A fully Bayesian approach for comprehensive mapping of magnitude and phase brain activation in complex-valued fMRI data.

Functional magnetic resonance imaging (fMRI) plays a crucial role in neuroimaging, enabling the exploration of brain activity through complex-valued signals. These signals, composed of magnitude and phase, offer a rich source of information for understanding brain functions. Traditional fMRI analyses have largely focused on magnitude information, often overlooking the potential insights offered by phase data. In this paper, we propose a novel fully Bayesian model designed for analyzing single-subject complex-valued fMRI (cv-fMRI) data. Our model, which we refer to as the CV-M&P model, is distinctive in its comprehensive utilization of both magnitude and phase information in fMRI signals, allowing for independent prediction of different types of activation maps. We incorporate Gaussian Markov random fields (GMRFs) to capture spatial correlations within the data, and employ image partitioning and parallel computation to enhance computational efficiency. Our model is rigorously tested through simulation studies, and then applied to a real dataset from a unilateral finger-tapping experiment. The results demonstrate the model's effectiveness in accurately identifying brain regions activated in response to specific tasks, distinguishing between magnitude and phase activation.

Construction of a corresponding empirical model to bridge thermal properties and synthesis of thermoresponsive poloxamines.

The thermoresponsive properties of poloxamine (tetra-branch PEO-PPO block copolymer) hydrogels are related to several variables. Of particular interest to this study were the molecular weight of the polymer, the molar ratio between PEO and PPO blocks, and the concentration of the aqueous solution. Accurately controlling the thermoresponsive behaviors of the polymer is critical to the application of such materials; therefore, the structure-property relationship of tetra-branch PEO-PPO block copolymer was studied by synthesis via anionic ring-opening polymerization (AROP). The structure-property relationships were studied by measuring the thermoresponsive behavior via differential scanning calorimetry (DSC) and developing an empirical model which statistically fit the collected data. This empirical model was then used for designing poloxamines that have critical micellization temperatures (CMT) between room temperature and physiological temperature. The model was validated with three polymers that targeted a CMT of 308 K (35°C). The empirical model showed great success in guiding the synthesis of poloxamines showing a temperature difference of less than 3 K between the predicted and the observed CMTs. This study showed a great potential of using an empirical model to set synthesis parameters to control the properties of the polymer products.

Bayesian Spatial Binary Regression for Label Fusion in Structural Neuroimaging.

Alzheimer's disease is a neurodegenerative condition that accelerates cognitive decline relative to normal aging. It is of critical scientific importance to gain a better understanding of early disease mechanisms in the brain to facilitate effective, targeted therapies. The volume of the hippocampus is often used in diagnosis and monitoring of the disease. Measuring this volume via neuroimaging is difficult since each hippocampus must either be manually identified or automatically delineated, a task referred to as segmentation. Automatic hippocampal segmentation often involves mapping a previously manually segmented image to a new brain image and propagating the labels to obtain an estimate of where each hippocampus is located in the new image. A more recent approach to this problem is to propagate labels from multiple manually segmented atlases and combine the results using a process known as label fusion. To date, most label fusion algorithms employ voting procedures with voting weights assigned directly or estimated via optimization. We propose using a fully Bayesian spatial regression model for label fusion that facilitates direct incorporation of covariate information while making accessible the entire posterior distribution. Our results suggest that incorporating tissue classification (e.g, gray matter) into the label fusion procedure can greatly improve segmentation when relatively homogeneous, healthy brains are used as atlases for diseased brains. The fully Bayesian approach also produces meaningful uncertainty measures about hippocampal volumes, information which can be leveraged to detect significant, scientifically meaningful differences between healthy and diseased populations, improving the potential for early detection and tracking of the disease.

Sampling Strategies for Fast Updating of Gaussian Markov Random Fields.

Gaussian Markov random fields (GMRFs) are popular for modeling dependence in large areal datasets due to their ease of interpretation and computational convenience afforded by the sparse precision matrices needed for random variable generation. Typically in Bayesian computation, GMRFs are updated jointly in a block Gibbs sampler or componentwise in a single-site sampler via the full conditional distributions. The former approach can speed convergence by updating correlated variables all at once, while the latter avoids solving large matrices. We consider a sampling approach in which the underlying graph can be cut so that conditionally independent sites are updated simultaneously. This algorithm allows a practitioner to parallelize updates of subsets of locations or to take advantage of 'vectorized' calculations in a high-level language such as R. Through both simulated and real data, we demonstrate computational savings that can be achieved versus both single-site and block updating, regardless of whether the data are on a regular or an irregular lattice. The approach provides a good compromise between statistical and computational efficiency and is accessible to statisticians without expertise in numerical analysis or advanced computing.

Regional and Local Temporal Trends of Borrelia burgdorferi and Anaplasma spp. Seroprevalence in Domestic Dogs: Contiguous United States 2013-2019.

