A new rapid landmark-based regional MRI segmentation method of the brain

A deep semantic segmentation correction network for multi-model tiny lesion areas detection

Background

Semantic segmentation of white matter hyperintensities related to focal cerebral ischemia (FCI) and lacunar infarction (LACI) is of significant importance for the automatic screening of tiny cerebral lesions and early prevention of LACI. However, existing studies on brain magnetic resonance imaging lesion segmentation focus on large lesions with obvious features, such as glioma and acute cerebral infarction. Owing to the multi-model tiny lesion areas of FCI and LACI, reliable and precise segmentation and/or detection of these lesion areas is still a significant challenge task.

Methods

We propose a novel segmentation correction algorithm for estimating the lesion areas via segmentation and correction processes, in which we design two sub-models simultaneously: a segmentation network and a correction network. The segmentation network was first used to extract and segment diseased areas on T2 fluid-attenuated inversion recovery (FLAIR) images. Consequently, the correction network was used to classify these areas at the corresponding locations on T1 FLAIR images to distinguish between FCI and LACI. Finally, the results of the correction network were used to correct the segmentation results and achieve segmentation and recognition of the lesion areas.

Results

In our experiment on magnetic resonance images of 113 clinical patients, our method achieved a precision of 91.76% for detection and 92.89% for classification, indicating a powerful method to distinguish between small lesions, such as FCI and LACI.

Conclusions

Overall, we developed a complete method for segmentation and detection of WMHs related to FCI and LACI. The experimental results show that it has potential clinical application potential. In the future, we will collect more clinical data and test more types of tiny lesions at the same time.

Hearing Impairment Is Associated with Smaller Brain Volume in Aging

Although recent studies show that age-related hearing impairment is associated with cerebral changes, data from a population perspective are still lacking. Therefore, we studied the relation between hearing impairment and brain volume in a large elderly cohort. From the population-based Rotterdam Study, 2,908 participants (mean age 65 years, 56% female) underwent a pure-tone audiogram to quantify hearing impairment. By performing MR imaging of the brain we quantified global and regional brain tissue volumes (total brain volume, gray matter volume, white matter (WM) volume, and lobe-specific volumes). We used multiple linear regression models, adjusting for age, sex, head size, time between hearing test and MR imaging, and relevant cognitive and cardiovascular covariates. Furthermore, we performed voxel-based morphometry to explore sub-regional differences. We found that a higher pure-tone threshold was associated with a smaller total brain volume [difference in standardized brain volume per decibel increase in hearing threshold in the age-sex adjusted model: -0.003 (95% confidence interval -0.004; -0.001)]. Specifically, WM volume was associated. Both associations were more pronounced in the lower frequencies. All associations were consistently present in all brain lobes in the lower frequencies and in most lobes in the higher frequencies, and were independent of cognitive function and cardiovascular risk factors. In voxel-based analyses we found associations of hearing impairment with smaller white volumes and some smaller and larger gray volumes, yet these were statistically non-significant. Our findings demonstrate that hearing impairment in elderly is related to smaller total brain volume, independent of cognition and cardiovascular risk factors. This mainly seems to be driven by smaller WM volume, throughout the brain.

The Rotterdam Scan Study: design update 2016 and main findings

Imaging plays an essential role in research on neurological diseases in the elderly. The Rotterdam Scan Study was initiated as part of the ongoing Rotterdam Study with the aim to elucidate the causes of neurological disease by performing imaging of the brain in a prospective population-based setting. Initially, in 1995 and 1999, random subsamples of participants from the Rotterdam Study underwent neuroimaging, whereas from 2005 onwards MRI has been implemented into the core protocol of the Rotterdam Study. In this paper, we discuss the background and rationale of the Rotterdam Scan Study. Moreover, we describe the imaging protocol, image post-processing techniques, and the main findings to date. Finally, we provide recommendations for future research, which will also be topics of investigation in the Rotterdam Scan Study.

The Rotterdam Scan Study: design and update up to 2012

Neuroimaging plays an important role in etiologic research on neurological diseases in the elderly. The Rotterdam Scan Study was initiated as part of the ongoing Rotterdam Study with the aim to unravel causes of neurological disease by performing neuroimaging in a population-based longitudinal setting. In 1995 and 1999 random subsets of the Rotterdam Study underwent neuroimaging, whereas from 2005 onwards MRI has been implemented into the core protocol of the Rotterdam Study. In this paper, we discuss the background and rationale of the Rotterdam Scan Study. We also describe the imaging protocol and post-processing techniques, and highlight the main findings to date. Finally, we make recommendations for future research, which will also be the main focus of investigation in the Rotterdam Scan Study.

