Enhancing Speech-to-Text Technology for Healthcare: Tackling Medical Jargon and Noisy Environments

**Speech-to-text (STT) technology is becoming increasingly important in the medical field**. Researcher Bożena Kostek from Gdańsk University of Technology is investigating how to optimize STT systems for healthcare applications. The primary focus is on improving transcription accuracy for medical terms, particularly in Polish, which poses challenges as STT models have been predominantly trained on English. **Kostek's work involves creating a detailed audio dataset comprising Polish medical terms spoken by specialists**, which is then analyzed using Automatic Speech Recognition models to convert speech into text. Key metrics such as Word Error Rate and Character Error Rate are used to gauge the model's performance. The study highlights difficulties in adapting STT models to handle medical jargon and operate effectively in noisy hospital environments. Kostek aims to address these issues by enhancing clarity in speech and is set to present her findings at the Acoustical Society of America meeting. The research is expected to expand to other languages, with ongoing collaborations in the Czech Republic. **The ultimate goal is to facilitate better data collection and increase face-time between doctors and patients.**