Investigation of a Novel 3D Dosimetry System Based on ClearView 3D Radiochromic Dosimeters

Purpose: To investigate and characterize a novel 3D dosimetry system consisting of ClearView radiochromic gel dosimeters and a state-of-the-art telecentric optical CT scanner: The Duke Large Field of View Optical-CT Scanner (DLOS). Methods and Materials: ClearView radiochromic dosimeters (Modus QA) are gellan gum based radiochromic dosimeters containing a water-soluble tetrazolium salt which reduces into an insoluble formazan dye (with associated color change) under ionizing radiation. Initial spectrophotometric studies investigated linearity of dose response on small volumes of ClearView in optical cuvettes. Simple, single beam benchmark radiation therapy treatments (central axis photon, lateral photon, and electron deliveries) were delivered to 10 and 15-cm diameter ClearView dosimeters. Additionally, a “stacking pyramid” delivery was developed consisting of overlaid fields of increasing size, delivered down the central axis of each dosimeter, with a 1x1 cm2 small field at the center. The treatments were modeled with a commissioned Eclipse treatment planning system. Dosimeters were scanned with the DLOS, submerged in a refractive index (RI) matching fluid (ca. 10% propylene glycol) both pre- and post-irradiation (within 24h) and 3D reconstructions of the change in linear-optical-attenuation was determined using in-house software and 3D Slicer. Percent depth-dose (PDD), cross plane and in-plane profiles, and relative 3D gamma analysis were performed and compared to the commissioned Eclipse dose, which served as ground truth. All experiments followed a standardized workflow for consistency. Results: Linearity of dose response was confirmed in the cuvette study with excellent agreement (R2 ≥ 0.9986) at two wavelengths (520-and 632 nm) at 3 post-irradiation time points: 21 hours, 6 and 10 days. Dosimeter reconstructions were performed at 1mm³ resolution in full 3D. Relative dose profiles of all irradiations, in both 15-and 10 cm dosimeters, show good agreement in comparison to Eclipse dose calculations, with root mean square errors (RMSE) 0.00107-0.006649, and R2 ≥ 0.9808. Relative 3D gamma analysis was performed at 7%4mm, 5%3mm, 3%3mm, 3%2mm, and 2%2mm for all deliveries on both 10-and 15-cm dosimeters. 15-cm benchmark irradiations passed with ≥ 94% at 2%2mm, ≥ 90% at 3%3mm, and ≥ 90% at 2%2mm, for the central axis, left lateral, and electron deliveries, respectively. 10-cm benchmark irradiations passed with ≥ 93% at 3%2mm, ≥ 91% at 3%3mm, and ≥ 90 at 3%2mm, for the central axis, left lateral, and electron deliveries, respectively. 15-cm stacking field irradiations passed with ≥ 94% at 3%2mm, and 10-cm stacking field irradiations passed with ≥ 96% at 2%2mm. Regions of known artifacts were excluded from gamma analysis (jar base, neck, wall). Some artifacts remain unaccounted for (e.g., ring and cupping artifacts). Conclusion: This work presents the first use of a telecentric optical-CT scanner with ClearView. The system shows substantive promise for a new, comprehensive 3D dosimetry system, and this effort lays the groundwork for further, more specialized applications. Both the benchmark irradiations and stacked field deliveries performed exceptionally well in the various gamma analyses and investigation of the profiles, even in the presence of artifacts that were not completely accounted for in the data analysis. General differences between the 15-cm and 10-cm dosimeters are not made abundantly clear with the small sample size of this work, but both seem viable. There seems to be no difference between photon and electron deliveries, and both are viable with the system given the experiments performed.