Evaluation and Prediction of X-ray Tube Failure in Medical Imaging Systems

X-ray tube failures are unfavorable and costly since they can stop an operation in its tracks. An essential part of an X-ray tube is the filament, which is heated to extremely elevated temperatures to emit electrons. The filament loses material through surface evaporation as a result of the elevated temperatures attained. The evaporation will cause the filament to get thinner and eventually wear out. When a filament fails, the entire X-ray tube needs to be replaced. In this research, we primarily addressed the condition-based failure prediction of a filament. We begin by describing the components and typical failure modes of the X-ray tube. The variable chosen to serve as a monitor for the filament condition was the C-factor. We suggest a model that uses the C-factor to track the state of a filament and anticipate when it will fail. Therefore, we attempt to construct the prediction model by fitting linear models piecewise with techniques like linear regression. Additionally, we found that small filaments commonly fail more than large filaments. The $\mathbf{C}$-factor gradually decreases as the filament gets old. However, we observed that the $\mathbf{C}$-factor gradually decreases at first, then abruptly jumps back to increase due to the automatic adaptation, and then gradually decreases again.