An operating theatre can be noisy and stressful, especially in critical situations. Error-free communication between everyone involved is a top priority. Speech is, however, not always clearly intelligible. That is why Fraunhofer IDMT in Oldenburg is working together with partners in the “METIOR” project on the development of a hearable system for error-free human-to-human communication as well as robust speech interaction and speech assistance. A further focus is the correct perception of important sounds, for example those emitted by medical devices.
Further elements of the project are the assessment and evaluation of (noise-induced) stress and the system’s use in practice.
The hearable is a smart earbud that facilitates clear communication in noisy workplaces. In addition, other solutions can be integrated, such as speech recognition and voice control. To test the hearable system under realistic conditions, the researchers are working in an operating theatre laboratory at Pius-Hospital in Oldenburg, one of the project partners. There, they can try out various test scenarios without affecting a real operation or jeopardising the procedure.
To do this, they are using various Fraunhofer IDMT technologies, such as algorithms for improving speech intelligibility. With the help of source separation methods based on neural networks, it is possible to separate the noise in the operating theatre into two components – speech and background noise – and speech can be transmitted to the addressees free of interference. One of the challenges here is that alarm sounds as well as other important and useful signals must of course remain audible.
A further aim is to integrate Fraunhofer technologies for individual hearing assistance into the system – because everyone hears differently. In this way, each and every person working in the operating theatre can adjust their hearable’s signal so that it optimally matches their own hearing preferences – not only the volume balance but also the sound.