Patent №
US 4,835,690
Granted
—
Owner
—
Lab
—
AI components
2
ml · hardware
Assignment
None on record
Dataset
AIPD
2023_r1 edition
Application
07181116
A magnetic resonance or other diagnostic imaging scanner (A) is connected with an image reconstruction module (B) for reconstructing diagnostic images. An integrated expert system (C) selects scan parameter settings for conducting the scan such that utility of the image is optimized for the intended diagnosis. A keyboard (12) receives subject and intended diagnostic application data such as the age of the patient, the region of the patient to be imaged, the anticipated size of the lesion, and the like. A first expert system (10) derives appropriate constraints on values for each performance index and priorities for each performance index from the subject and intended application data. The performance indices include contrast, resolution, scan suration, and the like. A scan parameter estimator look-up table (14) is addressed by the performance index constraints and priorities and retrieves corresponding estimated scan parameters. A second expert system (16) adjusts the estimated scan parameter settings to optimize a performance function (26). That is, the performance index values are predicted by a preselected mathematical model which relates scan parameters and performance indices. The scan parameters settings are adjusted until the predicted performance indices have an optimal fit with preselected ideal performance index values with higher priority performance indices being weighted more heavily than lower priority indices and with the predicted performance values being within the constraints. This application is a continuation of application Ser. No. 799,427, filed Nov. 19, 1985, to Raymond E. Gangarosa, et al., entitled "Integrated Expert System for Medical Imaging Scan, Set-up, and Scheduling" now abandoned.