Accuracy and Spatial Resolution Comparison of Two Proton Computed Tomography Scanners
G Dedes1*, H Drosten1, S Goetz1, J Dickmann1, C Sarosiek2, M Pankuch3, N Krah4, S Rit4, V Bashkirov5, R Schulte5, R Johnson6, K Parodi1, E DeJongh7, G Landry8, (1) LMU Munich, Munich, Bavaria, DE, (2) Medical College of Wisconsin, Milwaukee, WI, (3) Northwestern Medicine Chicago Proton Center, Warrenville, IL, (4) Universite de Lyon, Lyon, ,FR, (5) Loma Linda University, Loma Linda, CA, (6) U.C. Santa Cruz, Santa Cruz, CA, (7) ProtonVDA Inc, Naperville, IL, (8) University Hospital, LMU Munich, Munich, BV, DE
Presentations
SU-J-207-1 (Sunday, 7/10/2022) 4:00 PM - 5:00 PM [Eastern Time (GMT-4)]
Room 207
Purpose: Proton computed tomography (pCT) can provide superior relative stopping power (RSP) accuracy, which may allow reducing range margins in proton therapy. Several scanner prototypes have been built or are in the development phase. Different design approaches impact spatial resolution and RSP accuracy. This study conducted the first direct comparison of two conceptually different pCT scanners, using the same image reconstruction algorithm and the same scanned object.
Methods: A full tracking (FT) pCT prototype scanner and a commercially oriented positions only tracking (PO) scanner hosted at the same proton therapy facility were compared. The scanners also differ in their energy detector concepts. A phantom containing cylindrical plastic inserts of known RSP was scanned, and images were reconstructed with the same filtered backprojection algorithm accounting for curved proton paths. The RSP accuracy was expressed as the mean absolute percent error (MAPE) with respect to the reference RSP over all inserts. The spatial resolution was estimated from the radial edge spread function of high-density inserts.
Results: Both scanners achieved the same RSP MAPE (0.72%) when excluding the porous sinus insert. The overall MAPE of the PO scanner was better (0.81% vs. 1.14%) when all inserts and the body of the phantom were accounted for. The FT scanner achieved a higher spatial resolution (0.61 lp/mm) than the PO scanner (0.46 lp/mm), attributed to the lack of exit direction tracking of the latter.
Conclusion: The comparison between two prototype pCT scanners at the same clinical facility and with the same image reconstruction algorithm showed that they both fulfill the requirement of RSP accuracy around 1%. Their different design approaches were reflected in the spatial resolution, resulting in higher spatial resolution for the FT scanner than the for the more compact PO scanner, which only tracks the positions of protons but not their directions.
Funding Support, Disclosures, and Conflict of Interest: M. Pankuch and R. W. Schulte are members of the advisory board of ProtonVDA LLC. E.DeJongh is employed by ProtonVDA LLC.
Keywords
Stopping Power, Spatial Resolution, Protons
Taxonomy
IM- Particle (e.g., Proton) CT: Quantitative imaging/analysis
Contact Email