Each year, the Vanderbilt Kennedy Center (VKC) awards small grants to teams consisting of VKC faculty and their trainees and Vanderbilt Data Science Institute (VDSI) faculty and trainees. The IDD-READS (Intellectual and Developmental Disability REsearch And Data Science) Awards are an internal grant mechanism that supports research in intellectual and developmental disabilities that uses innovative data science tools and promotes collaboration among the VKC and VDSI.
Priority is always given to applications that propose innovative, multidisciplinary research and that are most likely to lead to extramural grant support. Applications proposing research that links biomedical (e.g., neuroscience or genetics) variables with behavioral, educational, or policy components (e.g., learning, behavior problems) are especially encouraged.
“The IDD-READS awards are an exciting way to bring together the wonderful VKC community of investigators, who are devoted to research to improving the lives of people with intellectual and developmental disabilities (IDD), and the remarkable and growing community of data scientists affiliated with the VDSI,” said Jeffrey Neul, M.D., Ph.D., Annette Schaffer Eskind Chair and VKC director. “Utilizing the advanced and innovative data science methods being developed in the VDSI to tackle challenging problems in IDD research provides new opportunities to advance our understanding of these disorders and identify methods to make meaningful impact on people with IDD. Additionally, the exposure and education about IDD to trainees in data science is incredibly valuable and we are hopeful that this exposure will cultivate a life-long interest in the use of data science to help accelerate IDD related research.”
Below are the 2022-23 IDD-READS Award recipients, with VKC members denoted with an asterisk.
2022-23 IDD-READS Awards
Mark Wallace, Ph.D.
Towards creating dynamic stimulus databases to uncover mental embeddings in neurotypical and neurodivergent populations
Mark Wallace, Ph.D.* (Hearing & Speech Sciences), David Tovar, M.D., Ph.D. (Neuroscience), [Team: Jessie Spencer-Smith, Ph.D. (Computer Science); VDSI student]
Alterations in sensory processing are commonly seen in autism. These changes are seen not only in responses to the individual senses (e.g., hearing, touch, etc.), but also in how the brain combines information across the different senses. One of the major barriers to the study of sensory function in autism (and other neurodevelopmental conditions) is the lack of rich and diverse sets of naturalistic sensory stimuli to use as probes to better understand how these stimuli are processed in order to create a coherent perceptual view of the world. In the proposed work, the team has the ambitious goal of being the first group to construct expansive stimulus sets across the sensory modalities using artificial intelligence (AI) methods. First steps and the immediate outcome of this IDD-READS project will be determining whether visual stimuli created using such AI methods are comparable to carefully curated visual stimulus sets used in neuroscience labs around the world. The validation procedure will include a combination of behavior, neural activity, and artificial neural networks. In performing this work, the team will seek to open the door to creating auditory, tactile, and audiovisual stimuli at scale using contemporary AI approaches, thus providing a powerful toolbox to probe how sensory stimuli can result in both normal and altered perceptual views of the world.
Rachael Muscatello, Ph.D.
Methods of Quantifying Cardiac Synchrony in Social Dyads
Rachael Muscatello, Ph.D.* (Psychiatry & Behavioral Sciences), Blythe Corbett, Ph.D.* (Psychiatry & Behavioral Sciences) [Team: Ji Noh, VDSI student]
The extent to which two individuals’ heartbeats synchronize during social conversation – interpersonal cardiac synchrony – may promote social cohesion and effective social communication. Many individuals with autism spectrum disorder (ASD) may experience atypical physical responses (e.g., elevated heart rate) to low-threat situations, including peer conversations. Previous research by the study team and others have demonstrated relationships between cardiovascular stress responses and social communication challenges in youth with ASD. However, it is not clear whether coordinated physiological responses, or lack thereof, between youth and a social partner (peer) may influence behavior and quality of interactions. As such, the study team aims to explore and identify best practices for developing cross-correlation models for analyzing cardiac synchrony in school-aged and adolescent youth with ASD during a social interaction. The study further seeks to extend previous methods in the field by using shorter, more precise time scales to capture real-time changes in physiological response to social demands. The proposed approach to analyzing interpersonal physiological synchrony from a real-life social interaction will inform practices for future research utilizing multiple methods of behavior and physiology to more precisely identify the underlying mechanisms of adaptive social functioning.
Miriam Lense, Ph.D.
Multiverse Analysis of EEG Biomarkers in Autism
Miriam Lense, Ph.D.* (Otolaryngology), Noah Fram, Ph.D. (Otolaryngology) [Team: Catie Chang Ph.D. (Computer Science), Zachary Williams (Neuroscience), Tiffany Woynaroski, Ph.D.* (Hearing & Speech Sciences), RA student]
Reliable and objective biologically based measurements (i.e., biomarkers) of social function and communication are needed to advance neuroscience-informed clinical trials in autism. Recent research from the Autism Biomarkers Consortium for Clinical Trials (ABC-CT) identified specific patterns of electrical activity in the brain, recorded using electroencephalography (EEG), as potential biomarkers. However, for EEG biomarkers to be meaningfully incorporated into clinical trials, it is imperative to maximize data quality, standardize and automate data processing, and minimize the number of individuals excluded from studies, longstanding data collection and processing issues in EEG research. In this IDD-READS grant, an interdisciplinary team of VKC and VDSI researchers aims to address this need via multiverse analysis. Multiverse analysis is a novel technique for finding the most accurate way to process neuroimaging data to find a particular pattern or signal if it is present. It uses machine learning techniques to find the best set of parameters to detect each biomarker, including choosing ideal algorithms for complex tasks like removing electrical activity generated by muscle movements. This project will optimize the ways that researchers measure different autism biomarkers and may help us understand the relationships between these biomarkers and behavioral measures such as developmental trajectories or responses to an intervention.
Hakmook Kang, Ph.D.
Quantitative Assessment of Brain-Behavior Associations in Children with Autism Spectrum Disorder via Bayesian Spatio-temporal Models
Hakmook Kang, Ph.D.* (Biostatistics), Carissa Cascio, Ph.D.* (Psychiatry & Behavioral Sciences) [Team: Yuting Mei, VDSI student]
Functional connectivity studies of autism have produced conflicting results and are often not integrated into a conceptual framework that takes symptomatology into account. The current conventional approaches to estimate functional connectivity (FC) tend to ignore the underlying spatial correlation within each region of interest, leading to biased estimator of FC. In turn, this contributes to the problem of producing unreliable results. The goal of this proposal is to employ Bayesian hierarchical spatio-temporal double fusion modeling to improve the ability to detect inter-network connectivity in a sample of individuals with and without autism. The analysis will include neural networks relevant for sensory and attentional features of autism: somatosensory networks, the salience network, and the default mode network. The functional connectivity of these networks will be examined in conjunction with somatosensory processing profiles derived from both questionnaire measures and laboratory experimental testing, providing a rich sensory phenotypic framework within which neural data will be interpreted.
Top photo by Getty Images
