Rudolf von Bünau, co-founder and CEO of Altavo: »The voice is an important part of our personality. Its loss is difficult for many patients to cope with and severely restricts their quality of life.« The technical implementation of the idea works like this: the movements of the tongue, palate and lips are recorded using radar technology. The sensors for this are attached to the cheeks and voice output is provided via a small loudspeaker near the mouth. To feed the AI of the »Promise-AI« project, which is also funded by the Federal Ministry of Education and Research, the Altavo team is diligently collecting donations of votes. »An important part of our work is to collect masses of training data for AI,« says Peter Birkholz, Professor of Language Technology and Cognitive Systems at TU Dresden. The test subjects read various texts aloud for more than an hour while the radar sensors in their face record the movements of their tongue, lips and palate. The company, which was founded in 2021, is now planning a clinical trial and wants to test the technology on patients for the first time. They should then speak in a voice that sounds like their former voice. »Ideally, there should be a recording of this – even short snippets of sound are often enough. In future, targeted voice recordings could be made before every laryngectomy.« You can find more information about Altavo on the website.