficient, transitivity, characteristic path length, and global efficiency, and we tested whether these measures are able to consistently identify group differences among healthy control (HC), mild cognitive impairment (MCI), and AD groups in the Alzheimer's Disease Neuroimaging Initiative (ADNI) cohort across 5 scales of the Lausanne parcellation. We found that the segregative measure of transtivity offered the greatest consistency across scales in distinguishing between healthy and diseased groups, while the other measures were impacted by the selection of scale to varying degrees. Global efficiency was the second most consistent measure that we tested, where the measure could distinguish between HC and MCI in all 5 scales and between HC and AD in 3 out of 5 scales. Characteristic path length was highly sensitive to the variation in scale, corroborating previous findings, and could not identify group differences in many of the scales. Average clustering coefficient was also greatly impacted by scale, as it consistently failed to identify group differences in the higher resolution parcellations. From these results, we conclude that many graph theoretical measures are sensitive to the selection of parcellation scale, and further development in methodology is needed to offer a more robust characterization of AD's relationship with disrupted connectivity.

Metric magnetic resonance imaging analysis reveals pronounced substantia-innominata atrophy in dementia with Lewy bodies with a psychiatric onset

Background

Dementia with Lewy bodies (DLB) is a type of dementia often diagnosed in older patients. Since its initial symptoms range from delirium to psychiatric and cognitive symptoms, the diagnosis is often delayed.

Objectives

In our study, we evaluated the magnetic resonance imaging (MRI) of patients suffering from DLB in correlation with their initial symptoms taking a new pragmatic approach entailing manual measurements in addition to an automated volumetric analysis of MRI.

Methods

A total of 63 patients with diagnosed DLB and valid 3D data sets were retrospectively and blinded evaluated. We assessed atrophy patterns (1) manually for the substantia innominata and (2) via FastSurfer for the most common supratentorial regions. Initial symptoms were categorized by (1) mild cognitive impairment (MCI), (2) psychiatric episodes, and (3) delirium.

Results

Manual metric MRI measurements revealed moderate, but significant substantia-innominata (SI) atrophy in patients with a psychiatric onset. FastSurfer analysis revealed no regional volumetric differences between groups.

Conclusion

The SI in patients with DLB and a psychiatric-onset is more atrophied than that in patients with initial MCI. Our results suggest potential differences in SI between DLB subtypes at the prodromal stage, which are useful when taking a differential-diagnostic approach. This finding should be confirmed in larger patient cohorts.

High-resolution atlasing and segmentation of the subcortex: Review and perspective on challenges and opportunities created by machine learning

This paper reviews almost three decades of work on atlasing and segmentation methods for subcortical structures in human brain MRI. In writing this survey, we have three distinct aims. First, to document the evolution of digital subcortical atlases of the human brain, from the early MRI templates published in the nineties, to the complex multi-modal atlases at the subregion level that are available today. Second, to provide a detailed record of related efforts in the automated segmentation front, from earlier atlas-based methods to modern machine learning approaches. And third, to present a perspective on the future of high-resolution atlasing and segmentation of subcortical structures in in vivo human brain MRI, including open challenges and opportunities created by recent developments in machine learning.

Machine learning-based automatic estimation of cortical atrophy using brain computed tomography images

Cortical atrophy is measured clinically according to established visual rating scales based on magnetic resonance imaging (MRI). Although brain MRI is the primary imaging marker for neurodegeneration, computed tomography (CT) is also widely used for the early detection and diagnosis of dementia. However, they are seldom investigated. Therefore, we developed a machine learning algorithm for the automatic estimation of cortical atrophy on brain CT. Brain CT images (259 Alzheimer’s dementia and 55 cognitively normal subjects) were visually rated by three neurologists and used for training. We constructed an algorithm by combining the convolutional neural network and regularized logistic regression (RLR). Model performance was then compared with that of neurologists, and feature importance was measured. RLR provided fast and reliable automatic estimations of frontal atrophy (75.2% accuracy, 93.6% sensitivity, 67.2% specificity, and 0.87 area under the curve [AUC]), posterior atrophy (79.6% accuracy, 87.2% sensitivity, 75.9% specificity, and 0.88 AUC), right medial temporal atrophy (81.2% accuracy, 84.7% sensitivity, 79.6% specificity, and 0.88 AUC), and left medial temporal atrophy (77.7% accuracy, 91.1% sensitivity, 72.3% specificity, and 0.90 AUC). We concluded that RLR-based automatic estimation of brain CT provided a comprehensive rating of atrophy that can potentially support physicians in real clinical settings.

