This research project focuses on the effectiveness, usability, and impact of training in the use of tools that integrate AI-driven speech-to-speech translation, aimed at facilitating communication with non-native speakers in healthcare settings. Based on insights from recent studies on AI translation tools in healthcare, including findings from Flanders, Belgium, this project assesses the role of these tools in promoting linguistic inclusion, ease of use, and accessibility in multilingual healthcare environments.

The research centers on three essential dimensions to ensure the quality standards of the tools and to strengthen collaboration in healthcare. First, we examine the usability and acceptance among healthcare providers: how do the tools influence user experience and satisfaction for providers, and to what extent do they improve communication with non-native speaking patients? This will be explored through user testing, interviews, and participant observation.

Second, we evaluate the impact of training: what is the effect of targeted training on the use and effectiveness of the tools by healthcare providers? Using pre- and post-training surveys, we will measure the results of the training, mapping potential improvements in the use of the tools and communication in healthcare.

Finally, we analyze the quality of AI-driven translations: how accurate and functional are the translations, especially for languages less well represented in AI training data, such as Arabic and Farsi? This analysis will identify areas for refinement, aiming to improve linguistic inclusion for underrepresented language groups.

By combining qualitative and quantitative methods, this research seeks to advance the integration of AI translation tools, balancing operational efficiency with privacy and data security, thereby contributing to an inclusive and accessible healthcare environment.