Petra Hedesiu

Image

Petra Hedesiu from Cluj-Napoca, Romania, graduated from “Emil Racovita” National College and plans to major in Computer Science with a minor in Economics at Columbia.

Petra was particularly involved in the intersection of Machine Learning (ML) and medical image analysis, especially in oncology and radiology. While conducting independent research through the Cambridge Future Scholars Program, she was struck by the challenge of accurately identifying carotid artery encasement—a critical cancer complication linked to low survival rates. To address this, she developed a novel ML model for automated detection of carotid artery encasement on Computed Tomography (CT) scans. She first-authored the resulting study and, as the youngest presenter, shared her findings at the European Congress of Radiology in February 2025 in Vienna, Austria.

Her first exposure to Machine Learning came during the summer of her sophomore year in a Data Science internship at National Instruments, a leading test and measurement company. Tasked with building a battery rejection prediction model, she took an innovative approach by analyzing the full I–V test curve, achieving 97% accuracy.

Curious to explore how AI could advance healthcare, Petra joined Pixeldata, a Romanian medical software company, as a Data Science Intern. There, she optimized deep learning algorithms such as U-Net, ResNet, and generative models to enhance CT image quality, reducing the required radiation dose by 20% and enabling more frequent scans for earlier disease detection.

Her first research experience came at the “Iuliu Hatieganu” University of Medicine and Pharmacy in Cluj-Napoca, where she applied AI to classify temporomandibular disc displacements on Magnetic Resonance Imaging (MRI). She was listed as a co-author of the study, published in August 2024 in the peer-reviewed Q1 journal Medicina.

In her free time, Petra enjoys skiing, tennis, and collecting records. A former ski racer, she competed in the 2023 Senior National Championships in Romania.

At Columbia, Petra hopes to continue exploring how AI can be applied across diverse fields. Building on her work in healthcare, she is curious about its potential in cybersecurity, finance, and climate science, working to support the community by finding ways to improve patient outcomes, protect data, or address global challenges.