Longitudinal whole-brain N-acetylaspartate concentration in healthy adults

BACKGROUND AND PURPOSE

Although NAA is often used as a marker of neural integrity and health in different neurologic disorders, the temporal behavior of WBNAA is not well characterized. Our goal therefore was to establish its normal variations in a cohort of healthy adults over typical clinical trial periods.

MATERIALS AND METHODS

Baseline amount of brain NAA, Q(NAA), was obtained with nonlocalizing proton MR spectroscopy from 9 subjects (7 women, 2 men; 31.2 ± 5.6 years old). Q(NAA) was converted into absolute millimole amount by using phantom-replacement. The WBNAA concentration was derived by dividing Q(NAA) with the brain parenchyma volume, V(B), segmented from MR imaging. Temporal variations were determined with 4 annual scans of each participant.

RESULTS

The distribution of WBNAA levels was not different among time points with respect to the mean, 12.1 ± 1.5 mmol/L (P > .6), nor was its intrasubject change (coefficient of variation = 8.6%) significant between any 2 scans (P > .5). There was a small (0.2 mL) but significant (P = .05) annual V(B) decline.

CONCLUSIONS

WBNAA is stable over a 3-year period in healthy adults. It qualifies therefore as a biomarker for global neuronal loss and dysfunction in diffuse neurologic disorders that may be well worth considering as a secondary outcome measure candidate for clinical trials.

Brain tissue volumes in relation to cognitive function and risk of dementia

We investigated in a population-based cohort study the association of global and lobar brain tissue volumes with specific cognitive domains and risk of dementia. Participants (n=490; 60-90 years) were non-demented at baseline (1995-1996). From baseline brain MRI-scans we obtained global and lobar volumes of CSF, GM, normal WM, white matter lesions and hippocampus. We performed neuropsychological testing at baseline to assess information processing speed, executive function, memory function and global cognitive function. Participants were followed for incident dementia until January 1, 2005. Larger volumes of CSF and WML were associated with worse performance on all neuropsychological tests, and an increased risk of dementia. Smaller WM volume was related to poorer information processing speed and executive function. In contrast, smaller GM volume was associated with worse memory function and increased risk of dementia. When investigating lobar GM volumes, we found that hippocampal volume and temporal GM volume were most strongly associated with risk of dementia, even in persons without objective and subjective cognitive deficits at baseline, followed by frontal and parietal GM volumes.

Alteration of brain metabolites in young alcoholics without structural changes

Using proton magnetic resonance spectroscopy and magnetic resonance imaging, we investigated concentrations of various brain metabolites, including glutamate, and measured brain volumes and neuropsychological performances in 13 recently abstinent young alcoholic men compared with 18 controls. No differences were found in volumetric variables between groups (intracranial volume, white matter, grey matter, anterior cingulate, insula, hippocampus, and amygdala). For the anterior cingulate, choline and creatine levels in the patient group were significantly lower than controls, and the glutamate to creatine ratio was significantly increased. These were correlated with altered short-term memory functions. Thus, neurochemical changes can occur even in the brains of young alcoholic men lacking brain atrophy.

Kidney function is related to cerebral small vessel disease

BACKGROUND AND PURPOSE

Poor kidney function, as measured by glomerular filtration rate (GFR), is closely associated with presence of glomerular small vessel disease. Given the hemodynamic similarities between the vascular beds of the kidney and the brain, we hypothesized an association between kidney function and markers of cerebral small vessel disease on MRI. We investigated this association in a population-based study of elderly persons.

METHODS

We measured GFR using the Cockcroft-Gault equation in 484 participants (60 to 90 years of age) from the Rotterdam Scan Study. Using automated MRI-analysis we measured global as well as lobar and deep volumes of gray matter and white matter, and volume of WML. Lacunar infarcts were rated visually. Volumes of deep white matter and WML and presence of lacunar infarcts reflected cerebral small vessel disease. We used linear and logistic regression models to investigate the association between GFR and brain imaging parameters. Analyses were adjusted for age, sex, and additionally for cardiovascular risk factors.

