ESTRO 2023 - Abstract Book

S1373

Digital Posters

ESTRO 2023

Conclusion The results reveal that deep learning-based approaches could overcome the undersampling challenge and bring the temperature difference closer to the ground truth. Even yet, the deep learning results indicate a 1°C temperature difference. This can be due to the models' performance differences between the magnitude and phase images. Future studies will concentrate on enhancing the networks' performance on phase images, which should result in a smaller temperature differential.

PO-1670 Initial Assessment of Adaptive Pancreas SBRT Treatment Efficiency using ViewRay MRIdian A3i

A. Gutierrez 1 , K. Mittauer 1 , R. Herrera 1 , J. McCulloch 1 , D. Alvarez 1 , N. Bassiri-Gharb 1 , R. Tolakanahalli 1 , A. Kaiser 1 , M. Hall 1 , R. Kotecha 1 , M. Mehta 1 , M. Chuong 1

1 Miami Cancer Institute - BHSF, Radiation Oncology, Miami, USA

Purpose or Objective Magnetic Resonance Image-guided Radiation Therapy (MRgRT) offers the ability to perform on-table adaptive radiotherapy using an automated beam gated delivery approach. A new upgrade was recently released for the ViewRay MRIdian system, consisting of substantial software and hardware changes, focusing on enhancing treatment delivery efficiency. Key developments include a novel software user interface that permits a parallel workflow amongst the physician, physicist, and therapist during the adaptive workflow process, multi-planar imaging and structure tracking, and an integrated patient visual feedback monitor to facilitate breath holds (BH). As one of the first worldwide A3i adopters beginning in July 2022, our primary objective in this analysis was to quantify time efficiency improvements observed in pancreas cancer patients treated in mid-inspiration BH with stereotactic MR-guided adaptive radiation therapy (SMART). Materials and Methods Since the launch of our MRgRT program in April 2018, treatment time information for all patients has been recorded for quality assessment and improvement purposes of our workflow efficiency. Specific metrics such as imaging time, contouring and adaptive plan reoptimization time (TAT), treatment delivery time (TDT)—inclusive of beam pauses due to respiratory motion, and total in-room time (TIRT) are recorded for each fraction of all patients. For this study, a subset of pancreas cancer patients receiving ablative SMART to 40-50 Gy/5 fx was selected. The last 19 patients (95 adapted fractions) treated on the previous ViewRay platform (v2) prior to the upgrade and the first 13 patients (65 adapted fractions) treated on A3i were evaluated. Specific time metrics investigated were TAT, TDT, and TIRT. A Mann-Whitney test (p<0.05) was used to evaluate statistical significance. Results We observed a 26.9% relative reduction in the median TIRT favoring A3i (49 vs. 67 mins; p<0.01) and a 50.0% reduction in median TAT time (20 vs. 40 mins; p<0.01) compared to v2. Regarding TAT, a reduction of 38.8% was also observed in the interfraction variability when using A3i over v2 as quantified by the reduction in variance (7 vs 12 mins; p<0.01). We found no significant difference in the median TDT (16 vs 16 mins) as expected since our standardized planning technique did not change after upgrading to A3i. TAT of ≤ 20 mins was rare with v2 (1%) although has become more common since we upgraded to A3i (48%). Likewise, TIRT of ≤ 50 mins have become significantly more common with A3i vs. v2 (52% vs. 4%).