Learning to detect boundary information for brain image segmentation

MRI brain images are always of low contrast, which makes it difficult to identify to which area the information at the boundary of brain images belongs. This can make the extraction of features at the boundary more challenging, since those features can be misleading as they might mix properties of different brain regions. Hence, to alleviate such a problem, image boundary detection plays a vital role in medical image segmentation, and brain segmentation in particular, as unclear boundaries can worsen brain segmentation results. Yet, given the low quality of brain images, boundary detection in the context of brain image segmentation remains challenging. Despite the research invested to improve boundary detection and brain segmentation, these two problems were addressed independently, i.e., little attention was paid to applying boundary detection to brain segmentation tasks. Therefore, in this paper, we propose a boundary detection-based model for brain image segmentation. To this end, we first design a boundary segmentation network for detecting and segmenting images brain tissues. Then, we design a boundary information module (BIM) to distinguish boundaries from the three different brain tissues. After that, we add a boundary attention gate (BAG) to the encoder output layers of our transformer to capture more informative local details. We evaluate our proposed model on two datasets of brain tissue images, including infant and adult brains. The extensive evaluation experiments of our model show better performance (a Dice Coefficient (DC) accuracy of up to \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$5.3\%$$\end{document}5.3% compared to the state-of-the-art models) in detecting and segmenting brain tissue images.

Brain atrophy patterns in multiple sclerosis patients treated with natalizumab and its clinical correlates

BACKGROUND

Multiple sclerosis (MS) is defined as a demyelinating disorder of the central nervous system, witnessing over the past years a remarkable progress in the therapeutic approaches of the inflammatory process. Yet, the ongoing neurodegenerative process is still ambiguous, under-assessed, and probably under-treated. Atrophy and cognitive dysfunction represent the radiological and clinical correlates of such process. In this study, we evaluated the effect of one specific MS treatment, which is natalizumab (NTZ), on brain atrophy evolution in different anatomical regions and its correlation with the cognitive profile and the physical disability.

METHODS

We recruited 20 patients diagnosed with relapsing-remitting MS (RR-MS) and treated with NTZ. We tracked brain atrophy in different anatomical structures using MRI scans processed with an automated image segmentation technique. We also assessed the progression of physical disability and the cognitive function and its link with the progression of atrophy.

RESULTS

During the first 2 years of treatment, a significant volume loss was noted within the corpus callosum and the cerebellum gray matter (GM). The annual atrophy rate of the cortical GM, the cerebellum GM, the thalamus, the amygdala, the globus pallidus, and the hippocampus correlated with greater memory impairment. As for the third and fourth years of treatment, a significant atrophy revolved around the gray matter, mainly the cortical one. We also noted an increase of the thalamus volume.

CONCLUSION

Atrophy in RR-MS patients treated with NTZ is regional and targeting highly cognitive regions mainly of the subcortical gray matter and the cerebellum. The cerebellum atrophy was a marker of physical disability progression. NTZ did not accelerate the atrophy process in MS and may play a neuroprotective role by increasing the thalamus volume.

Manual and automated analysis of atrophy patterns in dementia with Lewy bodies on MRI

Background

Dementia with Lewy bodies (DLB) is the second most common dementia type in patients older than 65 years. Its atrophy patterns remain unknown. Its similarities to Parkinson's disease and differences from Alzheimer's disease are subjects of current research.

Methods

The aim of our study was (i) to form a group of patients with DLB (and a control group) and create a 3D MRI data set (ii) to volumetrically analyze the entire brain in these groups, (iii) to evaluate visual and manual metric measurements of the innominate substance for real-time diagnosis, and (iv) to compare our groups and results with the latest literature. We identified 102 patients with diagnosed DLB in our psychiatric and neurophysiological archives. After exclusion, 63 patients with valid 3D data sets remained. We compared them with a control group of 25 patients of equal age and sex distribution. We evaluated the atrophy patterns in both (1) manually and (2) via Fast Surfers segmentation and volumetric calculations. Subgroup analyses were done of the CSF data and quality of 3D T1 data sets.

