This project presents a speech interface for voice entry of fluid balance data in the Patient Data Management System (PDMS) using automatic speech recognition and speech generation. This improved interface offers a reliable ``hands-free'' and ``eyes-free'' means for bedside data entry. The design supports multimodal user interaction originating from the keyboard, voice input, or the mouse depending on the user's choice. The system processes a vocabulary of fluid ingesta and excreta, their body sites, sets of numbers, units of measure, identification of the patient, as well as correction of data incorrectly input or not recognized. Validation of the software and user assessment are also involved.
R. El-Khoury, A.S. Malowany