Sparse unbiased analysis of anatomical variance in longitudinal imaging

Lesion-symptom mapping with NIHSS sub-scores in ischemic stroke patients

Background

Lesion-symptom mapping (LSM) is a statistical technique to investigate the population-specific relationship between structural integrity and post-stroke clinical outcome. In clinical practice, patients are commonly evaluated using the National Institutes of Health Stroke Scale (NIHSS), an 11-domain clinical score to quantitate neurological deficits due to stroke. So far, LSM studies have mostly used the total NIHSS score for analysis, which might not uncover subtle structure–function relationships associated with the specific sub-domains of the NIHSS evaluation. Thus, the aim of this work was to investigate the feasibility to perform LSM analyses with sub-score information to reveal category-specific structure–function relationships that a total score may not reveal.

Methods

Employing a multivariate technique, LSM analyses were conducted using a sample of 180 patients with NIHSS assessment at 48-hour post-stroke from the ESCAPE trial. The NIHSS domains were grouped into six categories using two schemes. LSM was conducted for each category of the two groupings and the total NIHSS score.

Results

Sub-score LSMs not only identify most of the brain regions that are identified as critical by the total NIHSS score but also reveal additional brain regions critical to each function category of the NIHSS assessment without requiring extensive, specialised assessments.

Conclusion

These findings show that widely available sub-scores of clinical outcome assessments can be used to investigate more specific structure–function relationships, which may improve predictive modelling of stroke outcomes in the context of modern clinical stroke assessments and neuroimaging.

Trial registration number

NCT01778335.

Strengths and challenges of longitudinal non-human primate neuroimaging

Longitudinal non-human primate neuroimaging has the potential to greatly enhance our understanding of primate brain structure and function. Here we describe its specific strengths, compared to both cross-sectional non-human primate neuroimaging and longitudinal human neuroimaging, but also its associated challenges. We elaborate on factors guiding the use of different analytical tools, subject-specific versus age-specific templates for analyses, and issues related to statistical power.

Similarity-driven multi-view embeddings from high-dimensional biomedical data

Diverse, high-dimensional modalities collected in large cohorts present new opportunities for the formulation and testing of integrative scientific hypotheses. Similarity-driven multi-view linear reconstruction (SiMLR) is an algorithm that exploits inter-modality relationships to transform large scientific datasets into smaller, more well-powered and interpretable low-dimensional spaces. SiMLR contributes an objective function for identifying joint signal, regularization based on sparse matrices representing prior within-modality relationships and an implementation that permits application to joint reduction of large data matrices. We demonstrate that SiMLR outperforms closely related methods on supervised learning problems in simulation data, a multi-omics cancer survival prediction dataset and multiple modality neuroimaging datasets. Taken together, this collection of results shows that SiMLR may be applied to joint signal estimation from disparate modalities and may yield practically useful results in a variety of application domains.

Reference-free brain template construction with population symmetric registration

Population registration has been proposed for normalizing a large group of images into a common space, which is important in many clinical and research studies, such as brain development, aging, and atlas construction. Different from pairwise registration problem that aligns the target image to the reference directly, determining the reference or the hidden common space with the least bias is important in population registration. In order to decrease this bias, a lot of work takes the arithmetic mean image as the reference. However, the arithmetic mean image is usually too smooth to guide the population registration. This work presents an efficient symmetric population registration strategy for brain template construction, which defines the symmetric population center guiding population registration. This is important because the population registration problem can be translated into a series of pairwise registration problem which is easier to optimize and implement. Another prominent merit of proposed population registration algorithm is reference-free, which eliminates the reference dependency-related bias in population registration. Based on this symmetric population registration, the brain template is constructed by approximating both the population's intensity and gradient information. In addition, we also present a new measurement named with average bias for evaluating the unbiasedness of brain template. Experiments were first carried out on four synthetic images created with controllable transforms, which aim at comparing the difference between conventional method and proposed method. Further experiment is designed for reference-free validation. Finally, in real inter-subject brain data, twenty MRI T1 volumes with size 256 × 256 × 176 are used to construct a symmetric brain template with proposed population registration method. The constructed brain template has a small bias and clear brain details comparing with DARTEL.