Results

Concordant with the literature, we detected moderate, symmetric atrophy of the hippocampus, entorhinal cortex and amygdala, as well as asymmetric atrophy of the right parahippocampal gyrus in DLB. The caudate nucleus was unaffected in patients with DLB, while all the other measured territories were slightly too moderately atrophied. The area under the curve analysis of the left hippocampus volume ratio (< 3646mm³) revealed optimal 76% sensitivity and 100% specificity (followed by the right hippocampus and left amygdala). The substantia innominata’s visual score attained a 51% optimal sensitivity and 84% specificity, and the measured distance 51% optimal sensitivity and 68% specificity in differentiating DLB from our control group.

Conclusions

In contrast to other studies, we observed a caudate nucleus sparing atrophy of the whole brain in patients with DLB. As the caudate nucleus is known to be the last survivor in dopamine-uptake, this could be the result of an overstimulation or compensation mechanism deserving further investigation. Its relative hypertrophy compared to all other brain regions could enable an imaging based identification of patients with DLB via automated segmentation and combined volumetric analysis of the hippocampus and amygdala.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12883-022-02642-0.

White Matter Connectivity between Structures of the Basal Ganglia using 3T and 7T

Analysis of the basal ganglia has been important in investigating the effects of Parkinson's disease as well as treatments for Parkinson's disease. One method of analysis has been using MRI for non-invasively segmenting the basal ganglia, then investigating significant parameters that involve the basal ganglia, such as fiber orientations and positional markers for deep brain stimulation (DBS). Following enhancements to optimizations and improvements to 3T and 7T MRI acquisitions, we utilized Lead-DBS on human connectome project data to automatically segment the basal ganglia of 49 human connectome project subjects, reducing the reliance on manual segmentation for more consistency. We generated probabilistic tractography streamlines between each segmentation pair using 3T and 7T human connectome diffusion data to observe any major differences in tractography streamline patterns that can arise due to tradeoffs from different field strengths and acquisitions. Tractography streamlines generated between basal ganglia structures using 3T images showed less standard deviation in streamline count than using 7T images. Mean tractography streamline counts generated using 3T diffusion images were all higher in count than streamlines generated using 7T diffusion images. We illustrate a potential method for analyzing the structural connectivity between basal ganglia structures, as well as visualize possible differences in probabilistic tractography that can arise from different acquisition protocols.

Brain Volumetric Measurements in Children With Attention Deficit Hyperactivity Disorder: A Comparative Study Between Synthetic and Conventional Magnetic Resonance Imaging

Objective: To investigate the profiles of brain volumetric measurements in children with attention deficit hyperactivity disorder (ADHD), and the consistency of these brain volumetric measurements derived from the synthetic and conventional T1 weighted MRI (SyMRI and cT1w MRI).

Methods: Brain SyMRI and cT1w images were prospectively collected for 38 pediatric patients with ADHD and 38 healthy children (HC) with an age range of 6–14 years. The gray matter volume (GMV), white matter volume (WMV), cerebrospinal fluid (CSF), non-WM/GM/CSF (NoN), myelin, myelin fraction (MYF), brain parenchyma volume (BPV), and intracranial volume (ICV) were automatically estimated from SyMRI data, and the four matching measurements (GMV, WMV, BPV, ICV) were extracted from cT1w images. The group differences of brain volumetric measurements were performed, respectively, using analysis of covariance. Pearson correlation analysis and interclass correlation coefficient (ICC) were applied to evaluate the association between synthetic and cT1w MRI-derived measurements.

Results: As for the brain volumetric measurements extracted from SyMRI, significantly decreased GMV, WMV, BPV, and increased NON volume (p < 0.05) were found in the ADHD group compared with HC; No group differences were found in ICV, CSF, myelin volume and MYF (p > 0.05). With regard to GMV, WMV, BPV, and ICV estimated from cT1w images, the group differences between ADHD and HC were consistent with the results estimated from SyMRI. And these four measurements showed noticeable correlation between the two approaches (r = 0.692, 0.643, 0.898, 0.789, respectively, p < 0.001; ICC values are 0.809, 0.782, 0.946, 0.873, respectively).

Conclusion: Our study demonstrated a global brain development disability, but normal whole-brain myelination in children with ADHD. Moreover, our results demonstrated the high consistency of brain volumetric indices between synthetic and cT1w MRI in children, which indicates the high reliability of SyMRI in the child-brain volumetric analysis.

Analysis of structural brain asymmetries in attention-deficit/hyperactivity disorder in 39 datasets

OBJECTIVE

Some studies have suggested alterations of structural brain asymmetry in attention-deficit/hyperactivity disorder (ADHD), but findings have been contradictory and based on small samples. Here, we performed the largest ever analysis of brain left-right asymmetry in ADHD, using 39 datasets of the ENIGMA consortium.

