A (Sub)field Guide to Quality Control in Hippocampal Subfield Segmentation on Highresolution T2-weighted MRI.

Inquiries into properties of brain structure and function have progressed due to developments in magnetic resonance imaging (MRI). To sustain progress in investigating and quantifying neuroanatomical details in vivo, the reliability and validity of brain measurements are paramount. Quality control (QC) is a set of procedures for mitigating errors and ensuring the validity and reliability of brain measurements. Despite its importance, there is little guidance on best QC practices and reporting procedures. The study of hippocampal subfields in vivo is a critical case for QC because of their small size, inter-dependent boundary definitions, and common artifacts in the MRI data used for subfield measurements. We addressed this gap by surveying the broader scientific community studying hippocampal subfields on their views and approaches to QC. We received responses from 37 investigators spanning 10 countries, covering different career stages, and studying both healthy and pathological development and aging. In this sample, 81% of researchers considered QC to be very important or important, and 19% viewed it as fairly important. Despite this, only 46% of researchers reported on their QC processes in prior publications. In many instances, lack of reporting appeared due to ambiguous guidance on relevant details and guidance for reporting, rather than absence of QC. Here, we provide recommendations for correcting errors to maximize reliability and minimize bias. We also summarize threats to segmentation accuracy, review common QC methods, and make recommendations for best practices and reporting in publications. Implementing the recommended QC practices will collectively improve inferences to the larger population, as well as have implications for clinical practice and public health.

Downstream effects of polypathology on neurodegeneration of medial temporal lobe subregions.

The medial temporal lobe (MTL) is a nidus for neurodegenerative pathologies and therefore an important region in which to study polypathology. We investigated associations between neurodegenerative pathologies and the thickness of different MTL subregions measured using high-resolution post-mortem MRI. Tau, TAR DNA-binding protein 43 (TDP-43), amyloid-ß and α-synuclein pathology were rated on a scale of 0 (absent)-3 (severe) in the hippocampus and entorhinal cortex (ERC) of 58 individuals with and without neurodegenerative diseases (median age 75.0 years, 60.3% male). Thickness measurements in ERC, Brodmann Area (BA) 35 and 36, parahippocampal cortex, subiculum, cornu ammonis (CA)1 and the stratum radiatum lacunosum moleculare (SRLM) were derived from 0.2 × 0.2 × 0.2 mm3 post-mortem MRI scans of excised MTL specimens from the contralateral hemisphere using a semi-automated approach. Spearman's rank correlations were performed between neurodegenerative pathologies and thickness, correcting for age, sex and hemisphere, including all four proteinopathies in the model. We found significant associations of (1) TDP-43 with thickness in all subregions (r = - 0.27 to r = - 0.46), and (2) tau with BA35 (r = - 0.31) and SRLM thickness (r = - 0.33). In amyloid-ß and TDP-43 negative cases, we found strong significant associations of tau with ERC (r = - 0.40), BA35 (r = - 0.55), subiculum (r = - 0.42) and CA1 thickness (r = - 0.47). This unique dataset shows widespread MTL atrophy in relation to TDP-43 pathology and atrophy in regions affected early in Braak stageing and tau pathology. Moreover, the strong association of tau with thickness in early Braak regions in the absence of amyloid-ß suggests a role of Primary Age-Related Tauopathy in neurodegeneration.

Pathological drivers of neurodegeneration in suspected non-Alzheimer's disease pathophysiology.

BACKGROUND: Little is known about the heterogeneous etiology of suspected non-Alzheimer's pathophysiology (SNAP), a group of subjects with neurodegeneration in the absence of ß-amyloid. Using antemortem MRI and pathological data, we investigated the etiology of SNAP and the association of neurodegenerative pathologies with structural medial temporal lobe (MTL) measures in ß-amyloid-negative subjects. METHODS: Subjects with antemortem MRI and autopsy data were selected from ADNI (n=63) and the University of Pennsylvania (n=156). Pathological diagnoses and semi-quantitative scores of MTL tau, neuritic plaques, α-synuclein, and TDP-43 pathology and MTL structural MRI measures from antemortem T1-weighted MRI scans were obtained. ß-amyloid status (A+/A-) was determined by CERAD score and neurodegeneration status (N+/N-) by hippocampal volume. RESULTS: SNAP reflects a heterogeneous group of pathological diagnoses. In ADNI, SNAP (A-N+) had significantly more neuropathological diagnoses than A+N+. In the A- group, tau pathology was associated with hippocampal, entorhinal cortex, and Brodmann area 35 volume/thickness and TDP-43 pathology with hippocampal volume. CONCLUSION: SNAP had a heterogeneous profile with more mixed pathologies than A+N+. Moreover, a role for TDP-43 and tau pathology in driving MTL neurodegeneration in the absence of ß-amyloid was supported.

