Respiratory Adaptive Computed Tomography (REACT) Helical Imaging for Reducing Artifacts in 4DCT Scans
Presentations
WE-C930-IePD-F2-1 (Wednesday, 7/13/2022) 9:30 AM - 10:00 AM [Eastern Time (GMT-4)]
Exhibit Hall | Forum 2
Purpose: Image artifacts in 4DCT scans lead to errors in the radiotherapy treatment workflow that propagate through treatment and cause negative clinical outcomes. REACT helical imaging employs a prospective couch velocity adaption and gating protocol to reduce artifacts that can occur due to breathing irregularity. In this digital phantom study, we demonstrate the benefits of our helical real-time gating protocol over conventional 4DCT imaging.
Methods: Breathing traces from 15 patients were selected from an existing tumor motion database based on trace variations with respect to its mean. 40 eXtended Cardiac And Thoracic (XCAT) phantom volumes each representing 2.5% of the breathing amplitude was generated for each breathing trace. The digital phantom volumes were used to simulate conventional and REACT helical acquisitions. Ten phase images were reconstructed with >180˚ fan angle filtered back projection with interpolation for each acquisition. Image quality was determined by computing Root Mean Squared Error (RMSE) between an acquired image and ground truth, Overlapping Tumor volume (OTV) of an acquired image to ground truth and identifying the number (>4mm, >6mm, >8mm) and type of artifacts (interpolation, double structure, truncation).
Results: Images acquired from REACT-helical acquisition decreased the overall number of artifacts (>4mm, >6mm, >8mm) around the tumor by 31% as compared to a conventional acquisition. The RMSE reduced by 41% and OTV increased by 4.2% for the REACT-helical acquisition as compared to conventional acquisition respectively.
Conclusion: Adaptive CT helical imaging has the potential to reduce errors in 4DCT images and potentially improve patient outcomes. The results demonstrate that REACT in helical mode outperforms conventional mode of 4DCT acquisition in reducing artifacts.
Funding Support, Disclosures, and Conflict of Interest: This work was supported by Cancer Australia funding (Application Number:1139268)
Keywords
Not Applicable / None Entered.
Taxonomy
Not Applicable / None Entered.
Contact Email