A Comparison of Partial Volume Correction Techniques for Measuring Change in Serial Amyloid PET SUVR

Longitudinal PET studies in aging and Alzheimer's disease populations rely on accurate and precise measurements of change over time from serial PET scans. Various methods for partial volume correction (PVC) are commonly applied to such studies, but existing comparisons and validations of these PVC methods have focused on cross-sectional measurements. Rate of change measurements inherently have smaller magnitudes than cross-sectional measurements, so levels of noise amplification due to PVC must be smaller, and it is necessary to re-evaluate methods in this context. Here we compare the relative precision in longitudinal measurements from serial amyloid PET scans when using geometric transfer matrix (GTM) PVC versus the traditional two-compartment (Meltzer-style), three-compartment (Müller-Gärtner-style), and no-PVC approaches. We used two independent implementations of standardized uptake value ratio (SUVR) measurement and PVC (one in-house pipeline based on SPM12 and ANTs, and one using FreeSurfer 6.0). For each approach, we also tested longitudinal-specific variants. Overall, we found that measurements using GTM PVC had significantly worse relative precision (unexplained within-subject variability ≈4-8%) than those using two-compartment, three-compartment, or no PVC (≈2-4%). Longitudinally-stabilized approaches did not improve these properties. This data suggests that GTM PVC methods may be less suitable than traditional approaches when measuring within-person change over time in longitudinal amyloid PET.

Longitudinal and cross-sectional structural magnetic resonance imaging correlates of AV-1451 uptake

We examined the relationship between in vivo estimates of tau deposition as measured by 18F-AV-1451 tau positron emission tomography imaging and cross-sectional cortical thickness, as well as rates of antecedent cortical thinning measured from magnetic resonance imaging in individuals with and without evidence of cerebral amyloid in 63 participants from the Alzheimer's Disease Neuroimaging Initiative study, including 32 cognitively normal individuals (mean age 74 years), 27 patients with mild cognitive impairment (mean age 76.8 years), and 4 patients diagnosed with Alzheimer's disease (mean age 80 years). We hypothesized that structural measures would correlate with 18F-AV-1451 in a spatially local manner and that this correlation would be stronger for longitudinal compared to cross-sectional measures of cortical thickness and in those with cerebral amyloid versus those without. Cross-sectional and longitudinal estimates of voxelwise atrophy were made from whole brain maps of cortical thickness and rates of thickness change. In amyloid-β-positive individuals, the correlation of voxelwise atrophy across the whole brain with a summary measure of medial temporal lobe (MTL) 18F-AV-1451 uptake demonstrated strong local correlations in the MTL with longitudinal atrophy that was weaker in cross-sectional analysis. Similar effects were seen in correlations between 31 bilateral cortical regions of interest. In addition, several nonlocal correlations between atrophy and 18F-AV-1451 uptake were observed, including association between MTL atrophy and 18F-AV-1451 uptake in parietal lobe regions of interest such as the precuneus. Amyloid-β-negative individuals only showed weaker correlations in data uncorrected for multiple comparisons. While these data replicate previous reports of associations between 18F-AV-1451 uptake and cross-sectional structural measures, the current results demonstrate a strong relationship with longitudinal measures of atrophy. These data support the notion that in vivo measures of tau pathology are tightly linked to the rate of neurodegenerative change.

The pediatric template of brain perfusion

Magnetic resonance imaging (MRI) captures the dynamics of brain development with multiple modalities that quantify both structure and function. These measurements may yield valuable insights into the neural patterns that mark healthy maturation or that identify early risk for psychiatric disorder. The Pediatric Template of Brain Perfusion (PTBP) is a free and public neuroimaging resource that will help accelerate the understanding of childhood brain development as seen through the lens of multiple modality neuroimaging and in relation to cognitive and environmental factors. The PTBP uses cross-sectional and longitudinal MRI to quantify cortex, white matter, resting state functional connectivity and brain perfusion, as measured by Arterial Spin Labeling (ASL), in 120 children 7-18 years of age. We describe the PTBP and show, as a demonstration of validity, that global summary measurements capture the trajectories that demarcate critical turning points in brain maturation. This novel resource will allow a more detailed understanding of the network-level, structural and functional landmarks that are obtained during normal adolescent brain development.

