eXplainable Deep Intelligence Lab

Partnership and Collaborative Hub

Research Projects

Our research lab is focused on developing cutting-edge technologies in the fields of Computer Vision (CV) and Natural Language Processing (NLP) for Computer Aided Detection (CAD) on local and Cloud computing clusters. Specifically, we are working on applied Deep Learning (DL) for localization and segmentation, sentiment analysis and response generation with the focus on explainability (XAI).

All
CV
NLP
XAI
Cloud

Question Answering from Electronic Health Records (EHR)

Pretrained language models can be utilized to address clinical questions from unstructured free form text in EHR. Providing potential to improve clinical decision making by accessing critical information in EHR notes.

LLMs for Air Traffic Control Safety

Leveraging pretrained language models to enhance operational safety in Air Traffic Control (ATC). The project focuses on overseeing ATC operations and training LLMs using historical data and contextual information provided by a digital twin of the ATC environment.

Fairness in Segmentation

Automatic segmentation offers the potential to help surgeons in preoperative planning. However, analyzing the fairness and any probable racial and gender bias in such models has been very limited. .

VR for Sundown Syndrome

Virtual Reality (VR) offers innovative approaches to support individuals with sundown syndrome, a condition often affecting those with dementia. The project explores creating tailored VR experiences and leveraging user feedback to improve mental well-being and overall quality of life.

Sentiment Analysis

The utilization of sentiment analysis techniques for depression detection can assist in identifying expressed emotions and opinions, enabling the identification of potential threats or opportunities.

Medication Effectiveness

ML models can be trained to predict an outcome of interest from a set of input features. XAI techniques and algorithms can be utilized to shed a light on feature importance to make better clinical decisions based on evaluated treatment efficacy.

Few-Shot Learning

Few-shot learning for segmentation provides the potential to significantly reduce the need for large amounts of annotated data, which is often costly and time-consuming to acquire.

LLMs for Security Testing

This project leverages Large Language Models (LLMs) to enhance security testing processes. By simulating attack scenarios and analyzing system vulnerabilities, the project aims to improve cybersecurity measures while ensuring adaptability with new attack patterns.

Cloud Performance Modeling

Many critical applications are running on the cloud, where scalable infinite processing power is provided. However, due to collocation of Virtual Machine (VM) and performance interferences, performance aware scheduling of critical and time sensitive jobs to the cloud is needed.

Knee Localization

Enabling more precise and efficient analysis of issues and injuries in medical imaging can be facilitated by localizing a region of interest within an image.

Transfer Learning for Retinal Disease Classification

To accurately predict the most significant abnormality among multiple categories of retinal disease, with limited data in each category, state-of-the-art pretrained multi-class models are required.

Hamidreza Moradi, PhD

Lab Director & PI North Carolina A&T State University

Our Research Team

Our research team is highly diverse and consists of individuals with a range of educational backgrounds and experience, including graduate and PhD students, doctors, and faculties, all of whom bring unique perspectives and expertise to the research projects. This diversity in the team ensures that there is a broad range of skills, knowledge, and experience that can be drawn upon to effectively tackle the research questions at hand. The team's diversity also promotes creativity, innovation, and inclusivity, and helps to ensure that the research outcomes are rigorous and well-rounded.

Faculty Members

(Alphabetical Order)

Hamidreza Arabnia, PhD

Academic Advisory Board University of Georgia

William B. Hillegass, MD, PhD

Clinical and Biostatistical Advisory Board University of Mississippi Medical Center

Sally Hodder, MD

Clinical Advisory Board University of West Virginia

Travonia Hughes, PhD

Cognitive Therapy Advisory Board, COAACH Director North Carolina A&T State University

Matthew Morris, PhD

Cognitive Therapy Advisory Board Vanderbilt University Medical Center

Kaushik Roy, PhD

Research Advisory Board North Carolina A&T State University

Shashank Shekhar, MD

Clinical Advisory Board Duke University

Sun Yi, PhD

Research Advisory Board North Carolina A&T State University

Student Members

Ehsan Alam

Graduate Research Assistant North Carolina A&T State University

Stefan Green

Graduate Research Assistant North Carolina A&T State University