Artificial intelligence to detect brain hemorrhages

The HCM supports joint project with the Clinic for Neuroradiology at the University Hospital Bonn

The project leaders - Alexander Radbruch, Director of the Clinic for Neuroradiology at the University Hospital Bonn (UKB), Daniel Paech, Senior Physician at the Clinic for Neuroradiology at the UKB and Alexander Effland from the Institute for Applied Mathematics. On the screen: CT images of a cerebral hemorrhage. © Universitätsklinikum Bonn (UKB)/J.F. Saba

 

Bonn, 05.08.2021. Cerebral hemorrhages are among the clinical emergencies in which rapid intervention is essential for the further course of the disease. In this context, radiology plays a central role, because only the reliable diagnosis of brain hemorrhage by means of CT (computed tomography) enables the correct classification of the hemorrhage and the initiation of further therapeutic steps. In order to be able to automatically detect brain hemorrhages in the future using artificial intelligence, mathematicians and physicians are working closely together.

The researchers intend to train self-learning artificial intelligence systems using large amounts of data so that they can learn to detect and qualify brain hemorrhages on their own. "The collaboration with medicine is so important because artificial intelligence systems can only get as good as the data they are trained on. Cooperation with the Clinic for Neuroradiology at the UKB, which has large amounts of image data on brain hemorrhages, is therefore essential," says Alexander Effland, one of the three project leaders. He is a group leader of HCM's Interdisciplinary Research Unit (IRU) Mathematics and Life Sciences and member of the Transdisciplinary Research Area (TRA) "Modelling" at the University of Bonn.

"The advantage of the new techniques is that they don't get tired and deliver the same performance even at three o'clock in the morning," emphasizes Daniel Paech, senior physician in the Clinic for Neuroradiology at UKB and also head of the project. While doctors run the risk of overlooking minor bleeding after long working days and especially during night duties, this does not happen to artificial intelligence, he adds.

"In Bonn, we have the great advantage of having an optimal partner for translational projects in the field of artificial intelligence in the form of Bonn's strong mathematics department," emphasizes Alexander Radbruch, Director of the Clinic for Neuroradiology at the UKB." Collaboration between mathematics and computer science as well as medicine is essential for the major future projects of medicine in the field of digitalization and artificial intelligence. Mathematical modeling, as well as theoretical and numerical analysis in conjunction with experimental data, is essential for future progress. Hence we decided some time ago to create internationally visible junior research groups in the field of mathematical modeling in life and medical sciences, the so-called Interdisciplinary Research Units (IRUs) in the area of “Mathematics and Life Sciences”.

To the press release of the University of Bonn