The Multimodal Medical Data Capture and Representation project enables the adoption of Electronic Medical Record (EMR) systems, especially in capturing and representing structured clinical information from unstructured data. To fully understand the workflow and requirements of EMR, a MITRE team evaluated medical Automatic Speech Recognition (ASR) engines, Handwriting Recognition (HR) tools, and as open source EMR systems; interviewed doctors; and visited hospitals. The team established a baseline performance using existing technology, built training materials for ASR/HR, and designed a multimodal research prototype that will interpret and convert captured free speech and handwriting into standard expressions.
To discuss licensing or collaboration activities, please contact us.