In Vivo Image-Based 4D Modeling of Competent and Regurgitant Mitral Valve Dynamics.

BACKGROUND: In vivo characterization of mitral valve dynamics relies on image analysis algorithms that accurately reconstruct valve morphology and motion from clinical images. The goal of such algorithms is to provide patient-specific descriptions of both competent and regurgitant mitral valves, which can be used as input to biomechanical analyses and provide insights into the pathophysiology of diseases like ischemic mitral regurgitation (IMR). OBJECTIVE: The goal is to generate accurate image-based representations of valve dynamics that visually and quantitatively capture normal and pathological valve function. METHODS: We present a novel framework for 4D segmentation and geometric modeling of the mitral valve in real-time 3D echocardiography (rt-3DE), an imaging modality used for pre-operative surgical planning of mitral interventions. The framework integrates groupwise multi-atlas label fusion and template-based medial modeling with Kalman filtering to generate quantitatively descriptive and temporally consistent models of valve dynamics. RESULTS: The algorithm is evaluated on rt-3DE data series from 28 patients: 14 with normal mitral valve morphology and 14 with severe IMR. In these 28 data series that total 613 individual 3DE images, each 3D mitral valve segmentation is validated against manual tracing, and temporal consistency between segmentations is demonstrated. CONCLUSIONS: Automated 4D image analysis allows for reliable non-invasive modeling of the mitral valve over the cardiac cycle for comparison of annular and leaflet dynamics in pathological and normal mitral valves. Future studies can apply this algorithm to cardiovascular mechanics applications, including patient-specific strain estimation, fluid dynamics simulation, inverse finite element analysis, and risk stratification for surgical treatment.

Hippocampal subfield pathologic burden in Lewy body diseases vs. Alzheimer's disease.

AIMS: Lewy body diseases (LBD) are characterized by alpha-synuclein (SYN) pathology, but comorbid Alzheimer's disease (AD) pathology is common and the relationship between these pathologies in microanatomic hippocampal subfields is understudied. Here we use digital histological methods to test the association between hippocampal SYN pathology and the distribution of tau and amyloid-beta (Aß) pathology in LBD and contrast with AD subjects. We also correlate pathologic burden with antemortem episodic memory testing. METHODS: Hippocampal sections from 49 autopsy-confirmed LBD cases, 30 with no/low AD copathology (LBD - AD) and 19 with moderate/severe AD copathology (LBD + AD), and 30 AD patients were stained for SYN, tau, and Aß. Sections underwent digital histological analysis of subfield pathological burden which was correlated with antemortem memory testing. RESULTS: LBD - AD and LBD + AD had similar severity and distribution of SYN pathology (P > 0.05), CA2/3 being the most affected subfield (P < 0.02). In LBD, SYN correlated with tau across subfields (R = 0.49, P < 0.001). Tau burden was higher in AD than LBD + AD (P < 0.001), CA1/subiculum and entorhinal cortex (ERC) being most affected regions (P = 0.04 to <0.01). However, tau pathology in LBD - AD was greatest in CA2/3, which was equivalent to LBD + AD. Aß severity and distribution was similar between LBD + AD and AD. Total hippocampal tau and CA2/3 tau was inversely correlated with memory performance in LBD (R = -0.52, -0.69, P = 0.04, 0.009). CONCLUSIONS: Our findings suggest that tau burden in hippocampal subfields may map closely with the distribution of SYN pathology in subfield CA2/3 in LBD diverging from traditional AD and contribute to episodic memory dysfunction in LBD.

Quantitative MRI of Perivascular Spaces at 3T for Early Diagnosis of Mild Cognitive Impairment.

