The overall goal of this research is to build a structured reporting system
that reduces the cost, delays, and inconvenience associated with conventio
nal dictation and speech recognition systems. We have implemented such a st
ructured reporting system for radiology that replaces current dictation and
transcription processes by allowing radiologists and other imaging profess
ionals to select imaging findings from a medical lexicon. The system uses a
n imaging-specific information model, called a "description set,' to organi
ze selected terms in a relational database. Unique features of the knowledg
e representation that enhance its expressiveness include its ability to cod
ify uncertainty about an imaging observation and to represent explicitly th
e logical relationships among imaging findings. In addition, the system doe
s not require the user to fill in "blanks' in a static text template. Inste
ad, it allows entry of terms in arbitrary order and uses automated text-gen
eration techniques to create a text report that referring physicians are ac
customed to receiving. In parallel, the system also produces a multimedia r
eport that the referring physician can use as a quick reference. Unlike the
results of conventional dictation or speech recognition, each finding is c
oded in a relational database for later information processing. Thus, the s
tructured report database can be used to index images by content, to provid
e real-time decision support, to enhance radiologists' performance, to cond
uct exploratory clinical research, and to transmit imaging report data to c
omputer-based patient record systems. Copyright (C) 2000 by W.B. Saunders C
ompany.