The ScholarShip

The ScholarShip is a digital archive for the scholarly output of the ECU community. Its mission is to capture, preserve, and make available the intellectual output of East Carolina University's faculty, staff, and students.

Communities & Collections in the ScholarShip

Select a community to browse its collections.

Now showing 1 - 8 of 8

Recent Submissions

Access status: Restricted ,
Difference in Brain Connectivity and Working Memory in College StudentsBefore and After a Silent Rest Using EEG
(2025-05-27) Allen, Riley; Wozniak, Megan Carol
This pilot study investigated the effects of a silent rest period on neural connectivity and working memory in college students using EEG analysis. Participants completed a series of Backwards Corsi Tasks before and after a five-minute rest while brain activity was monitored. The results showed consistent engagement of the occipital lobe across all trials and notable shifts in parietal and frontal lobe activation after the rest. Post-rest performance showed an increase in working memory scores along with signs of increased cognitive control and decreased mental fatigue in early trials. These findings suggest that short periods of silent rest may enhance cognitive performance and could inform future educational and neurological research.
Access status: Restricted ,
Influence of Cognitive Measures on Social Perception in Parkinson's Disease
(2025-05-01) Dawson, Mary L; Mather, Olivia T.
Parkinson’s Disease (PD) is a progressive neurodegenerative disorder that significantly impacts an individual’s mobility, cognition, and overall quality of life. In addition to motor impairments, non-motor symptoms such as cognitive and communicative impairments significantly affect social interactions for individuals with PD. The current study analyzes the effects of PD on perceiving social interactions and how that relates to cognitive skills. A series of dynamic language and cognitive tests are utilized to observe aspects of non-literal language perception in individuals with PD. Neuropsychological testing of PD participants was conducted using the Cognitive Linguistic Quick Test (CLQT) to assess general cognitive skills, in relation to accuracy on the RISC social perception task. The social communication task consisted of video stimuli pulled from the Relational Inference in Social Communication (RISC) database (Rothermich & Pell, 2015). The participants were to identify the speaker’s intent, including whether the speaker was sincere or insincere. The results show that all participants, both participants with PD and healthy controls (HC), had more difficulties in identifying nonliteral statements compared to literal statements. Participants with PD had a harder time recognizing social intent compared to the HC, which may be due to changes in cognitive abilities. For the PD group, analyses showed positive correlations between many cognitive domains and the social communication task. In summary, our study emphasizes the relationship between cognitive skills and social perception impairments in patients with PD and highlights the need to develop diagnostic tools and ways to treat these impairments.
Access status: Restricted ,
EVALUATING THE POTENTIAL OF AI CHAT BOTS FOR CONSTRUCTION ESTIMATING
(2025-04-30) Lambert, Davis A; Languell, Cole
Artificial intelligence (AI) promises a new age of automation within the U.S. economy. AI leverages advanced machine learning algorithms to predict patterns and manage tasks ranging from simple to complex. AI chatbots like ChatGPT showcase the expanding potential of artificial intelligence by simplifying the implementation of the technology and providing an accessible method for enhancing productivity and reducing workloads. In construction, the application of AI can provide efficiency and reduced timelines in the bidding process compared to traditional methods. The efficiency benefits associated with AI are particularly helpful in cost estimating, where contractors can be more competitive by providing cost expectations faster and with reduced human error. However, the performance and accuracy of the AI output require testing to evaluate the accuracy of the automated results before adoption as an industry standard and incorporation into the bidding process. This study investigates the accuracy of these automated construction outputs through a comparison of AI and traditional approaches to estimating and scheduling. The challenges associated with the incorporation of AI in construction are also investigated to determine the impact of implementation costs, additional AI training requirements, and resistance to change in the industry. After acknowledging the current barriers, the continued development of artificial intelligence and the potential applications of the technology are discussed to reshape the industry, streamline project timelines, and improve overall safety and efficiency in construction operations.
ECU Research Data Inventory
Access status: Open Access ,
Superpatching for Image Analysis Using Transformers and Superpixels
(East Carolina University, July 2025) McCutcheon, Brannon Brannon
Transformers have revolutionized Computer Vision, offering robust performance across diverse tasks. However, their reliance on uniform pixel patching presents limitations, including computational inefficiency for larger images, suboptimal handling of local features, and an inability to process non-uniform patches. Addressing these constraints allows for new opportunities to expand their utility in demanding fields, such as medical imaging. This work proposes a novel architecture combining Convolutional Neural Networks (CNNs) and Transformers to leverage superpixels, clusters of pixels with shared characteristics that capture local feature boundaries effectively. We propose an architecture that segments images into a collection of superpixels, vectorizes these superpixels using a CNN, and passes the resulting tokenized vector representations to a standard Transformer. By removing the uniformity constraint in patching, our approach aims to enhance Transformer performance on tasks requiring large-scale image analysis and fine-grained local feature understanding, potentially opening a way for broader Transformer applications in Computer Vision.