The innovation of Artificial Intelligence (AI) once again managed to come up with another technological breakthrough in the medical sector. Recently students from the Faculty of Engineering at Universitas Dian Nuswantoro (Udinus) developed an automatic segmentation model of brain blood vessels, including veins and arteries. The model was developed using the MONAI framework combined with Re-Net.

The research could improve the detection accuracy of arteries, as well as reduce the dependency on manual segmentation by radiologists. The research was published in the Journal of Electronics, Electromedical Engineering, and Medical Informatics, on December 5, 2024. The research team is currently preparing the article to be forwarded to the Q3 Journal.

The research was entitled ‘The Automatic Dicom Data Segmentation of Brain Veins and Arteries Utilizing the MONAI Framework.’ It involves two faculty members, namely, Menik Dwi Kurniatie, S.Si., M.Biotech —a lecturer from the Engineering Faculty at Udinus, and a lecturer from the Medicine Faculty at Udinus, dr. Andreas Wilson Setiawan, M.Kes.

Meanwhile, the research also involved students of the Engineering Faculty at Udinus, including Reica Diva Jacinda, and Nebrisca Patriana Yossy. Not only that, Udinus also involved a researcher from the National Innovation Research Agency (BRIN), Talitha Asmaria.

Reica, the leader of the research team, explained that the model was able to diagnose cerebrovascular diseases, like stroke, aneurysm, and arteriovenous malformation. According to her, using the MONAI technology would allow the segmentation process to be done automatically in under a minute.

“This can certainly help medical professionals in analyzing the patients’ condition more quickly and accurately. This is because the model also involves the utilization of Artificial Intelligence,” Reica uttered.

At the research, Reica highlighted the primary challenges during the research, including the limited access to medical data and computation resources. 

“We had to look for optimization strategies, including dataset augmentation and cloud computing utilization to handle the GPU limitation,” she explained.

Improving Work Efficiency

Meanwhile, one of the faculty members supervising the project, Menik, stated that the innovation had indeed the potential to enhance the quality of learning media, particularly in supporting and planning better pre-surgery procedures. With their success, the MONAI-based Automatic segmentation model of brain vessels is expected to be widely adopted in the medical sector.

She continued that with the benefits, the research could serve as a huge step in AI implementation, particularly in the Medical Sector. This is because the innovation could improve the analytical accuracy of medical images and medical professional work efficiency in the near future.

“We want our research to be beneficial, not only in diagnosing but also in the development of more advanced medical technology in general,” Menik concluded. (Humas Udinus/Alex. Foto: Dok. Humas FT)