Reducing 4D CT imaging artifacts at the source: first experimental results from the respiratory adaptive computed...

Area of Science:

• Medical Imaging
• Radiotherapy Physics
• Computational Imaging

Background:

• Breathing variations during 4D CT imaging cause geometric irregularities (respiratory-induced image artifacts).
• These artifacts negatively impact radiotherapy treatment efficacy.
• Reducing image artifacts is crucial for accurate radiation targeting.

Purpose of the Study:

• To develop and evaluate Respiratory Adaptive Computed Tomography (REACT) for reducing respiratory-induced image artifacts in 4D CT.
• To integrate REACT with clinical hardware for the first time.
• To hypothesize that REACT reduces artifacts by ≥ 4 mm compared to conventional 4D CT.

Main Methods:

• 4D CT image sets acquired using REACT and conventional methods on a Siemens Somatom scanner.
• REACT triggered CT acquisition based on real-time breathing regularity evaluation using the Varian RPM system.
• Images compared to a static-phantom ground truth for geometric differences; breathing irregularity assessed via RMSE_Beam_on.

Main Results:

• REACT significantly reduced artifacts ≥ 4 mm by 70% for tumors (p=0.003) and 76% for lungs (p=0.0002).
• Tumor volume reduction observed (10% to 6%), and lung volume reduction (2% to 1%) compared to conventional 4D CT.
• Breathing irregularity during imaging (RMSE_Beam_on) reduced by 27% (p=0.013) with REACT.

Conclusions:

• REACT was successfully integrated with clinical hardware, demonstrating its clinical feasibility.
• The study supports the hypothesis that REACT significantly reduces respiratory-induced image artifacts compared to conventional 4D CT.
• REACT shows promise for providing clearer images in clinical radiotherapy, warranting further investigation.