By Katherine Liverman
This past year, three high school students with a shared passion for positive impact came together and designed an app that could change the lives of veterans across the nation.
Ethan Ocasio, a Falls Church local who goes to the New School of Northern Virginia, along with Shreeja Kikkisetti and Neeyanth Kopparapu, found that they all shared a desire to help veterans find medical aid when they met through the Girls Computing League, a nonprofit dedicated to creating equal access to modern technology education.
Kopparapu is the co-founder of GCL with his sister, Kavya Kopparapu. Ocasio remembers playing the Kopparapu siblings in chess tournaments as a kid and reunited with the two upon joining the group about three years ago. Around a year later, Kikkisetti entered the picture; and the three of them began brainstorming on how to change the world through tech.
Ultimately, the trio went on to successfully create an app designed to help veterans access clinical trials and presented it at the U.S. Department of Veteran Affairs’ AI Tech Sprint.
Their app, known as The Clinical Trial Selector, securely collects patient records from the Center for Medicaid and Medicare System, juxtaposes it with trial information from the National Cancer Institute’s Clinical Trials Database and automatically determines trial eligibility for veteran patients. This saves veteran patients not only the time of both manually gathering and inputting their medical information into the system but also the labor in understanding how to retrieve information from these complex systems.
“Our app removes barriers to finding clinical trials by automating the process of looking through a patient’s medical records,” Ocasio states.
In their research, they found that veteran patients are suffering because a complex system and its niche medical lexicon make it difficult to identify clinical trials. So the group set out to streamline the process for the average veteran patient.
“We envisioned how beneficial an accessible clinical trial selector would be to Medicare and Medicaid patients so that they could quickly and easily find matching clinical trials without having specialized medical knowledge,” Kopparapu said.
The app began as a pilot project, and when the group determined its feasibility they pushed forward with the idea; working over school breaks and in their spare free time while still acting as full-time students.
The process of app design is no walk in the park. The group faced the challenges of user-friendly interfaces, codesets and processing features to ensure the accuracy of matching information to trials. One of the main issues they encountered was the accurate and reliable translation of the medical codes that make up patient information.
To preface, “code” is essentially the language of computers. It acts as a set of instructions for a computer to follow. These instructions make up programs, operating systems and mobile apps like The Clinical Trial Selector. Now, things get more complicated when translating codes because, similar to different languages, not all codes have direct translations. Considering the app pulls different medical codes from different databases, the group was faced with the challenge of identifying a consistent and reliable way to translate multiple medical codes. Ultimately, the students were able to use a system that rendered translations in real-time.
“We applied the artificial intelligence technique called Natural Language Processing (NLP) to these descriptions in order to extract computable criteria that we then used to further filter the available trials and improve the matches,” said Ocasio.
Once the group felt confident about the app, they enlisted in the AI Tech Sprint as the sole student group participants, and were selected to present their results on the sprint’s demo day at the U.S. Census Bureau, U.S. Department of Commerce.
The app has not yet reached production, so it cannot be said whether or not it is producing intended results. The group is currently working with a number of organizations to bring the app to production at no cost. Kikkisetti described the group’s goal as “To make it an open-source tool which other organizations could integrate into their own medical systems.”
Looking forward, the group aims to open up the app and expand its accessibility to the general public. Ocasio states: “Although the app has been marketed towards veterans, it could also benefit the general public because it could help trials gain participants which would result in treatments being available to the general public faster.”