System requirements to improve adaptive 4-dimensional computed tomography (4D CT) imaging.

Area of Science:

• Medical Imaging
• Radiotherapy Physics
• Computational Imaging

Background:

• Four-Dimensional Computed Tomography (4D CT) is crucial for stereotactic body radiotherapy (SBRT) but suffers from respiratory motion artifacts.
• Irregular breathing during 4D CT acquisition can lead to artifacts, negatively impacting treatment efficacy.
• REspiratory Adaptive CT (REACT) is a gating method designed to minimize motion artifacts by avoiding imaging during irregular breathing phases.

Purpose of the Study:

• To investigate the contribution of hardware and software latencies in 4D CT and REACT to image artifacts beyond those caused by irregular breathing.
• To quantify the impact of various latencies, including gantry rotation time, couch shifts, acquisition delays, and phase calculation methods, on image quality.

Main Methods:

• Simulated imaging using the digital extended cardiac-torso (XCAT) phantom with fifty patient-measured respiratory traces.
• Compared conventional cine 4D CT with five REACT protocols featuring systematically varied parameters to assess latency effects.
• Quantified artifacts by measuring image normalized cross-correlation and volume error against a static ground truth.

Main Results:

• Gantry rotation time and acquisition delay were identified as the most impactful parameters for reducing image artifacts.
• The phase calculation method also influenced motion artifacts, suggesting potential for patient-specific optimization.
• Increased artifacts correlated with baseline drift, indicating that longer scan times may degrade image quality.

Conclusions:

• Optimizing gantry rotation time and acquisition delay is critical for future REACT and 4D CT development to minimize motion artifacts.
• Patient-specific assessment of the phase calculation method may further improve artifact reduction.
• Minimizing scan time and addressing baseline drift are important considerations for maintaining image quality in motion-managed CT acquisitions.