Related Experiment Videos
Expert system-controlled image display.
H A Swett1, P R Fisher, A I Cohn
1Department of Diagnostic Radiology, Yale University School of Medicine, New Haven, CT 06510.
Radiology
|August 1, 1989
Summary
IMAGE/ICON is a novel expert computer system for diagnostic radiology. It visually displays sorted images to aid physicians in pattern recognition and medical education, enhancing diagnostic accuracy.
Area of Science:
- Medical Imaging
- Artificial Intelligence in Medicine
- Radiology
Background:
- Traditional text-based medical expert systems have limitations in visually-dependent fields like radiology.
- Radiological diagnosis relies heavily on pattern recognition and subtle morphological feature analysis.
Purpose of the Study:
- To develop a prototype expert computer system, IMAGE/ICON, for diagnostic radiology.
- To enhance diagnostic aid and medical education through a visual, interactive platform.
Main Methods:
- Developed IMAGE/ICON, a system that displays image groups sorted by similarities (features, causes, clinical setting).
- The system presents morphologic variations and disease spectrums.
- It also generates written analyses of key case features.
Main Results:
- IMAGE/ICON can display images based on shared features, causes, or clinical contexts.
- The system illustrates morphologic variations and disease-related abnormality spectra.
- It provides accompanying written analyses of case specifics.
Conclusions:
- IMAGE/ICON serves as a valuable diagnostic aid and educational tool in radiology.
- The system has potential as an intelligent radiologic workstation, integrating with picture archiving and communication systems.