Prof. Dr. Şükriye Bilge Gürsel, a faculty member of the Radiation Oncology Department at Ondokuz Mayıs University's Faculty of Medicine, has made a significant contribution to the use of artificial intelligence in breast cancer treatment.
What sets Prof. Dr. Gürsel apart from her colleagues is that she completed this study not only as a doctor but also as a computer engineer, as part of her graduation thesis. While an associate professor at the Ondokuz Mayıs Faculty of Medicine, Gürsel re-entered university exams and won a place in computer engineering. Three years later, she graduated with high honors in engineering and was also promoted to a professorship in medicine. In her engineering thesis, she drew on her medical expertise and signed off on a pilot study capable of predicting recurrences in cervical and breast cancer post-treatment with 92% accuracy using artificial intelligence.
At the 15th National Radiation Oncology Congress organized by the Turkish Radiation Oncology Association (TROD), one of the most exciting topics was the session where Prof. Dr. Gürsel discussed her work on using artificial intelligence in breast cancer. What distinguishes her from her peers is that she accomplished this study not just as a doctor but also as a computer engineer through her own thesis.
Prof. Dr. Gürsel, who completed her engineering degree in three years with high honors thanks to the numerical courses she had previously taken, said, "I was a faculty member during the day and attended classes until late at night. In my 40s, I felt like I was reliving my 20s. There's no other doctor in Turkey who has graduated in computer engineering in the field of radiation oncology. I started in 2019 and finished in 2022. I graduated with high honors. We did two projects for the thesis. Both were in the health field.
"One project was about recurrence prediction in breast and cervical tumors using artificial intelligence applications. With artificial intelligence, collecting various patient data and parameters for breast or cervical cancer, it's possible to predict whether the disease will recur and how much the recurrence will decrease with certain treatments. This was demonstrated in my pilot study," she explained.
Prof. Dr. Gürsel shared details about her pilot study for her final project, which included 1,090 cancer patients from two centers with five-year follow-ups: "After teaching the AI the data of patients who had recurrences and those who didn't, the test group had a program that predicted with a 92% accuracy rate. Cancer has many parameters, such as molecular, genetic, and radiometric data. The human brain can't make these predictions for each parameter, but artificial intelligence will enable us. It will present data like 'this patient is at very high risk, the recurrence risk in 5 years is this high or low.' This will allow us to recommend more intense, more aggressive treatments for this patient, thus minimizing the risk of recurrence."
Positioning herself as an interface between medicine and engineering, Prof. Dr. Gürsel concluded: "The collection and processing of health data and developments in cancer are rapidly increasing; there's a lot of data. For example, just for one patient, data of 8 GB is generated. The human brain can't convert this much data into knowledge. Türkiye's data is also very important; we can't process 97% of our data. For example, that 8 GB of data for a patient isn't being used much. It stays in archives on computers. Data mining and artificial intelligence applications will be our hands and feet in converting this into knowledge and will be a method that must definitely be used in cancer treatments."
DİNLE