Automated selection of clinical data to support radiographic interpretation

Modern computerized hospital information systems can store much of the clinical information required for patient care. In some settings the amount of this information is great enough to interfere with its efficient use. Supplying information in support of radiographic interpretation provides an example of this problem. We describe an automated process that abstracts select data from a clinical database and creates a report recounting salient facts designed to assist the radiologist in interpreting the chest x-ray. The reporting system selects data using a combination of expert systems technology and information theory. It produces a brief summary of relevant clinical data. The original version of this process required run-time use of expert systems techniques that were time-prohibitive in the clinical setting. Through the use of pre-compiled tables of finding specific information contents we have reduced the time requirement to a point where the applications will soon be clinically feasible.