A semi-automated pipeline for the segmentation of rhesus macaque hippocampus: validation across a wide age range

This report outlines a neuroimaging pipeline that allows a robust, high-throughput, semi-automated, template-based protocol for segmenting the hippocampus in rhesus macaque (Macaca mulatta) monkeys ranging from 1 week to 260 weeks of age. The semiautomated component of this approach minimizes user effort while concurrently maximizing the benefit of human expertise by requiring as few as 10 landmarks to be placed on images of each hippocampus to guide registration. Any systematic errors in the normalization process are corrected using a machine-learning algorithm that has been trained by comparing manual and automated segmentations to identify systematic errors. These methods result in high spatial overlap and reliability when compared with the results of manual tracing protocols. They also dramatically reduce the time to acquire data, an important consideration in large-scale neuroradiological studies involving hundreds of MRI scans. Importantly, other than the initial generation of the unbiased template, this approach requires only modest neuroanatomical training. It has been validated for high-throughput studies of rhesus macaque hippocampal anatomy across a broad age range.

Sparse canonical correlation analysis relates network-level atrophy to multivariate cognitive measures in a neurodegenerative population

This study establishes that sparse canonical correlation analysis (SCCAN) identifies generalizable, structural MRI-derived cortical networks that relate to five distinct categories of cognition. We obtain multivariate psychometrics from the domain-specific sub-scales of the Philadelphia Brief Assessment of Cognition (PBAC). By using a training and separate testing stage, we find that PBAC-defined cognitive domains of language, visuospatial functioning, episodic memory, executive control, and social functioning correlate with unique and distributed areas of gray matter (GM). In contrast, a parallel univariate framework fails to identify, from the training data, regions that are also significant in the left-out test dataset. The cohort includes164 patients with Alzheimer's disease, behavioral-variant frontotemporal dementia, semantic variant primary progressive aphasia, non-fluent/agrammatic primary progressive aphasia, or corticobasal syndrome. The analysis is implemented with open-source software for which we provide examples in the text. In conclusion, we show that multivariate techniques identify biologically-plausible brain regions supporting specific cognitive domains. The findings are identified in training data and confirmed in test data.

Converging evidence for the processing costs associated with ambiguous quantifier comprehension

Traditional neuroanatomic models of language comprehension have emphasized a core language network situated in peri-Sylvian cortex. More recent evidence appears to extend the neuroanatomic network beyond peri-Sylvian cortex to encompass other aspects of sentence processing. In this study, we evaluate the neuroanatomic basis for processing the ambiguity in doubly-quantified sentences. For example, a sentence like "All the dogs jumped in a lake" can be interpreted with a collective interpretation (e.g., several dogs jumping into a single lake) or a distributive interpretation (e.g., several dogs each jumping into a different lake). In Experiment 1, we used BOLD fMRI to investigate neuroanatomic recruitment by young adults during the interpretation of ambiguous doubly-quantified sentences in a sentence-picture verification task. We observed that young adults exhibited a processing cost associated with interpreting ambiguous sentences and this was related to frontal and parietal cortex recruitment. In Experiment 2, we investigate ambiguous sentence processing with the identical materials in non-aphasic patients with behavioral variant frontotemporal dementia (bvFTD) who have frontal cortex disease and executive and decision-making limitations. bvFTD patients are insensitive to ambiguity associated with doubly-quantified sentences, and this is related to the magnitude of their frontal cortex disease. These studies provide converging evidence that cortical regions that extend beyond peri-Sylvian cortex help support the processing costs associated with the interpretation of ambiguous doubly-quantified sentences.

Methodological considerations in longitudinal morphometry of traumatic brain injury

Traumatic brain injury (TBI) has recently been reconceptualized as a chronic, evolving disease process. This new view necessitates quantitative assessment of post-injury changes in brain structure that may allow more accurate monitoring and prediction of recovery. In particular, TBI is known to trigger neurodegenerative processes and therefore quantifying progression of diffuse atrophy over time is currently of utmost interest. However, there are various methodological issues inherent to longitudinal morphometry in TBI. In this paper, we first overview several of these methodological challenges: lesion evolution, neurosurgical procedures, power, bias, and non-linearity. We then introduce a sensitive, reliable, and unbiased longitudinal multivariate analysis protocol that combines dimensionality reduction and region of interest approaches. This analysis pipeline is demonstrated using a small dataset consisting of four chronic TBI survivors.

