Radiographic testing system with learning-based performance prediction

Abstract

Disclosed herein is a radiographic imaging system which performs system performance monitoring by (1) using automatic exposure control (AEC) components to predict the average image gray level; (2) obtaining measured average image gray levels from the portions of the X-ray detector situated in the X-ray shadow of the AEC components; and then (3) comparing the predicted and measured values. The predicted values are determined by use of a prediction model which is modified by a learning system over successive exposures to provide more accurate predictions. After the learning system has sufficiently developed the prediction model, the error between the predicted and measured gray level values may be monitored in later exposures and an error routine can be activated if the error exceeds a predetermined threshold. In this case, the error may indicate that system components in the imaging chain (e.g., the detector or AEC components) require maintenance.

Dependencies are inferred from the claim text.