CCIR alumna Petra Hedesiu’s research began in the Cambridge Future Scholar Programme and was later published in Nature’s Scientific Reports.
How can a high school student contribute to scientific research? For CCIR alumna Petra Hedesiu, the journey began with an interest in AI and healthcare. She pursued a research question during CCIR’s Summer 2024 Cambridge Future Scholar Programme. We are proud to celebrate the publication of her work in Scientific Reports, a peer-reviewed journal within the Nature Portfolio.
Nature Portfolio publishes the internationally renowned journal Nature. Its journal Scientific Reports evaluates original research through expert peer review. Petra’s publication marks a significant milestone in her academic journey. She began the study while still a high school student researcher at CCIR.
Her paper, Automated Detection of Carotid Artery Encasement on CT Scans Using an AI-Driven Approach, investigates an automated method for assessing the relationship between a tumour and the carotid artery. The study appeared online on October 10, 2026.
Before publication, Petra presented her findings at the European Congress of Radiology in 2025. She was the youngest presenter at the event. She now studies at Columbia University as an Egleston Scholar.
Turning an interest in AI into research at CCIR
Petra joined CCIR’s Summer 2024 Cambridge Future Scholar Programme with an interest in machine learning and healthcare. Before joining, she had completed a data science internship and explored AI applications for MRI images. These experiences gave her a foundation to build on during the programme.
Through the Artificial Intelligence in Medical Image Analysis course, Petra worked under the mentorship of Dr. Kutsev Ozyoruk. A teaching assistant also supported her work. Her independent research explored whether AI could help detect carotid artery encasement on CT scans of patients with head and neck cancer.
In her CCIR Student Spotlight interview, Petra highlighted the consistent mentorship and technical support she received. She had not experienced the same level of support in previous independent school projects. Throughout the investigation, her mentor and teaching assistant helped her work through technical challenges.
Petra also valued the programme’s progression from lectures to practical research. First, she developed her understanding of AI concepts. Then, she applied that knowledge to a specific medical imaging problem.
Interested in exploring AI through your own research?
Turn your interests in artificial intelligence, healthcare, and medical imaging into an independent research project. At CCIR Academy, you can explore research questions with guidance from experienced professor mentors.
Discover CCIR Academy’s research programmes, professor mentors, and student opportunities through our research enquiry form.
Investigating carotid artery encasement with AI
Carotid artery encasement occurs when a tumour surrounds the carotid artery. This major blood vessel supplies blood to the brain. In head and neck cancer, encasement can complicate tumour removal and increase surgical risks. Doctors therefore rely on specialist interpretation of CT scans to assess the relationship.
Petra investigated whether AI could assist with this assessment. In her Student Spotlight interview, she described a key challenge: she did not have a pre-annotated dataset. Such a dataset would contain images with the relevant structures already labelled for model training.
To address this challenge, Petra developed and trained two separate models. One identified the carotid artery, while the other identified the tumour. This process is known as segmentation. It outlines specific structures within a medical image.
The resulting outlines allowed her to examine the distance between the tumour and artery. She could also assess how far the tumour wrapped around the vessel.
These measurements helped her study whether AI could support radiologists in evaluating a factor that affects surgical planning. Petra envisioned the method as a potential aid for less experienced radiologists. However, further development and evaluation are needed before researchers can establish its usefulness in clinical practice.
The youngest presenter at the European Congress of Radiology
In 2025, Petra presented her findings at the European Congress of Radiology in Vienna, Austria. The European Society of Radiology organises this major international meeting. That year, more than 20,500 participants from 131 countries attended.
The audience included radiologists, radiographers, researchers, and other medical imaging experts. As the youngest presenter, Petra shared research she had started in high school. She brought her work on AI-assisted CT analysis to an international audience of professionals.
Petra also served as the study’s first author. Her presentation marked another step in the development of research that began during the CCIR programme.
Publication in Nature’s Scientific Reports
The journal Scientific Reports accepted Petra’s study on September 17, 2026. It published the paper on October 10, 2026.
Petra is the first-listed author. An author note recognises her equal contribution alongside Sorana Eftimie. Dr. Kutsev Ozyoruk also co-authored the paper and continued supporting the research after the CCIR programme.
Publishing in a peer-reviewed Nature Portfolio journal marks an important achievement for Petra. Her study also adds findings to the wider scientific discussion of AI-assisted medical image analysis. Other researchers can examine the results and build on the work.
Beginning undergraduate studies at Columbia as an Egleston Scholar
Petra now studies at Columbia University, where she was selected for the Egleston Scholars Program. Columbia describes Egleston Scholars as representing the top 1% of Columbia Engineering applicants. The programme recognises their achievements and potential as students, researchers, and leaders.
Her Columbia Engineering profile highlights her experience in machine learning and medical image analysis. It also features the research she pursued through CCIR.
Reflecting on her experience, Petra said:
“The research experience it gave me helped me earn a likely letter to Columbia University through the Egleston Scholars Program.”
The Egleston Scholars Program offers research opportunities, faculty mentorship, and support for academic and professional development. These resources will allow Petra to build on the skills she developed at CCIR. She also hopes to explore AI applications beyond healthcare, including cybersecurity, finance, and climate science.
How curiosity and commitment can lead to meaningful research
Can a high school student make a meaningful contribution to science? Petra’s experience shows how a research question can develop through sustained effort and academic guidance.
During CCIR’s Cambridge Future Scholar Programme, she explored the intersection of AI and healthcare under the guidance of Dr. Kutsev Ozyoruk and her teaching assistant. However, her work did not end when the programme finished. She continued developing her research, presented at the European Congress of Radiology, and published her study in Scientific Reports.
Now an undergraduate and Egleston Scholar at Columbia University, Petra can continue building on this foundation. Her next steps may lead to new questions and applications across several fields.
At CCIR, we are proud to celebrate Petra’s achievements. Her journey illustrates how research mentorship can help young scholars turn curiosity into substantive academic work. We look forward to seeing how her research develops in the years ahead.
Ready to take your research interests further?
Turn your academic interests into an original research project with guidance from experienced professor mentors at CCIR.
Download the CCIR Research Programme Prospectus to explore our 170+ research areas, discover our Cambridge Future Scholar and 1-on-1 Research Mentorship programmes, learn about our professor mentors from leading universities, and explore programme formats, student outcomes, scholarships, and the application process.
