Real Time Kilovolt Feedback and Image Artifact Suppression of High Voltage Power Supply for Industrial Cone Beam CT
Industrial cone beam computed tomography has revolutionized nondestructive testing and inspection across manufacturing, aerospace, and materials science applications. The technique reconstructs three dimensional internal structure from a series of two dimensional X-ray projections acquired as the object rotates relative to the X-ray source and detector. The high voltage power supply driving the X-ray tube determines the X-ray energy spectrum and intensity, with voltage stability being critical for achieving artifact free reconstructions. Real time feedback control of the kilovolt value combined with appropriate image processing enables suppression of artifacts arising from voltage fluctuations, enhancing the diagnostic capability of cone beam CT systems.
The X-ray tube in a cone beam CT system operates by accelerating electrons from a cathode to an anode at high potential, typically ranging from tens to hundreds of kilovolts depending on the object density and thickness being inspected. The electron kinetic energy upon impact with the anode determines the maximum X-ray energy through the bremsstrahlung process, with the peak X-ray energy in kiloelectron volts equaling the tube voltage in kilovolts. The X-ray spectrum also includes characteristic X-rays from the anode material, but the continuous bremsstrahlung spectrum dominates the imaging relevant radiation. The tube voltage therefore directly determines the penetrating power of the X-ray beam and the contrast characteristics of the resulting images.
Voltage fluctuations during CT data acquisition cause corresponding variations in the X-ray spectrum and intensity. When projections acquired at different voltages are combined in the reconstruction algorithm, the inconsistent attenuation measurements create artifacts in the reconstructed volume. Ring artifacts appear as circular variations in density around the rotation axis, arising from systematic differences in detector response that may be linked to voltage variations at specific projection angles. Streak artifacts radiate from high density features when the attenuation measurements in different projections are inconsistent due to voltage changes. These artifacts can obscure real features or create false indications, degrading the inspection reliability.
Real time kilovolt feedback systems continuously monitor the tube voltage and adjust the high voltage power supply to maintain the desired setpoint. The voltage measurement typically employs resistive divider networks that scale the high voltage to levels measurable by precision analog to digital converters. The divider ratio must be precisely calibrated and stable over temperature and time to ensure accurate voltage measurement. High voltage dividers designed for measurement applications use precision resistor networks with low temperature coefficients and guard structures to minimize leakage currents that could affect measurement accuracy.
The feedback controller processes the voltage measurement and generates correction signals to the power supply. Proportional integral derivative control algorithms provide the basic regulation function, with the controller gains tuned for the specific dynamics of the high voltage system. The proportional gain determines the immediate response to voltage errors, while the integral gain eliminates steady state offset. The derivative gain can improve transient response but may amplify measurement noise if not carefully limited. Digital implementation of the controller enables sophisticated algorithms including adaptive gain scheduling, feedforward compensation, and notch filtering to address specific disturbance frequencies.
The bandwidth of the feedback system determines the frequency range over which voltage disturbances are suppressed. Higher bandwidth enables correction of faster disturbances but requires faster measurement and control execution. The limiting factors for bandwidth include the measurement system latency, the power supply response time, and the sampling rate of the digital controller. High performance systems may achieve bandwidths of hundreds or thousands of hertz, enabling suppression of disturbances from mains frequency ripple, switching power supply artifacts, and other sources within this frequency range.
Image based artifact suppression techniques complement the voltage feedback control by correcting for residual voltage variations in the projection data. Normalization of projection intensities based on reference regions or air measurements can partially compensate for intensity variations from voltage changes. More sophisticated approaches model the relationship between voltage and X-ray spectrum, enabling correction of the spectral changes that affect contrast and beam hardening behavior. These post processing techniques can improve image quality even with imperfect voltage stability, but work best when combined with effective real time control.
The X-ray intensity varies with the square of the tube voltage for a constant tube current, making intensity a sensitive indicator of voltage stability. Monitoring the X-ray intensity with the detector or dedicated reference detectors provides an additional measure of voltage stability that can be used for artifact correction or feedback control. The intensity measurement responds to the effective voltage averaged over the exposure time, while direct voltage measurement captures the instantaneous value. Combining both measurements provides comprehensive characterization of voltage behavior during CT acquisition.
Environmental factors affecting voltage stability include temperature variations, mains voltage fluctuations, and load changes from X-ray tube conditioning or aging. Temperature changes affect the resistance values in voltage dividers and the characteristics of power supply components. Mains voltage variations propagate through the power supply to the output unless adequately rejected by the regulation system. X-ray tube characteristics change with aging and conditioning, potentially affecting the load presented to the power supply. The feedback system must accommodate these variations while maintaining stable output voltage.
The integration of voltage feedback with the CT system control enables coordinated optimization of imaging performance. The voltage setpoint can be adjusted between scans or even during scans for specialized techniques such as dual energy imaging. The feedback parameters can be optimized for the specific imaging requirements, with tighter regulation for high contrast applications and relaxed requirements for screening applications where artifacts are less critical. Communication between the voltage control system and the reconstruction software enables artifact correction algorithms to use the measured voltage history for improved correction accuracy.
