A multimodal interface for the immediate transcription of radiology dictation

We present the design and implementation of an integrated multimodal interface that delivers instant turnaround on transcribing a radiology dictation. This instant turnaround time virtually eliminates a hospital´s liability with respect to improper transcriptions of oral dictations and all but eliminates the need for transcribers. The multimodal interface seamlessly integrates three modes of input - speech, handwriting, and written gestures to provide an easy-to-use system for the radiologist.