BACKGROUND AND PURPOSE: The limitations inherent in the current methods of diagnosing mild cognitive impairment have constrained the use of early therapeutic interventions to delay the progression of mild cognitive impairment to dementia. This study evaluated whether quantifying enlarged perivascular spaces observed on MR imaging can help differentiate those with mild cognitive impairment from cognitively healthy controls and, thus, have an application in the diagnosis of mild cognitive impairment. MATERIALS AND METHODS: We automated the identification of enlarged perivascular spaces in brain MR Images using a custom quantitative program designed with Matlab. We then quantified the densities of enlarged perivascular spaces for patients with mild cognitive impairment (n = 14) and age-matched cognitively healthy controls (n = 15) and compared them to determine whether the density of enlarged perivascular spaces can serve as an imaging surrogate for mild cognitive impairment diagnosis. RESULTS: Quantified as a percentage of volume fraction (v/v%), densities of enlarged perivascular spaces were calculated to be 2.82 ± 0.40 v/v% for controls and 4.17 ± 0.57 v/v% for the mild cognitive impairment group in the subcortical brain (P < .001), and 2.74 ± 0.57 v/v% for the controls and 3.90 ± 0.62 v/v% for the mild cognitive impairment cohort in the basal ganglia (P < .001). Maximum intensity projections exhibited a visually conspicuous difference in the distributions of enlarged perivascular spaces for a patient with mild cognitive impairment and a control patient. By means of receiver operating characteristic curve analysis, we determined the sensitivity and specificity of using enlarged perivascular spaces as a differentiating biomarker between mild cognitive impairment and controls to be 92.86% and 93.33%, respectively. CONCLUSIONS: The density of enlarged perivascular spaces was found to be significantly higher in those with mild cognitive impairment compared with age-matched healthy control subjects. The density of enlarged perivascular spaces, therefore, may be a useful imaging biomarker for the diagnosis of mild cognitive impairment.

Defining SNAP by cross-sectional and longitudinal definitions of neurodegeneration.

Introduction: Suspected non-Alzheimer's pathophysiology (SNAP) is a biomarker driven designation that represents a heterogeneous group in terms of etiology and prognosis. SNAP has only been identified by cross-sectional neurodegeneration measures, whereas longitudinal measures might better reflect "active" neurodegeneration and might be more tightly linked to prognosis. We compare neurodegeneration defined by cross-sectional 'hippocampal volume' only (SNAP/L-) versus both cross-sectional and longitudinal 'hippocampal atrophy rate' (SNAP/L+) and investigate how these definitions impact prevalence and the clinical and biomarker profile of SNAP in Mild Cognitive Impairment (MCI). Methods: 276 MCI patients from ADNI-GO/2 were designated amyloid "positive" (A+) or "negative" (A-) based on their florbetapir scan and neurodegeneration 'positive' or 'negative' based on cross-sectional hippocampal volume and longitudinal hippocampal atrophy rate. Results: 74.1% of all SNAP participants defined by the cross-sectional definition of neurodegeneration also met the longitudinal definition of neurodegeneration, whereas 25.9% did not. SNAP/L+ displayed larger white matter hyperintensity volume, a higher conversion rate to dementia over 5 years and a steeper decline on cognitive tasks compared to SNAP/L- and the A- CN group. SNAP/L- had more abnormal values on neuroimaging markers and worse performance on cognitive tasks than the A- CN group, but did not show a difference in dementia conversion rate or longitudinal cognition. Discussion: Using a longitudinal definition of neurodegeneration in addition to a cross-sectional one identifies SNAP participants with significant cognitive decline and a worse clinical prognosis for which cerebrovascular disease may be an important driver.

A protocol for manual segmentation of medial temporal lobe subregions in 7 Tesla MRI.

Recent advances in MRI and increasing knowledge on the characterization and anatomical variability of medial temporal lobe (MTL) anatomy have paved the way for more specific subdivisions of the MTL in humans. In addition, recent studies suggest that early changes in many neurodegenerative and neuropsychiatric diseases are better detected in smaller subregions of the MTL rather than with whole structure analyses. Here, we developed a new protocol using 7 Tesla (T) MRI incorporating novel anatomical findings for the manual segmentation of entorhinal cortex (ErC), perirhinal cortex (PrC; divided into area 35 and 36), parahippocampal cortex (PhC), and hippocampus; which includes the subfields subiculum (Sub), CA1, CA2, as well as CA3 and dentate gyrus (DG) which are separated by the endfolial pathway covering most of the long axis of the hippocampus. We provide detailed instructions alongside slice-by-slice segmentations to ease learning for the untrained but also more experienced raters. Twenty-two subjects were scanned (19-32 yrs, mean age = 26 years, 12 females) with a turbo spin echo (TSE) T2-weighted MRI sequence with high-resolution oblique coronal slices oriented orthogonal to the long axis of the hippocampus (in-plane resolution 0.44 × 0.44 mm2) and 1.0 mm slice thickness. The scans were manually delineated by two experienced raters, to assess intra- and inter-rater reliability. The Dice Similarity Index (DSI) was above 0.78 for all regions and the Intraclass Correlation Coefficients (ICC) were between 0.76 to 0.99 both for intra- and inter-rater reliability. In conclusion, this study presents a fine-grained and comprehensive segmentation protocol for MTL structures at 7 T MRI that closely follows recent knowledge from anatomical studies. More specific subdivisions (e.g. area 35 and 36 in PrC, and the separation of DG and CA3) may pave the way for more precise delineations thereby enabling the detection of early volumetric changes in dementia and neuropsychiatric diseases.

