A Self-Contained Calibration Scheme for Micro-CT With Irregular Trajectories Based on Phantom Auto-Measurement

Cone-beam CT (CBCT) with irregular trajectories has been widely used in various fields such as clinical medicine and scientific research. Precise geometric calibration is a crucial prerequisite for determining the imaging quality of these systems. Current primary calibration methods rely on highly accurate geometric phantoms with complex patterns. However, to reduce the reliance on manufacturing these phantoms with specialized equipment, a common alternative method is to use a well-calibrated CBCT system with a matching resolution to measure the phantom. However, when calibrating high-resolution CBCT (Micro CT) with irregular trajectories using this strategy, finding a Micro CT with a matching resolution and field of view is often inconvenient, and may even be challenging when the desired Micro CT resolution exceeds the available ones. To address this issue, we propose a method for measuring phantoms based on uncalibrated systems with irregular trajectories. This method is based on a non-convex model constructed with multiple projection views and multiple coordinate systems. After solving the non-convex model efficiently by using the proposed simulated annealing algorithm with a hook, the phantom can be measured accurately. The advantage of this method is that it does not rely on a third-party well-calibrated Micro CT system and enables precise measurements of phantoms using the desired uncalibrated system itself. Consequently, by combining existing geometric calibration algorithms, we can achieve self-contained calibration independent of third-party measurement or manufacturing instruments. This significantly simplifies the dependence on objective conditions during the geometric calibration process for Micro CT systems with irregular trajectories.