RESULTS

Persons with lower GFR had less deep white matter volume (difference in standardized volume per SD decrease in GFR: -0.15 [95% CI -0.26 to -0.04]), more WML (difference per SD decrease in GFR: 0.14 [95% CI 0.03 to 0.25]), and more often lacunar infarcts, although the latter was not significant. GFR was not associated with gray matter volume or lobar white matter volume. Additional adjustment for cardiovascular risk factors yielded similar results.

CONCLUSIONS

Impaired kidney function is associated with markers of cerebral small vessel disease as assessed on MRI.

Whole-brain N-acetylaspartate as a surrogate marker of neuronal damage in diffuse neurologic disorders

Proton MR spectroscopy (1H-MR spectroscopy) is a quantitative MR imaging technique often used to complement the sensitivity of conventional MR imaging with specific metabolic information. A key metabolite is the amino acid derivative N-acetylaspartate (NAA), which is almost exclusive to neurons and their processes and is, therefore, an accepted marker of their health and attenuation. Unfortunately, most 1H-MR spectroscopy studies only account for small 1- to 200-cm volumes of interest (VOI), representing less than 20% of the total brain volume. These VOIs have at least 5 additional restrictions: 1) To avoid contamination from subcutaneous and bone marrow lipids, they must be placed away from the skull, thereby missing most of the cortex. 2) They must be image-guided onto MR imaging-visible pathology, subjecting them to the implicit assumption that metabolic changes occur only there. 3) They encounter misregistration errors in serial studies. 4) The time needed to accumulate sufficient signal-intensity quality is often restrictive, and 5) they incur (unknown) T1- and T2-weighting. All these issues are avoided (at the cost of specific localization) by measuring the nonlocalized average NAA concentration over the entire brain. Indeed, whole-brain NAA quantification has been applied to several diffuse neurodegenerative diseases (where specific localization is less important than the total load of the pathology), and the results are presented in this review.

From genes to brain: understanding brain development in neurogenetic disorders using neuroimaging techniques

For almost two decades, a considerable amount of work has been devoted to the accurate delineation of normal and abnormal brain development using cerebral MRI. In the broad field of neuroimaging research, specific genetic conditions associated with impaired cognitive performances or with psychiatric symptoms have received increased attention because of their potential for revealing insight on the biologic correlates of behavior. First delineated by volumetric measurements of cerebral lobes or regions of interest, new image processing techniques are currently defining cerebral phenotypes associated with neurogenetic disorders with increasing precision. In this article the authors review the contribution of structural brain imaging in advancing our understanding of the pathogenic processes underlying altered brain development in Down, fragile X, and velocardiofacial (22q11DS) syndromes.

Magnetic resonance approaches to brain aging and Alzheimer disease-associated neuropathology

The noninvasive, nonradioactive, quantitative nature of magnetic resonance techniques has propelled them to the forefront of neuroscience and neuropsychiatric research. In particular, recent advances have confirmed their enormous potential in patients with Alzheimer disease (AD). Structural and functional magnetic resonance (MR) imaging have demonstrated significant correlation with clinical outcomes and underlying pathology and are used increasingly in the AD clinic. This review will highlight the role of high-resolution structural MR imaging and functional magnetic resonance imaging in the identification of atrophic and hemodynamic changes in AD and their potential as diagnostic biomarkers and surrogates of therapeutic response. Advanced MR techniques based on diffusion, perfusion, and neurochemical abnormalities in the aging brain will be presented briefly. These newer techniques continue to expand our understanding of neuropathology in the aging brain and are likely to play an important clinical role in the future.

A knowledge-guided active model method of cortical structure segmentation on pediatric MR images

PURPOSE

To develop an automated method for quantification of cortical structures on pediatric MR images.

MATERIALS AND METHODS

A knowledge-guided active model (KAM) approach was proposed with a novel object function similar to the Gibbs free energy function. Triangular mesh models were transformed to images of a given subject by maximizing entropy, and then actively slithered to boundaries of structures by minimizing enthalpy. Volumetric results and image similarities of 10 different cortical structures segmented by KAM were compared with those traced manually. Furthermore, the segmentation performances of KAM and SPM2, (statistical parametric mapping, a MATLAB software package) were compared.