METHODS

We analyzed asymmetry of subcortical and cerebral cortical structures in up to 1,933 people with ADHD and 1,829 unaffected controls. Asymmetry Indexes (AIs) were calculated per participant for each bilaterally paired measure, and linear mixed effects modeling was applied separately in children, adolescents, adults, and the total sample, to test exhaustively for potential associations of ADHD with structural brain asymmetries.

RESULTS

There was no evidence for altered caudate nucleus asymmetry in ADHD, in contrast to prior literature. In children, there was less rightward asymmetry of the total hemispheric surface area compared to controls (t = 2.1, p = .04). Lower rightward asymmetry of medial orbitofrontal cortex surface area in ADHD (t = 2.7, p = .01) was similar to a recent finding for autism spectrum disorder. There were also some differences in cortical thickness asymmetry across age groups. In adults with ADHD, globus pallidus asymmetry was altered compared to those without ADHD. However, all effects were small (Cohen's d from -0.18 to 0.18) and would not survive study-wide correction for multiple testing.

CONCLUSION

Prior studies of altered structural brain asymmetry in ADHD were likely underpowered to detect the small effects reported here. Altered structural asymmetry is unlikely to provide a useful biomarker for ADHD, but may provide neurobiological insights into the trait.

The Association Between Perivascular Spaces and Cerebral Blood Flow, Brain Volume, and Cardiovascular Risk

Background: Basal ganglia perivascular spaces are associated with cognitive decline and cardiovascular risk factors. There is a lack of studies on the cardiovascular risk burden of basal ganglia perivascular spaces (BG-PVS) and their relationship with gray matter volume (GMV) and GM cerebral blood flow (CBF) in the aging brain. Here, we investigated these two issues in a large sample of cognitively intact older adults.

Methods: A total of 734 volunteers were recruited. MRI was performed with 3.0 T using a pseudo-continuous arterial spin labeling (pCASL) sequence and a sagittal isotropic T1-weighted sequence for CBF and GMV analysis. The images obtained from 406 participants were analyzed to investigate the relationship between the severity of BG-PVS and GMV/CBF. False discovery rate-corrected P-values (P_FDR) of <0.05 were considered significant. The images obtained from 254 participants were used to study the relationship between the severity of BG-PVS and cardiovascular risk burden. BG-PVS were rated using a 5-grade score. The severity of BG-PVS was classified as mild (grade <3) and severe (grade ≥3). Cardiovascular risk burden was assessed with the Framingham General Cardiovascular Risk Score (FGCRS).

Results: Severe basal ganglia perivascular spaces were associated with significantly smaller GMV and CBF in multiple cortical regions (P_FDR <0.05), and were associated with significantly larger volume in the bilateral caudate nucleus, pallidum, and putamen (P_FDR <0.05). The participants with severe BG-PVS were more likely to have a higher cardiovascular risk burden than the participants with mild BG-PVS (60.71% vs. 42.93%; P =0.02).

Conclusion: In cognitively intact older adults, severe BG-PVS are associated with smaller cortical GMV and CBF, larger subcortical GMV, and higher cardiovascular risk burden.

CERMEP-IDB-MRXFDG: a database of 37 normal adult human brain [18F]FDG PET, T1 and FLAIR MRI, and CT images available for research

We present a database of cerebral PET FDG and anatomical MRI for 37 normal adult human subjects (CERMEP-IDB-MRXFDG). Thirty-nine participants underwent static [18F]FDG PET/CT and MRI, resulting in [18F]FDG PET, T1 MPRAGE MRI, FLAIR MRI, and CT images. Two participants were excluded after visual quality control. We describe the acquisition parameters, the image processing pipeline and provide participants' individual demographics (mean age 38 ± 11.5 years, range 23-65, 20 women). Volumetric analysis of the 37 T1 MRIs showed results in line with the literature. A leave-one-out assessment of the 37 FDG images using Statistical Parametric Mapping (SPM) yielded a low number of false positives after exclusion of artefacts. The database is stored in three different formats, following the BIDS common specification: (1) DICOM (data not processed), (2) NIFTI (multimodal images coregistered to PET subject space), (3) NIFTI normalized (images normalized to MNI space). Bona fide researchers can request access to the database via a short form.

