With AI against bleeding gums and periodontitis

"90 percent of the population brushes their teeth every day, yet over 70 percent suffer from inflammation of the gums or tooth bed. That means everyone knows how important brushing teeth is, but almost no one does it properly," says Prof. Dr. Bernhard Sick, head of the Department of Intelligent Embedded Systems at the University of Kassel.

There is a lot of interest in better understanding tooth brushing: Dentistry sees the consequences of poor oral hygiene, and health insurers incur costs. In geriatric care, too, it is relevant to know how best to brush other people's teeth, and the movement sciences are interested in how to train and automate certain movement sequences.

Brushing teeth is not trivial

For this reason, scientists working on the "Brushalyze" project are developing a new research device that performs a detailed analysis of the toothbrushing process in an automated manner, thus enabling a more detailed study. In this way, Prof. Dr. Renate Deinzer from JLU Giessen, Prof. Dr. Keywan Sohrabi from THM and Prof. Dr. Bernhard Sick and their colleagues are providing important impetus for basic research in dentistry. With the new device, dentistry will be able to analyze tooth brushing in greater depth in the future. Dentistry is supporting the project during development as part of the Scientific Advisory Board.

To date, the most accurate analysis of the toothbrushing process has been provided by video analysis by personnel trained in observation methods. "Training this staff and then conducting the observations is extremely time-consuming," explains Prof. Dr. Renate Deinzer, head of the Institute of Medical Psychology at JLU Giessen. "Analyzing a toothbrushing process of a few minutes requires several hours. At the same time, important details cannot be analyzed in a purely visual observation."

An intelligent toothbrush for research is emerging

Dr. Keywan Sohrabi, professor of medical informatics at the Technical University of Central Hesse, designs the hardware. Sensors will be integrated into a manual toothbrush to measure direction of movement, force and other physical variables. "With this prototype, we can conduct initial studies and collect data," Sohrabi explains. Sick's colleagues are using these sensor data to develop a model using machine learning techniques to map the toothbrushing process: Which tooth surface is currently being brushed? Is the brush being moved in a circular or horizontal motion? What pressure is being applied and was the plaque successfully removed? This replaces video analysis. The Medical Psychology team, which specializes in observing toothbrushing processes, in turn provides basic data and checks the plausibility of the automated evaluations. This is how a new research device for basic research in dentistry is created.

The German Research Foundation is funding "Brushalyze" for three years with more than 750,000 euros. The University of Kassel will receive 293,250 euros, the JLU Giessen 232,900 euros and the THM 251,050 euros. The project is accompanied by a high-ranking Scientific Advisory Board with 16 members from dentistry, psychology and computer science.

Contact:

University of Kassel
Prof. Dr. rer. Nat. Bernhard Sick
Head of Department of Intelligent Embedded Systems
Tel.: 0561 804 6020
E-mail: bsick[at]uni-kassel[dot]de

Justus Liebig University Giessen
Prof. Dr. Renate Deinzer
Head of the Institute of Medical Psychology
E-Mail: Renate.Deinzer@mp.med.uni-giessen.de

Technische Hochschule Mittelhessen
Prof. Dr. rer. physiol. Keywan Sohrabi
Department of Health, Medical Informatics
Tel.: 0641 309 6600
E-mail: keywan.sohrabi[at]ges.thm[dot]de