Analyzing image data to detect a CT system’s error state and identify the corresponding root cause

Especially in a fully automated inline setup of an X-ray system, human supervision of the measurements is reduced.This makes it crucial to avoid undetected error states of the X-ray system and to ensure that the measurement results can be trusted.This paper presents an approach to automate the step of visually inspecting the image data using algorithms, aiming to detect anomalies in the image data which indicate error states of the X-ray system.Furthermore, an analysis of the image data based on machine learning (trained by realistic simulations), which helps to identify the corresponding root cause of an error state, is being investigated.For both, first promising results are being presented.