An open source multivariate framework for n-tissue segmentation with evaluation on public data

We introduce Atropos, an ITK-based multivariate n-class open source segmentation algorithm distributed with ANTs ( http://www.picsl.upenn.edu/ANTs). The Bayesian formulation of the segmentation problem is solved using the Expectation Maximization (EM) algorithm with the modeling of the class intensities based on either parametric or non-parametric finite mixtures. Atropos is capable of incorporating spatial prior probability maps (sparse), prior label maps and/or Markov Random Field (MRF) modeling. Atropos has also been efficiently implemented to handle large quantities of possible labelings (in the experimental section, we use up to 69 classes) with a minimal memory footprint. This work describes the technical and implementation aspects of Atropos and evaluates its performance on two different ground-truth datasets. First, we use the BrainWeb dataset from Montreal Neurological Institute to evaluate three-tissue segmentation performance via (1) K-means segmentation without use of template data; (2) MRF segmentation with initialization by prior probability maps derived from a group template; (3) Prior-based segmentation with use of spatial prior probability maps derived from a group template. We also evaluate Atropos performance by using spatial priors to drive a 69-class EM segmentation problem derived from the Hammers atlas from University College London. These evaluation studies, combined with illustrative examples that exercise Atropos options, demonstrate both performance and wide applicability of this new platform-independent open source segmentation tool.

Within-subject template estimation for unbiased longitudinal image analysis

Longitudinal image analysis has become increasingly important in clinical studies of normal aging and neurodegenerative disorders. Furthermore, there is a growing appreciation of the potential utility of longitudinally acquired structural images and reliable image processing to evaluate disease modifying therapies. Challenges have been related to the variability that is inherent in the available cross-sectional processing tools, to the introduction of bias in longitudinal processing and to potential over-regularization. In this paper we introduce a novel longitudinal image processing framework, based on unbiased, robust, within-subject template creation, for automatic surface reconstruction and segmentation of brain MRI of arbitrarily many time points. We demonstrate that it is essential to treat all input images exactly the same as removing only interpolation asymmetries is not sufficient to remove processing bias. We successfully reduce variability and avoid over-regularization by initializing the processing in each time point with common information from the subject template. The presented results show a significant increase in precision and discrimination power while preserving the ability to detect large anatomical deviations; as such they hold great potential in clinical applications, e.g. allowing for smaller sample sizes or shorter trials to establish disease specific biomarkers or to quantify drug effects.

Sparse dose painting based on a dual-pass kinetic-oxygen mapping of dynamic PET images

Development of molecular imaging such as positron-emission tomography (PET) offers an opportunity to optimize radiotherapy treatment planning by conforming the dose distribution to physiological details within tumors, so called dose painting. Quantification of the acquired images and an efficient and practical dose prescription remain two key questions in this field. This paper proposes a novel framework to optimize the dose prescription based on dual-pass modeling of dynamic [18F]FMISO PET images. An optimization algorithm for sparse dose painting (SDP) is developed by minimizing a linear combination of two terms corresponding to the efficiency and total variation of the dose distribution with the constraint of a constant mean dose. Dose efficiency is defined using the linear-quadratic model. The radiosensitivity given by the oxygen tension is estimated using a dual-pass kinetic-oxygen mapping strategy. This is achieved by integrating a realistic [18F]FMISO PET imaging simulation model, which can simulate the distribution of oxygen and tracer under the same tumor microenvironment setting. The algorithm was compared with a typical dose painting by number (DPBN) method in one data set of a patient with head and neck cancer.

Accurate measurement of brain changes in longitudinal MRI scans using tensor-based morphometry

This paper responds to Thompson and Holland (2011), who challenged our tensor-based morphometry (TBM) method for estimating rates of brain changes in serial MRI from 431 subjects scanned every 6 months, for 2 years. Thompson and Holland noted an unexplained jump in our atrophy rate estimates: an offset between 0 and 6 months that may bias clinical trial power calculations. We identified why this jump occurs and propose a solution. By enforcing inverse-consistency in our TBM method, the offset dropped from 1.4% to 0.28%, giving plausible anatomical trajectories. Transitivity error accounted for the minimal remaining offset. Drug trial sample size estimates with the revised TBM-derived metrics are highly competitive with other methods, though higher than previously reported sample size estimates by a factor of 1.6 to 2.4. Importantly, estimates are far below those given in the critique. To demonstrate a 25% slowing of atrophic rates with 80% power, 62 AD and 129 MCI subjects would be required for a 2-year trial, and 91 AD and 192 MCI subjects for a 1-year trial.