Effects of Head Motion on the Evaluation of Age-related Brain Network Changes Using Resting State Functional MRI

Reproducibility of Diffusion MRI-Based Tractography in the Fetal Brain

BACKGROUND

Tractography based on diffusion MRI (dMRI) is a useful tool to study white matter of the developing brain. However, its application in fetal brains is limited due to motion artifacts and low resolution of in utero dMRI, leading to reduced reliability, which was scarcely investigated in previous studies.

PURPOSE

To identify reliably traceable fibers in fetal brains and assess whether reproducibility varies with gestational age (GA) and varies between brain regions.

STUDY TYPE

Prospective cohort study.

SUBJECTS

A total of 44 healthy fetuses with GAs between 25 and 37 (31 ± 6).

FIELD STRENGTH/SEQUENCE

3-T, diffusion-weighted echo-planar imaging sequence (2-5 repeated dMRI scans within the same session per subject).

ASSESSMENT

We fitted dMRI with constrained spherical deconvolution model and conducted tractography on eight fibers. We extracted volume, fractional anisotropy, and fiber count for each fiber and assessed the reproducibility of these metrics between repeated scans within each subject. Data were divided into two age-based subgroups (≤30 weeks, N = 28, and >30 weeks, N = 16) for further tests.

STATISTICAL TESTS

The reproducibility were compared between fibers by analysis of variance and two-sample t tests. Multiple comparisons were corrected by the false discovery rate (5% was accepted).

RESULTS

The reproducibility of the anterior thalamic radiation, inferior longitudinal fasciculus (ILF), genu of the corpus callosum (GCC), and body of the corpus callosum (BCC) significantly decreased with advancing GA (correlation coefficient = 0.525-0.823), as confirmed by group comparisons between fetuses in early GA (≤30 weeks) and late GA (>30 weeks) groups. Corticospinal tract, inferior fronto-occipital fasciculus, and GCC showed high reproducibility for fiber count (weighted dice average = 0.846 vs. 0.814), while BCC and ILF exhibited the lowest reproducibility in both age groups.

DATA CONCLUSION

The study indicates that the reliability of fetal brain tractography depends on GA and varies among different fibers.

LEVEL OF EVIDENCE

2 TECHNICAL EFFICACY: Stage 2.

Changes in white matter functional networks across late adulthood

Introduction

The aging brain is characterized by decreases in not only neuronal density but also reductions in myelinated white matter (WM) fibers that provide the essential foundation for communication between cortical regions. Age-related degeneration of WM has been previously characterized by histopathology as well as T2 FLAIR and diffusion MRI. Recent studies have consistently shown that BOLD (blood oxygenation level dependent) effects in WM are robustly detectable, are modulated by neural activities, and thus represent a complementary window into the functional organization of the brain. However, there have been no previous systematic studies of whether or how WM BOLD signals vary with normal aging. We therefore performed a comprehensive quantification of WM BOLD signals across scales to evaluate their potential as indicators of functional changes that arise with aging.

Methods

By using spatial independent component analysis (ICA) of BOLD signals acquired in a resting state, WM voxels were grouped into spatially distinct functional units. The functional connectivities (FCs) within and among those units were measured and their relationships with aging were assessed. On a larger spatial scale, a graph was reconstructed based on the pair-wise connectivities among units, modeling the WM as a complex network and producing a set of graph-theoretical metrics.

Results

The spectral powers that reflect the intensities of BOLD signals were found to be significantly affected by aging across more than half of the WM units. The functional connectivities (FCs) within and among those units were found to decrease significantly with aging. We observed a widespread reduction of graph-theoretical metrics, suggesting a decrease in the ability to exchange information between remote WM regions with aging.

Discussion

Our findings converge to support the notion that WM BOLD signals in specific regions, and their interactions with other regions, have the potential to serve as imaging markers of aging.

Latent Similarity Identifies Important Functional Connections for Phenotype Prediction

OBJECTIVE

Endophenotypes such as brain age and fluid intelligence are important biomarkers of disease status. However, brain imaging studies to identify these biomarkers often encounter limited numbers of subjects but high dimensional imaging features, hindering reproducibility. Therefore, we develop an interpretable, multivariate classification/regression algorithm, called Latent Similarity (LatSim), suitable for small sample size but high feature dimension datasets.

METHODS

LatSim combines metric learning with a kernel similarity function and softmax aggregation to identify task-related similarities between subjects. Inter-subject similarity is utilized to improve performance on three prediction tasks using multi-paradigm fMRI data. A greedy selection algorithm, made possible by LatSim's computational efficiency, is developed as an interpretability method.

RESULTS

LatSim achieved significantly higher predictive accuracy at small sample sizes on the Philadelphia Neurodevelopmental Cohort (PNC) dataset. Connections identified by LatSim gave superior discriminative power compared to those identified by other methods. We identified 4 functional brain networks enriched in connections for predicting brain age, sex, and intelligence.

CONCLUSION

We find that most information for a predictive task comes from only a few (1-5) connections. Additionally, we find that the default mode network is over-represented in the top connections of all predictive tasks.

SIGNIFICANCE

We propose a novel prediction algorithm for small sample, high feature dimension datasets and use it to identify connections in task fMRI data. Our work can lead to new insights in both algorithm design and neuroscience research.

Resting State Networks Related to the Maintenance of Good Cognitive Performance During Healthy Aging

Purpose: Maintenance of cognitive performance is important for healthy aging. This study aims to elucidate the relationship between brain networks and cognitive function in subjects maintaining relatively good cognitive performance.

Methods: A total of 120 subjects, with equal number of participants from each age group between 20 and 70 years, were included in this study. Only participants with Addenbrooke’s Cognitive Examination – Revised (ACE-R) total score greater than 83 were included. Anatomical T1-weighted MR images and resting-state functional MR images (rsfMRIs) were taken from all participants using a 3-tesla MRI scanner. After preprocessing, several factors associated with age including the ACE-R total score, scores of five domains, sub-scores of ACE-R, and brain volumes were tested. Morphometric changes associated with age were analyzed using voxel based morphometry (VBM) and changes in resting state networks (RSNs) were examined using dual regression analysis.

Results: Significant negative correlations with age were seen in the total gray matter volume (GMV, r = −0.58), and in the memory, attention, and visuospatial domains. Among the different sub-scores, the score of the delayed recall (DR) showed the highest negative correlation with age (r = −0.55, p < 0.001). In VBM analysis, widespread regions demonstrated negative correlation with age, but none with any of the cognitive scores. Quadratic approximations of cognitive scores as functions of age showed relatively delayed decline compared to total GMV loss. In dual regression analysis, some cognitive networks, including the dorsal default mode network, the lateral dorsal attention network, the right / left executive control network, the posterior salience network, and the language network, did not demonstrate negative correlation with age. Some regions in the sensorimotor networks showed positive correlation with the DR, memory, and fluency scores.

Conclusion: Some domains of the cognitive test did not correlate with age, and even the highly correlated sub-scores such as the DR score, showed delayed decline compared to the loss of total GMV. Some RSNs, especially involving cognitive control regions, were relatively maintained with age. Furthermore, the scores of memory, fluency, and the DR were correlated with the within-network functional connectivity values of the sensorimotor network, which supported the importance of exercise for maintenance of cognition.