RESULTS

The averaged volumetric agreements between KAM- and manually-defined structures (both 0.95 for structures in healthy children and children with medulloblastoma) were higher than the volumetric agreement for SPM2 (0.90 and 0.80, respectively). The similarity measurements (kappa) between KAM- and manually-defined structures (0.95 and 0.93, respectively) were higher than those for SPM2 (both 0.86).

CONCLUSION

We have developed a novel automatic algorithm, KAM, for segmentation of cortical structures on MR images of pediatric patients. Our preliminary results indicated that when segmenting cortical structures, KAM was in better agreement with manually-delineated structures than SPM2. KAM can potentially be used to segment cortical structures for conformal radiation therapy planning and for quantitative evaluation of changes in disease or abnormality.

Neuroanatomy of Down syndrome in vivo: a model of preclinical Alzheimer's disease

Aging in Down syndrome (DS) is accompanied by neuropathological features of Alzheimer's disease (AD). Therefore, DS has been proposed as a model to study predementia stages of AD. MRI-based measurement of grey matter atrophy is an in vivo surrogate marker of regional neuronal density. A range of neuroimaging studies have described the macroscopic neuroanatomy of DS. Recent studies using sensitive quantitative measures of region-specific atrophy based on high-resolution MRI suggest that age-related atrophy in DS resembles the pattern of brain atrophy in early stages of AD. The pattern of atrophy determined in predementia DS supports the notion that AD-type pathology leads to neuronal degeneration not only in allocortical, but also in neocortical brain areas before onset of clinical dementia. This has major implications for our understanding of the onset and progression of AD-type pathology both in DS and in sporadic AD.

Interactive segmentation of the cerebral lobes with fuzzy inference in 3T MR images

Measurement of volume and surface area of the frontal, parietal, temporal and occipital lobes from magnetic resonance (MR) images shows promise as a method for use in diagnosis of dementia. This article presents a novel computer-aided system for automatically segmenting the cerebral lobes from 3T human brain MR images. Until now, the anatomical definition of cerebral lobes on the cerebral cortex is somewhat vague for use in automatic delineation of boundary lines, and there is no definition of cerebral lobes in the interior of the cerebrum. Therefore, we have developed a new method for defining cerebral lobes on the cerebral cortex and in the interior of the cerebrum. The proposed method determines the boundaries between the lobes by deforming initial surfaces. The initial surfaces are automatically determined based on user-given landmarks. They are smoothed and deformed so that the deforming boundaries run along the hourglass portion of the three-dimensional shape of the cerebrum with fuzzy rule-based active contour and surface models. The cerebrum is divided into the cerebral lobes according to the boundaries determined using this method. The reproducibility of our system with a given subject was assessed by examining the variability of volume and surface area in three healthy subjects, with measurements performed by three beginners and one expert user. The experimental results show that our system segments the cerebral lobes with high reproducibility.

Selective atrophy of left hemisphere and frontal lobe of the brain in old men

In this study, volumes of the whole brain, hemispheres, and frontal lobes of young and elderly adults were quantified by an automated method. Effects of age, sex, and side on absolute and relative volumes of the brain structures were evaluated. Compared with the young group, elderly participants showed a 15% volume loss in the whole brain and hemispheres, and a 22% volume loss in the frontal lobes. The relative volume of the left hemisphere in the elderly group decreased more than that of the right hemisphere. Elderly men showed significantly greater left hemisphere and left frontal lobe volume losses than did elderly women, indicating that the larger left hemisphere relative volume reduction is largely contributed to by selective atrophy of the left frontal lobe volume in elderly men. These results may reflect age- and sex-related functional deterioration in the left brain.

Reliable manual segmentation of the frontal, parietal, temporal, and occipital lobes on magnetic resonance images of healthy subjects

BACKGROUND

It is a challenge to reliably measure the lobar volumes from magnetic resonance imaging (MRI) data.

OBJECTIVE

Description of a landmark-based method for volumetric segmentation of the brain into the four cerebral lobes from MR images.

METHOD

The segmentation method relies on a combination of anatomical landmarks and geometrical definitions. The first step, described previously, is a segmentation of the four lobes on the surface of the brain. The internal borders between the lobes are defined on the axial slices of the brain. The intra- and inter- rater reliability was determined from the MRI scans of a group of 10 healthy control subjects measured by 2 independent raters.