Comparison of In Vivo and Ex Vivo MRI of the Human Hippocampal Formation in the Same Subjects.

Multiple techniques for quantification of hippocampal subfields from in vivo MRI have been proposed. Linking in vivo MRI to the underlying histology can help validate and improve these techniques. High-resolution ex vivo MRI can provide an intermediate modality to map information between these very different imaging modalities. This article evaluates the ability to match information between in vivo and ex vivo MRI in the same subjects. We perform rigid and deformable registration on 10 pairs of in vivo (3 T, 0.4 × 0.4 × 2.6 mm3) and ex vivo (9.4 T, 0.2 × 0.2 × 0.2 mm3) scans, and describe differences in MRI appearance between these modalities qualitatively and quantitatively. The feasibility of using this dataset to validate in vivo segmentation is evaluated by applying an automatic hippocampal subfield segmentation technique (ASHS) to in vivo scans and comparing SRLM (stratum/radiatum/lacunosum/moleculare) surface to manual tracing on corresponding ex vivo scans (and in 2 cases, histology). Regional increases in thickness are detected in ex vivo scans adjacent to the ventricles and were not related to scanner, resolution differences, or susceptibility artefacts. Satisfactory in vivo/ex vivo registration and subvoxel accuracy of ASHS segmentation of hippocampal SRLM demonstrate the feasibility of using this dataset for validation, and potentially, improvement of in vivo segmentation methods.

Automated Hippocampal Subfield Segmentation at 7T MRI.

BACKGROUND AND PURPOSE: High resolution 7T MRI is increasingly used to investigate hippocampal subfields in vivo, but most studies rely on manual segmentation which is labor intensive. We aimed to evaluate an automated technique to segment hippocampal subfields and the entorhinal cortex at 7T MRI. MATERIALS AND METHODS: The cornu ammonis (CA)1, CA2, CA3, dentate gyrus, subiculum, and entorhinal cortex were manually segmented, covering most of the long axis of the hippocampus on 0.70-mm(3) T2-weighted 7T images of 26 participants (59 ± 9 years, 46% men). The automated segmentation of hippocampal subfields approach was applied and evaluated by using leave-one-out cross-validation. RESULTS: Comparison of automated segmentations with corresponding manual segmentations yielded a Dice similarity coefficient of >0.75 for CA1, the dentate gyrus, subiculum, and entorhinal cortex and >0.54 for CA2 and CA3. Intraclass correlation coefficients were >0.74 for CA1, the dentate gyrus, and subiculum; and >0.43 for CA2, CA3, and the entorhinal cortex. Restricting the comparison of the entorhinal cortex segmentation to a smaller range along the anteroposterior axis improved both intraclass correlation coefficients (left: 0.71; right: 0.82) and Dice similarity coefficients (left: 0.78; right: 0.77). The accuracy of the automated segmentation versus a manual rater was lower, though only slightly for most subfields, than the intrarater reliability of an expert manual rater, but it was similar to or slightly higher than the accuracy of an expert-versus-manual rater with ∼170 hours of training for almost all subfields. CONCLUSIONS: This work demonstrates the feasibility of using a computational technique to automatically label hippocampal subfields and the entorhinal cortex at 7T MRI, with a high accuracy for most subfields that is competitive with the labor-intensive manual segmentation. The software and atlas are publicly available: http://www.nitrc.org/projects/ashs/.

Fully automatic segmentation of the mitral leaflets in 3D transesophageal echocardiographic images using multi-atlas joint label fusion and deformable medial modeling.

Comprehensive visual and quantitative analysis of in vivo human mitral valve morphology is central to the diagnosis and surgical treatment of mitral valve disease. Real-time 3D transesophageal echocardiography (3D TEE) is a practical, highly informative imaging modality for examining the mitral valve in a clinical setting. To facilitate visual and quantitative 3D TEE image analysis, we describe a fully automated method for segmenting the mitral leaflets in 3D TEE image data. The algorithm integrates complementary probabilistic segmentation and shape modeling techniques (multi-atlas joint label fusion and deformable modeling with continuous medial representation) to automatically generate 3D geometric models of the mitral leaflets from 3D TEE image data. These models are unique in that they establish a shape-based coordinate system on the valves of different subjects and represent the leaflets volumetrically, as structures with locally varying thickness. In this work, expert image analysis is the gold standard for evaluating automatic segmentation. Without any user interaction, we demonstrate that the automatic segmentation method accurately captures patient-specific leaflet geometry at both systole and diastole in 3D TEE data acquired from a mixed population of subjects with normal valve morphology and mitral valve disease.

