Related Experiment Videos
Internist-1, an experimental computer-based diagnostic consultant for general internal medicine
The New England Journal of Medicine
|August 19, 1982
Summary
The experimental INTERNIST-I program shows promise for computer-assisted diagnosis in internal medicine but requires further development. Its current limitations prevent reliable clinical application due to reasoning and diagnostic deficiencies.
Area of Science:
- Medical Informatics
- Artificial Intelligence in Medicine
- Clinical Decision Support Systems
Background:
- INTERNIST-I is an advanced experimental computer program designed for comprehensive internal medicine diagnoses.
- It distinguishes itself through a broad approach and extensive knowledge base compared to other diagnostic programs.
Purpose of the Study:
- To systematically evaluate the strengths and weaknesses of the INTERNIST-I program.
- To assess its diagnostic capabilities and identify areas for improvement.
Main Methods:
- A systematic evaluation was conducted using 19 clinicopathological exercises.
- Performance was compared against hospital clinicians and expert case discussants.
Main Results:
- INTERNIST-I's performance was qualitatively similar to hospital clinicians but inferior to expert discussants.
- The program demonstrated limitations in anatomical/temporal reasoning, multi-area differential diagnoses, and explaining its diagnostic process.
Conclusions:
- The current version of INTERNIST-I is not reliable enough for clinical use.
- Key deficiencies include reasoning limitations and the inability to provide transparent explanations for its diagnostic conclusions.