RESULTS

The intra-rater relative error (and intra-class correlation coefficient) of the lobar volume measures ranged from 0.81% to 3.85% (from 0.97 to 0.99). The inter-rater relative error (and intra-class correlation coefficient) ranged from 0.55% to 3.09% (from 0.94 to 0.99).

CONCLUSION

This technique has been shown to have high intra- and inter-rater reliability. The current method provides a method to obtain volumetric estimates of the 4 cerebral lobes.

Regional distribution of white matter hyperintensities in vascular dementia, Alzheimer's disease and healthy aging

BACKGROUND

White matter hyperintensities (WMH) on MRI scans indicate lesions of the subcortical fiber system. The regional distribution of WMH may be related to their pathophysiology and clinical effect in vascular dementia (VaD), Alzheimer's disease (AD) and healthy aging.

METHODS

Regional WMH volumes were measured in MRI scans of 20 VaD patients, 25 AD patients and 22 healthy elderly subjects using FLAIR sequences and surface reconstructions from a three-dimensional MRI sequence.

RESULTS

The intraclass correlation coefficient for interrater reliability of WMH volume measurements ranged between 0.99 in the frontal and 0.72 in the occipital lobe. For each cerebral lobe, the WMH index, i.e. WMH volume divided by lobar volume, was highest in VaD and lowest in healthy controls. Within each group, the WMH index was higher in frontal and parietal lobes than in occipital and temporal lobes. Total WMH index and WMH indices in the frontal lobe correlated significantly with the MMSE score in VaD. Category fluency correlated with the frontal lobe WMH index in AD, while drawing performance correlated with parietal and temporal lobe WMH indices in VaD.

CONCLUSIONS

A similar regional distribution of WMH between the three groups suggests a common (vascular) pathogenic factor leading to WMH in patients and controls. Our findings underscore the potential of regional WMH volumetry to determine correlations between subcortical pathology and cognitive impairment.

Neurologic Injury Associated with CABG Surgery: Outcomes, Mechanisms, and Opportunities for Improvement

Neurologic injuries, whether subtle or overt, are a major source of morbidity secondary to coronary artery bypass graft (CABG) surgery. A comprehensive review of research in the area of neurologic injury is provided. We conclude this article by providing insight regarding areas requiring further investigation in order to reduce sustainably the risk of these iatrogenic events among patient undergoing CABG surgery.

Neuroanatomical localisation and clinical correlates of white matter lesions in the elderly

BACKGROUND

White matter lesions (WML) in elderly people co-occur with hypertension, depression, and cognitive impairment. Little is known about the density and distribution of WML in normal elderly people, whether they occur randomly in the aging brain or tend to cluster in certain areas, or whether patterns of WML aggregation are linked to clinical symptoms.

OBJECTIVES

To describe patterns of WML distribution in a large representative population of elderly people using non-inferential cluster analysis; and to determine the extent to which such patterns are associated with clinical symptomatology.

METHOD

A population sample of 1077 elderly people was recruited. Multiple analysis of correspondence followed by automatic classification methods was used to explore overall patterns of WML distribution. Correspondence was then sought between these patterns and a range of cerebrovascular, psychiatric, and neurological symptoms.

RESULTS

Three distinct patterns of spatial localisation within the brain were observed, corresponding to distinct clusters of clinical symptoms. In particular WML aggregation in temporal and occipital areas was associated with greater age, hypertension, late onset depressive disorder, poor global cognitive function, and overall WML frequency.

CONCLUSIONS

WML localisation is not random in the aging brain, and their distribution is associated with age and the presence of clinical symptoms. Age differences suggest there may be patterns of progression across time; however, this requires confirmation from longitudinal imaging studies.