White matter imaging contributes to the multimodal diagnosis of frontotemporal lobar degeneration.

OBJECTIVE: To evaluate the distribution of white matter (WM) disease in frontotemporal lobar degeneration (FTLD) and Alzheimer disease (AD) and to evaluate the relative usefulness of WM and gray matter (GM) for distinguishing these conditions in vivo. METHODS: Patients were classified as having FTLD (n = 50) or AD (n = 42) using autopsy-validated CSF values of total-tau:ß-amyloid (t-tau:Aß(1-42)) ratios. Patients underwent WM diffusion tensor imaging (DTI) and volumetric MRI of GM. We employed tract-specific analyses of WM fractional anisotropy (FA) and whole-brain GM density analyses. Individual patient classification was performed using receiver operator characteristic (ROC) curves with FA, GM, and a combination of the 2 modalities. RESULTS: Regional FA and GM were significantly reduced in FTLD and AD relative to healthy seniors. Direct comparisons revealed significantly reduced FA in the corpus callosum in FTLD relative to AD. GM analyses revealed reductions in anterior temporal cortex for FTLD relative to AD, and in posterior cingulate and precuneus for AD relative to FTLD. ROC curves revealed that a multimodal combination of WM and GM provide optimal classification (area under the curve = 0.938), with 87% sensitivity and 83% specificity. CONCLUSIONS: FTLD and AD have significant WM and GM defects. A combination of DTI and volumetric MRI modalities provides a quantitative method for distinguishing FTLD and AD in vivo.

3D mesh based wall thickness measurement: identification of left ventricular hypertrophy phenotypes.

Left ventricular hypertrophy (LVH) is a complex cardiac condition mainly identified by the thickening of the myocardial wall. Although most of the contemporary cardiac imaging modalities provide high resolution 3D images, the wall thickness (WT) is still measured within the acquired planes. This way of measurement may introduce an error as cardiac wall is not necessarily orthogonal to the plane. In this study we analyze how different approaches to measure WT can affect an automatic identification of hypertrophy. The compared approaches are: WT measured along surface normal and the one provided by a medial surface. For both approaches we evaluated their ability to identify LVH phenotypes by testing with two classifiers: Transductive Confidence Machine-k Nearest Neighbor (TCM-kNN) and Linear Discriminant Analysis (LDA). Fifty three subjects were included in this study: 18 patients with hypertrophic cardiomyopathy (HCM), 13 patients with hypertensive heart disease (HDD) and 22 sedentary subjects (CG). Medial surface based approach allowed obtaining higher classification accuracy in HDD patients, while normal based approach allowed for higher classification accuracy in HCM patients.

An automated approach to connectivity-based partitioning of brain structures.

We present an automated approach to the problem of connectivity-based partitioning of brain structures using diffusion imaging. White-matter fibres connect different areas of the brain, allowing them to interact with each other. Diffusion-tensor MRI measures the orientation of white-matter fibres in vivo, allowing us to perform connectivity-based partitioning non-invasively. Our new approach leverages atlas-based segmentation to automate anatomical labeling of the cortex. White-matter connectivities are inferred using a probabilistic tractography algorithm that models crossing pathways explicitly. The method is demonstrated with the partitioning of the corpus callosum of eight healthy subjects.

Segmentation, registration, and measurement of shape variation via image object shape.

A model of object shape by nets of medial and boundary primitives is justified as richly capturing multiple aspects of shape and yet requiring representation space and image analysis work proportional to the number of primitives. Metrics are described that compute an object representation's prior probability of local geometry by reflecting variabilities in the net's node and link parameter values, and that compute a likelihood function measuring the degree of match of an image to that object representation. A paradigm for image analysis of deforming such a model to optimize a posteriori probability is described, and this paradigm is shown to be usable as a uniform approach for object definition, object-based registration between images of the same or different imaging modalities, and measurement of shape variation of an abnormal anatomical object, compared with a normal anatomical object. Examples of applications of these methods in radiotherapy, surgery, and psychiatry are given.