Mom's Cancer Battle Inspires Teen To Develop AI Detection Tool

VERNON HILLS, IL — When Aarnav Garg's mother was diagnosed with stage IV ovarian cancer about five years ago, he was just 10 years old.

Watching her battle the disease would eventually inspire Garg, who is now 16 years old and a junior at Vernon Hills High School, to combine his passion for computer science with a deeply personal mission: developing technology that could help improve ovarian cancer detection.

Garg has since earned national recognition for OvaScan, an artificial intelligence system he developed to analyze ovarian ultrasound images and identify tumors. Garg was recently named a 2026 Davidson Fellows Honorable Mention for the project, making him one of two Illinois students to receive the distinction.

"Around 5 years ago, my mother was diagnosed with stage IV ovarian cancer," Garg told Patch. "Seeing her journey through recovery inspired me to take on a project dealing with this lethal disease."

Garg said his interest in computer science and a desire to make a meaningful difference led him to explore how artificial intelligence could be used in medical imaging.

"After reading several recently published research papers and identifying gaps in the field, I decided to pursue a project at the intersection of AI and ovarian cancer," he said.

The resulting technology, OvaScan, uses two deep-learning models to analyze ultrasound images. The first identifies an area that may contain a tumor, while the second creates a more precise outline of it.

Garg developed the system as a prototype iPhone application that allows users to upload ultrasound scans and receive tumor outlines. The technology also extracts numerical information about the identified areas that could help researchers better understand the images, according to Garg.

The project took about seven months to complete, beginning during the summer of 2025 and concluding in March 2026.

Garg said much of that time was spent learning unfamiliar technology, troubleshooting problems and refining the system.

"This project pushed me to navigate a steep learning curve," he said. "The biggest challenge I faced during the study was entering an unfamiliar field with little experience, particularly in medical imaging and deep learning segmentation."

He also encountered technical challenges while adapting the artificial intelligence models for use on an iPhone, including memory limitations and difficulties converting the models to run on a mobile device.

Despite those obstacles, Garg said his research produced promising results.

His best model achieved a Dice score of 0.916, a measurement of how closely the computer-generated tumor outline matches a reference outline. A score of 1 represents a perfect match. Another model achieved a mean average precision score of 0.935 at a specified detection threshold, reflecting its ability to locate areas of interest in ultrasound images.

Garg said the mobile application can complete its analysis in less than 300 milliseconds. Although the results are encouraging, OvaScan remains a research prototype and has not been validated for clinical diagnosis or routine patient care.

Garg completed most of the work independently but received feedback from a professor at Northwestern University's Feinberg School of Medicine and a faculty member at the Illinois Mathematics and Science Academy.

His research has already earned recognition at both the state and national levels.

In April, Garg was named the 2026 Illinois BioGENEius Champion, winning a statewide research competition organized by the Illinois Biotechnology Innovation Organization. He received a $1,500 prize for his project, which was selected from a field of 33 semifinalists.

More recently, he was named a Davidson Fellows Honorable Mention in the technology category. The national program recognizes young people who have completed significant work in science, technology, engineering, mathematics and other fields.

"Being named a Davidson Fellows Honorable Mention means a lot to me because it recognizes the work I put into developing OvaScan," Garg said. "The project required many hours of reading research papers, debugging code, testing different models, and refining the pipeline."

Garg said he is especially proud of how much he learned while tackling a field in which he initially had little experience.

His work was also accepted for presentation at MICAD 2026, an international conference focused on medical imaging and computer-aided diagnosis. Looking ahead, Garg hopes to continue developing OvaScan and eventually explore ways to make the technology more accessible for clinical use.

He also plans to pursue additional research involving artificial intelligence and health care, including the use of large language models.

After graduating from high school, Garg hopes to pursue an interdisciplinary career combining mathematics, computer engineering and biomedical research.

For now, he sees OvaScan as an early step toward a future in which technology could help address complex medical challenges — including the disease that affected his own family.

"It's encouraging to have this work recognized nationally, and I hope to build on what I've learned through future research," Garg said.