In 2019, in the United States, over 220,000 and 350,000 dogs tested positive for exposure to Anaplasma spp. and Borrelia burgdorferi, respectively. To evaluate regional and local temporal trends of pathogen exposure we used a Bayesian spatio-temporal binomial regression model, analyzing serologic test results for these pathogens from January 2013 to December 2019. Regional trends were not static over time, but rather increased within and beyond the borders of historically endemic regions. Increased seroprevalence was observed as far as North Carolina and North Dakota for both pathogens. Local trends were estimated to evaluate the heterogeneity of underlying changes. A large cluster of counties with increased B. burgdorferi seroprevalence centered around West Virginia, while a similar cluster of counties with increased Anaplasma spp. seroprevalence centered around Pennsylvania and extended well into Maine. In the Midwest, only a small number of counties experienced an increase in seroprevalence; instead, most counties had a decrease in seroprevalence for both pathogens. These trends will help guide veterinarians and pet owners in adopting the appropriate preventative care practices for their area. Additionally, B. burgdorferi and A. phagocytophilum cause disease in humans. Dogs are valuable sentinels for some vector-borne pathogens, and these trends may help public health providers better understand the risk of exposure for humans.

Local and regional temporal trends (2013-2019) of canine Ehrlichia spp. seroprevalence in the USA.

BACKGROUND: In the USA, there are several Ehrlichia spp. of concern including Ehrlichia canis, Ehrlichia ewingii, Ehrlichia chaffeensis, Ehrlichia muris eauclarensis, and "Panola Mountain Ehrlichia". Of these, E. canis is considered the most clinically relevant for domestic dogs, with infection capable of causing acute, subclinical, and chronic stages of disease. Changes in climate, land use, habitats, and wildlife reservoir populations, and increasing contact between both human and dog populations with natural areas have resulted in the increased risk of vector-borne disease throughout the world. METHODS: A Bayesian spatio-temporal binomial regression model was applied to serological test results collected from veterinarians throughout the contiguous USA between January 2013 and November 2019. The model was used to quantify both regional and local temporal trends of canine Ehrlichia spp. seroprevalence and identify areas that experienced significant increases in seroprevalence. RESULTS: Regionally, increasing seroprevalence occurred within several states throughout the central and southeastern states, including Missouri, Arkansas, Mississippi, Alabama, Virginia, North Carolina, Georgia and Texas. The underlying local trends revealed increasing seroprevalence at a finer scale. Clusters of locally increasing seroprevalence were seen from the western Appalachian region into the southern Midwest, along the Atlantic coast in New England, parts of Florida, Illinois, Wisconsin and Minnesota, and in a couple areas of the Mountain region. Clusters of locally decreasing seroprevalence were seen throughout the USA including New York and the mid-Atlantic states, Texas, the Midwest, and California. CONCLUSIONS: Canine Ehrlichia spp. seroprevalence is increasing in both endemic and non-endemic areas of the USA. The findings from this study indicate that dogs across a wide area of the USA are at risk of exposure and these results should provide veterinarians and pet owners with the information they need to make informed decisions about prevention of tick exposure.

Regional and local temporal trends in the prevalence of canine heartworm infection in the contiguous United States: 2012-2018.

BACKGROUND: Canine heartworm disease is a potentially fatal disease for which treatment is financially burdensome for many pet owners. Prevention is strongly advocated by the veterinary community along with routine testing for infection during annual wellness examinations. Despite the availability of efficacious chemoprophylaxis, recent reports have suggested that the incidence of heartworm disease in domestic dogs is increasing. RESULTS: Using data from tests for heartworm infection in the USA from January 2012 through September 2018, a Bayesian spatio-temporal binomial regression model was used to estimate the regional and local temporal trends of heartworm infection prevalence. The area with the largest increase in regional prevalence was found in the Lower Mississippi River Valley. Regional prevalence increased throughout the southeastern states and northward into Illinois and Indiana. Local (county-level) prevalence varied across the USA, with increasing prevalence occurring along most of the Atlantic coast, central United States, and western states. Clusters of decreasing prevalence were present along the Mississippi Alluvial Plain (a historically endemic area), Oklahoma and Kansas, and Florida. CONCLUSIONS: Canine heartworm infection prevalence is increasing in much of the USA, both regionally and locally, despite veterinarian recommendations on prevention and testing. Additional steps should be taken to protect dogs, cats and ferrets. Further work is needed to identify the driving factors of the locally decreasing prevalence present along the Mississippi Alluvial plain, Florida, and other areas.

Incorporating spatial dependence into Bayesian multiple testing of statistical parametric maps in functional neuroimaging.

The analysis of functional neuroimaging data often involves the simultaneous testing for activation at thousands of voxels, leading to a massive multiple testing problem. This is true whether the data analyzed are time courses observed at each voxel or a collection of summary statistics such as statistical parametric maps (SPMs). It is known that classical multiplicity corrections become strongly conservative in the presence of a massive number of tests. Some more popular approaches for thresholding imaging data, such as the Benjamini-Hochberg step-up procedure for false discovery rate control, tend to lose precision or power when the assumption of independence of the data does not hold. Bayesian approaches to large scale simultaneous inference also often rely on the assumption of independence. We introduce a spatial dependence structure into a Bayesian testing model for the analysis of SPMs. By using SPMs rather than the voxel time courses, much of the computational burden of Bayesian analysis is mitigated. Increased power is demonstrated by using the dependence model to draw inference on a real dataset collected in a fMRI study of cognitive control. The model also is shown to lead to improved identification of neural activation patterns known to be associated with eye movement tasks.