Semiautomatic brain region extraction: a method of parcellating brain regions from structural magnetic resonance images

Structural MR imaging has become essential to the evaluation of regional brain changes in both healthy aging and disease-related processes. Several methods have been developed to measure structure size and regional brain volumes, but many of these methods involve substantial manual tracing and/or landmark identification. We present a new technique, semiautomatic brain region extraction (SABRE), for the rapid and reliable parcellation of cortical and subcortical brain regions. We combine the SABRE parcellation with tissue compartment segmentation [NeuroImage 17 (2002) 1087] to produce measures of gray matter (GM), white matter (WM), ventricular CSF, and sulcal CSF for 26 brain regions. Because SABRE restricts user input to a few easily identified landmarks, inter-rater reliability is high for all volumes, with all coefficients between 0.91 and 0.99. To assess construct validity, we contrasted SABRE-derived volumetric data from healthy young and older adults. Results from the SABRE parcellation and tissue segmentation showed significant differences in multiple brain regions in keeping with regional atrophy described in the literature by researchers using lengthy manual tracing methods. Our findings show that SABRE is a reliable semiautomatic method for assessing regional tissue volumes that provides significant timesavings over purely manual methods, yet maintains information about individual cortical landmarks.

Automated human frontal lobe identification in MR images based on fuzzy-logic encoded expert anatomic knowledge

Identification of human brain structures in MR images comprises an area of increasing interest, which also presents numerous methodological challenges. Here we describe a new knowledge-based automated method designed to identify several major brain sulci and then to define the frontal lobes by using the identified sulci as landmarks. To identify brain sulci, sulcal images were generated by morphologic operations and then separated into different components based on connectivity analysis. Subsequently, the individual anatomic features were evaluated by using fuzzy membership functions. The crisp decisions, i.e., the identification of sulci, were made by taking the maximum of the summation of all the membership functions. The identification was designed in a hierarchical order. The longitudinal fissure was extracted first. The left and right central sulci were then identified based on the left and right hemispheres. Next, the lateral sulci were identified based on the central sulci and hemispheres. Finally, the left and right frontal lobes were defined from the two hemispheres. The method was evaluated by visual inspection, comparison with manual segmentation, and comparison with manually volumetric results in references. The average Jaccard similarities of left and right frontal lobes between the automated and manual segmentation were 0.89 and 0.91, respectively. The average Kappa indices of left and right frontal lobes between the automated and manual segmentation were 0.94 and 0.95, respectively. These results show relatively high accuracy of using this novel method for human frontal lobe identification and segmentation.

Diffusion MRI studies in vascular cognitive impairment and dementia

Since its introduction more than two decades ago, Magnetic Resonance Imaging (MRI) has not only allowed for visualization of the macrostructure of the CNS, but also has been able to study dynamic processes which constitute the substrate of currently available MRI variants. While conventional Diffusion Weighted Imaging (DWI) permits a robust visualization of lesions just a few minutes after the onset of cerebral ischemia, Diffusion Tensor Imaging (DTI) measures the magnitude and direction of diffusion, leading to the characterization of cerebral white matter (WM) microstructural integrity. In this paper, the potential role of MRI techniques, particularly DTI, for the study of the relationship between changes in the microstructural integrity of WM and cognitive impairment in the context of cerebrovascular disease are discussed. Significant correlations between scores of behavioral measures of cognitive function and regional anisotropy values are an example of the potential efficacy of DTI for in vivo studies of brain connectivity in vascular neurodegenerative conditions.

[Volumetric MRI for evaluation of regional pattern and progressin of neocortical degeneration in Alzheimer's disease]

PURPOSE

Volumetric analysis of the corpus callosum and hippocampus using MRI in Alzheimer's disease (AD) to evaluate the regional pattern and progression of neocortical neurodegeneration.

METHODS

In subsequent studies we investigated patients with AD and healthy controls. Volumetry was based on MRI-data from a sagittal 3D T1w-gradient echo sequence. The corpus callosum (CC) was measured in a midsagittal slice, and subdivided into 5 subregions. Volumetry of the hippocampus/amygdala-formation (HAF) was performed by segmentation in coronary reoriented slices.

RESULTS

In AD patients we found a significant atrophy in the rostrum und splenium of CC. The atrophy was correlated with the severity of dementia, but no correlation was found with the load of white matter lesions. In comparison with (18)FDG-PET, we found a significant correlation of regional CC-atrophy with the regional decline of cortical glucose metabolism. A ROC-analysis demonstrated no significant differences in the diagnostic accuracy of HAF volumetry and regional CC volumetry of the splenium (region C5) even in mild stages of dementia.

CONCLUSION

Regional atrophy of CC can be used as a marker of neocortical degeneration even in early stages of dementia in AD.