CSE Wins Best Full Paper Award at Eurographic Workshop

CSE Wins Best Full Paper Award at Eurographic Workshop

06 Oct 2022

CSE Wins Best Full Paper Award at Eurographic Workshop

The College of Science and Engineering (CSE) demonstrated its leadership in innovative visual computing solutions by winning the Best Full Paper Award during the Eurographics Workshop on Visual Computing for Biology and Medicine 2022, in Vienna, Austria.

Titled "HistoContours: a framework for visual annotation of histopathology whole slide images", the paper was co-authored by three CSE students: Khaled Al-Thelaya and Nauman Gilal - both PhD in Computer Science and Engineering candidates, and Faaiz Joad, Master of Science in Data Science student.

They presented the paper on behalf of the team leading the project: Dr. Marco Agus and Dr. Jens Schneider, both Assistant Professors at CSE, and Sidra Medicine Scientist Dr. William Mifsud.

The CSE research team is working with Sidra Medicine and CRS4 in Italy with the aim of optimizing histopathologists' workflows using artificial intelligence (AI).

Histopathologists usually have little time to dedicate to each tissue sample they need to analyze, but generally work to have first an overview of the condition of the tissue before attaining more accurate statistics.

The project uses AI and computer vision techniques to localize and classify individual nuclei (otherwise known as cell cores). The information extracted in this fashion is then aggregated across space to suppress noise and give the histopathologist an overview, while at the same time providing accurate statistics on the distribution of cell conditions.

What sets this project apart from the research happening globally is that Sidra Medicine is a pediatric hospital. The main attention of the international research community is on cancer, whereas children more often suffer from allergies and inflammations. This also means that the data sets commonly available are biased towards cancer detection. The researchers use these data sets and a process called transfer learning to adapt their AI models to the focus area Sidra Medicine is interested in.

The team will also plan to extend their work towards decision support systems for rare diseases, such as Wilms’ tumor. For more information about the College of Science and Engineering, visit cse.hbku.edu.qa.