Association between structural brain features and gene expression by weighted gene co-expression network analysis in conversion from MCI to AD

Alzheimer's disease (AD) is a neurodegenerative disease. Mild cognitive impairment (MCI) represents a state of cognitive function between normal cognition and dementia. Longitudinal studies showed that some MCI patients remained in a state of MCI, and some developed AD. The reason for these different conversions from MCI remains to be investigated. 180 MCI participants were followed for eight years. 143 MCI patients maintained the MCI state (MCI_S), and the remaining thirty-seven MCI patients were re-evaluated as having AD (MCI_AD). We obtained 1,036 structural brain characteristics and 15,481 gene expression values from the 180 MCI participants and applied weighted gene co-expression network analysis (WGCNA) to explore the relationship between structural brain features and gene expression. Regulating mediator effect analysis was employed to explore the relationships among gene expression, brain region measurements and clinical phenotypes. We found that 60 genes from the MCI_S group and 18 genes from the MCI_AD group respectively had the most significant correlations with left paracentral lobule and sulcus (L.PTS) and right subparietal sulcus (R.SubPS) thickness; CTCF, UQCR11 and WDR5B were the mutual genes between the two groups. The expression of CTCF gene and clinical score are completely mediated by L.PTS thickness, and the UQCR11 and WDR5B gene expression levels significantly regulate the mediating effect pathway. In conclusion, the factors affecting the different conversions from MCI are closely related to L.PTS thickness and the CTCF, UQCR11 and WDR5B gene expression levels. Our results add a theoretical foundation of imaging genetics for conversion from MCI to AD.

A Review of Publicly Available Automatic Brain Segmentation Methodologies, Machine Learning Models, Recent Advancements, and Their Comparison

BACKGROUND

The noninvasive study of the structure and functions of the brain using neuroimaging techniques is increasingly being used for its clinical and research perspective. The morphological and volumetric changes in several regions and structures of brains are associated with the prognosis of neurological disorders such as Alzheimer's disease, epilepsy, schizophrenia, etc. and the early identification of such changes can have huge clinical significance. The accurate segmentation of three-dimensional brain magnetic resonance images into tissue types (i.e., grey matter, white matter, cerebrospinal fluid) and brain structures, thus, has huge importance as they can act as early biomarkers. The manual segmentation though considered the "gold standard" is time-consuming, subjective, and not suitable for bigger neuroimaging studies. Several automatic segmentation tools and algorithms have been developed over the years; the machine learning models particularly those using deep convolutional neural network (CNN) architecture are increasingly being applied to improve the accuracy of automatic methods.

PURPOSE

The purpose of the study is to understand the current and emerging state of automatic segmentation tools, their comparison, machine learning models, their reliability, and shortcomings with an intent to focus on the development of improved methods and algorithms.

METHODS

The study focuses on the review of publicly available neuroimaging tools, their comparison, and emerging machine learning models particularly those based on CNN architecture developed and published during the last five years.

CONCLUSION

Several software tools developed by various research groups and made publicly available for automatic segmentation of the brain show variability in their results in several comparison studies and have not attained the level of reliability required for clinical studies. The machine learning models particularly three dimensional fully convolutional network models can provide a robust and efficient alternative with relation to publicly available tools but perform poorly on unseen datasets. The challenges related to training, computation cost, reproducibility, and validation across distinct scanning modalities for machine learning models need to be addressed.

Ensemble of ROI-based convolutional neural network classifiers for staging the Alzheimer disease spectrum from magnetic resonance imaging

Patches from three orthogonal views of selected cerebral regions can be utilized to learn convolutional neural network (CNN) models for staging the Alzheimer disease (AD) spectrum including preclinical AD, mild cognitive impairment due to AD, and dementia due to AD and normal controls. Hippocampi, amygdalae and insulae were selected from the volumetric analysis of structured magnetic resonance images (MRIs). Three-view patches (TVPs) from these regions were fed to the CNN for training. MRIs were classified with the SoftMax-normalized scores of individual model predictions on TVPs. The significance of each region of interest (ROI) for staging the AD spectrum was evaluated and reported. The results of the ensemble classifier are compared with state-of-the-art methods using the same evaluation metrics. Patch-based ROI ensembles provide comparable diagnostic performance for AD staging. In this work, TVP-based ROI analysis using a CNN provides informative landmarks in cerebral MRIs and may have significance in clinical studies and computer-aided diagnosis system design.