Undergraduates shape the future at Watson College with innovative research

At Watson College, under- graduate students are advancing research with real-world impact. Working with faculty mentors, students contribute original ideas, develop technical skills, and gain hands-on experience through collaborative projects and research. The projects highlighted here demonstrate the range of innovation taking place across Watson and the numerous opportunities that Binghamton University has to offer.

Fighting deepfakes with detection tools

As AI-generated images and videos become increasingly realistic, verifying the authenticity of digital media has become a challenge.

Supported by the EXCEED Inclusive Innovation Internship, Mark Cortez ’26, an electrical engineering major, is collaborating with Professor Yu Chen on CerVaLens, a research project that integrates topological frequency analysis with lightweight entropy-based methods to identify AI-generated and manipulated images.

“Our project aims to create a mobile app that can detect whether an image is AI-generated or altered using AI editing tools,” Cortez says. “Eventually, we want to verify audio and video as well.”

Cortez, who transferred to Watson from Mohawk Valley Community College in 2024 and is now pursuing a 4+1 master’s degree focused on cybersecurity, plays a hands-on role in building and testing the system. His work includes automating the creation of synthetic image datasets used to train and evaluate detection models.

“One of the biggest hurdles is cost,” Cortez says. “Some of the AI models we need to generate training data are expensive per token, which forces you to be very intentional about efficiency and optimization.”

Cortez believes the importance of this work is clear. He often scrolls through social media and encounters AI-generated content within seconds. Some examples are fun and harmless, but more realistic deepfakes can be used for fraud, misinformation, and extortion.

“My hope is for the app to be a working prototype that is robust, accounting for varying image generators, to confidently classify static images, audio, and video elements,” Cortez says.

Building a better heart model

Manar Mabrouk ’26 knows she doesn’t want to be a doctor, but she hopes to contribute to the medical field by discovering and building new technology.

As a biomedical engineering student at Watson, she pursued that goal as vice president of the Biomedical Engineering Society and through hands-on research in Professor Gretchen Mahler’s lab.

Mabrouk studied the heart’s aortic valve, which regulates blood flow. As people age, calcium deposits can build up in and around the valve, impairing movement and circulation. Current treatment often requires invasive heart surgery, which is risky for older patients.

Mabrouk and her team developed an in-vitro model to mimic both healthy and diseased conditions, hoping to understand how heart tissue changes over time and develop new treatments.

She learned that research is never linear — it requires trial and error that can prolong the process.

“It’s definitely the evolving learning ideology that is really cool,” she says. “You never get bored in the lab. There’s always something that pops up, whether it be an issue with your experiment or a contamination issue. There’s always something that you have to troubleshoot, and it keeps life interesting while teaching me new skills.”

Mabrouk’s research was supported by the Collegiate Science and Technology Entry Program (CSTEP), a program that helps students from different backgrounds succeed in STEM fields through academic enrichment and research experience.

While at Watson, she also attended the national Biomedical Engineering Society conference, which was a transformative experience: “It was fun to be in a space with so many smart and talented individuals doing phenomenal research.”

After Binghamton, Mabrouk plans to attend Boston University to pursue a master’s in biomedical engineering to work in research and development.

Searching for new cancer clues

Triple-negative breast cancer (TNBC) is one of the most aggressive forms of breast cancer and remains especially difficult to treat because of its resistance to hormone-based therapies.

Deen Kaakour ’26, a biomedical engineering major, studied this disease in Associate Professor Fake “Frank” Lu’s lab, where his research focused on a lesser-known phenomenon observed in TNBC tumors: ossification, or the formation of bone-like deposits.

“I find this work meaningful because of the impact it could have on the future of oncology,” Kaakour says. “Breast cancer is a terrible disease, and I know people in my life personally who have been affected by it.”

Kaakour studied how and why ossification occurs by growing cancer cells under specialized conditions and observing how they changed using advanced imaging technology. The latest tool in Lu’s lab is stimulated Raman scattering microscopy, which analyzes how cells reflect light on a quantum scale to identify anomalies.

“This type of microscopy has not yet been applied to this model in any research, so it provides new insights into cellular changes in real time,” he says.

After observing positive results in two-dimensional cell cultures, Kaakour plans to expand the project to three-dimensional organoids that include additional cell types found in real tumors. His long-term goal is to explore whether inducing controlled cell differentiation could reduce tumor aggressiveness and lessen dependence on chemotherapy.

“If the work I am doing, even in some minor way, could positively impact patient care in the future, I feel really proud knowing that I was a part of something larger than me,” Kaakour says.

Accelerating drug discovery

Understanding how molecules behave during chemical reactions is central to fields ranging from drug development to materials science.

