Lauren Jones is a SURF-funded honors student in the College of Engineering at the University of Arkansas, majoring in Industrial Engineering and Operations Analytics.

Author: Lauren Jones | Major: Industrial Engineering and Operational Analytics | Semester: Spring 2025

My name is Lauren Jones. I am a senior Industrial Engineering and Operations Analytics major at the University of Arkansas College of Engineering. This spring (2025), I began a research project under the mentorship of Dr. Shengfan Zhang in the Department of Industrial Engineering. Our project focuses on applying deep learning to healthcare data to improve early detection of pediatric sepsis. This semester has been centered around literature review and preparation for data access. After spending the summer interning abroad in Rome, Italy, I will return in the fall to complete the model implementation and analysis phase of my research.

Sepsis remains one of the most serious and potentially fatal conditions among children, especially in underserved populations. By applying Long Short-Term Memory (LSTM) neural networks, a powerful class of deep learning models, to Arkansas All-Payer Claims (AAPC) data, our research aims to predict the onset of sepsis earlier than traditional methods allow. Early detection could allow healthcare providers to intervene before symptoms worsen, potentially saving lives and reducing hospital burdens. While this semester focused on foundational research and preparation, the broader goal is to help clinicians in Arkansas (and potentially beyond) monitor vulnerable patients more effectively using advanced artificial intelligence tools.

I chose this research topic because of its intersection between data science, operations research, and real-world impact on public health. Pediatric sepsis is a growing concern, especially in Arkansas, where rural health disparities are significant. The idea of combining machine learning with healthcare resonated with me because of its direct application to saving lives and improving care for some of the most vulnerable patients. I connected with Dr. Shengfan Zhang after learning about her work in healthcare analytics and predictive modeling through our department’s faculty research highlights. After expressing my interest and sharing my background in data analytics, she welcomed me onto her research team and has been an incredible mentor since.

Because access to real-world electronic health record data is sensitive and highly regulated, my first semester of research primarily consisted of reviewing published literature. I explored a wide range of topics: the biology and progression of sepsis, data preprocessing techniques, time-series modeling in healthcare, and the architecture of LSTM networks. I also studied prior research projects that used the AAPC database and learned about the ethical considerations and approval processes for using de-identified patient data.

One of the biggest challenges I faced this semester was navigating the volume of technical literature, particularly in deep learning. Academic papers often assume a strong background in computer science or medicine, so I had to spend extra time breaking down concepts, rewatching lectures, and writing summary notes in plain language. It was humbling but also empowering: I’ve become a stronger independent learner and a more confident technical reader.

Dr. Zhang has played a crucial role in shaping my research process. She not only guided my literature review but also encouraged me to ask hard questions and connect the dots across fields. Our meetings gave me the opportunity to brainstorm modeling approaches, clarify methods I didn’t fully understand, and begin developing an ethical framework for this work. While this semester didn’t involve lab work or travel, I was able to engage with other undergraduate researchers in our department who are also working on industrial engineering applications. Their input helped me feel connected and collaborative even in an otherwise solitary research phase.

This summer, I’ll be participating in an internship abroad in Rome, Italy, an opportunity I’m incredibly excited about. When I return in Fall 2025, I will begin the data analysis phase of my project. Using a cleaned, de-identified dataset from the AAPC database, I will build and train an LSTM model to predict sepsis onset in children. If successful, this model could support early healthcare decisions in pediatric ICU’s

Through the Student Undergraduate Research Fellowship, I have been able to focus my efforts on developing a structured and ethical plan. Ultimately, I hope this experience will shape my future career in healthcare analytics or operations research. This project has shown me how data science can be used not just for prediction, but for ethical, life-saving interventions. I’m excited to see where it leads me next.