In the School of Computing, undergraduate researcher Obadiah “Obie” Smolenski — a triple-major in computer science, physics, and mathematics — is helping advance this work by applying machine learning to molecular discovery.

Working with Associate Professor Kenneth Chiu, Smolenski contributes to a project that uses machine learning to model quantum chemistry at a scale that would be impractical using traditional methods.

“Our goal is to scale up the current state-of-the-art models to use them to simulate chemical reactions at a realistic and useful scale,” he says.

Conventional quantum chemistry calculations can take days to evaluate a single molecule, limiting how many compounds can be tested. Machine learning, however, dramatically accelerates that process.

“With machine learning, this can be sped up from several days to less than a second per molecule,” Smolenski says, allowing researchers to test far more molecules for desirable drug-like properties.

In the long term, this approach could improve the efficiency and quality of drug development.

“This could lead to the discovery of drug treatments with any number of useful properties, such as being easier to make in the lab or having fewer side effects,” he says.

The project is part of a broader collaboration between Binghamton University and Lawrence Livermore National Laboratory in California, and the team’s findings were published at NeurIPS, a leading computer science conference. In December, Smolenski also presented the work at a major conference in San Diego, an opportunity that underscored the real-world relevance of the research.

“It was very rewarding,” he says. “I’m proud of all the work that I put into seeing it through.”

Improving electronics reliability

Modern electronics rely on microscopic components being securely bonded together, making reliability a critical concern in semiconductor manufacturing.

Malcolm John, a mechanical engineering student, is working with Professor Junghyun Cho on advanced semiconductor packaging research, with a focus on evaluating the mechanical strength of hybrid bonding.

“Bond strength is important in hybrid bonding because it determines whether stacked chips can stay together through everyday use without cracking or separating,” John says.

John’s research examines how factors such as contamination, surface preparation, and material behavior affect bond strength. He focuses on mechanical testing and failure analysis to understand how and where bonds fail, information that can help engineers improve manufacturing processes and device durability.

To support this work, John designed custom fixtures for a double cantilever beam test, commonly used to evaluate bond strength.

“At the time of me joining the research group, we didn’t have fixtures to run these types of tests,” John says. “The fixtures I designed allow us to run this specific type of test using a universal testing machine that the group already owns.”

The fixtures were designed using Autodesk Inventor and manufactured through the Watson Fabrication Lab, allowing the research group to perform tests using existing equipment. At the same time, John says the experience has strengthened his understanding of core engineering concepts before encountering them in class.

“Being exposed to topics like bonding, surface preparation, and material behavior in a real research setting helped me understand them more deeply once they appeared in lectures,” he says.

Helping new moms get proper care

Getting proper medical treatment is essential during pregnancy and after the baby is born — but maternity care in a rural region can face extra challenges.

A senior capstone project at the School of Systems Science and Industrial Engineering wants to improve the patient experience and close the gap between the help that mothers need and what they actually receive.

Partnering with United Health Services’ Chenango Memorial Hospital, the students surveyed doctors, staff, and patients about three key areas: communication clarity and consistency across care teams; coordination and continuity across prenatal, delivery, and postpartum care; and patient education and hospital discharge preparedness.

“We found a communication barrier between patients and their providers, which can be a problem even outside of pregnancy care,” says Loren Cuomo ’26. “Sometimes doctors can be in a rush — they have very busy schedules.”

The Albany native felt drawn to industrial and systems engineering during Watson’s Engineering Design Division program, which gives first-year students a taste of the disciplines that the college offers: “I liked how it’s a mix of business and engineering. You can do so much with it. Also, I’ve always liked figuring out the process of things and how to make something better.”

For the hospital project, Cuomo and her teammates — Anna Jiang ’26, Isaac Joseph ’26, Jonathan Lazo ’26, and Prithvi Noronha ’26 — worked on a solution with Assistant Professors Kimberly Harry and Safa ElKefi and UHS Coordinator of Strategic Initiatives and Performance Improvement Keana Schrag.

Because issues like new public transportation options are outside of the students’ ability to fix, they built a detailed website to address patients’ other concerns. It includes contact information, a trimester-by-trimester pregnancy guide, tips on postpartum care, and a community section where mothers can post questions and connect with one another.

Although the students have graduated (and Cuomo starts working at Regeneron Pharmaceuticals’ warehouse operation this summer), they handed over their data and website to UHS, and the hospital plans to integrate them into its maternity program.

“It’s nice to have an opportunity where you’re doing something meaningful,” Cuomo says. “Working with pregnant women is just so special. You have to be careful — they’re trusting you with their healthcare processes.”

— Christina Forte '27 and Chris Kocher contributed to this story.