Search

• Technological innovation in Radiotherapy led to Magnetic Resonance guided RT systems. • Artificial Intelligence can play a significant role in MR guided RT. • Many manual procedures can be automatized and standardized thanks to AI. • Medical Physicists are expected to be major actors in AI clinical implementation. • The current status of AI integration in MRgRT will be presented and discussed. Over the last years, technological innovation in Radiotherapy (RT) led to the introduction of Magnetic Resonance-guided RT (MRgRT) systems. Due to the higher soft tissue contrast compared to on-board CT-based systems, MRgRT is expected to significantly improve the treatment in many situations. MRgRT systems may extend the management of inter- and intra-fraction anatomical changes, offering the possibility of online adaptation of the dose distribution according to daily patient anatomy and to directly monitor tumor motion during treatment delivery by means of a continuous cine MR acquisition. Online adaptive treatments require a multidisciplinary and well-trained team, able to perform a series of operations in a safe, precise and fast manner while the patient is waiting on the treatment couch. Artificial Intelligence (AI) is expected to rapidly contribute to MRgRT, primarily by safely and efficiently automatising the various manual operations characterizing online adaptive treatments. Furthermore, AI is finding relevant applications in MRgRT in the fields of image segmentation, synthetic CT reconstruction, automatic (on-line) planning and the development of predictive models based on daily MRI. This review provides a comprehensive overview of the current AI integration in MRgRT from a medical physicist's perspective. Medical physicists are expected to be major actors in solving new tasks and in taking new responsibilities: their traditional role of guardians of the new technology implementation will change with increasing emphasis on the managing of AI tools, processes and advanced systems for imaging and data analysis, gradually replacing many repetitive manual tasks. [ABSTRACT FROM AUTHOR]

Purpose: Dark current radiation produced during linac beam‐hold has the potential to lead to unplanned dose delivered to the patient. With the increased usage of motion management and step‐and‐shoot IMRT deliveries for MR‐guided systems leading to increased beam‐hold time, it is necessary to consider the impact of dark current radiation on patient treatments. Methods: The relative dose rate due to dark current for the ViewRay MRIdian linac was measured longitudinally over 15 months (June 2018‐August 2019). Ion chamber measurements were acquired with the linac in the beam‐hold state and the beam‐on state, with the ratio representing the relative dark current dose rate. The potential contribution of the dark current dose to the overall prescription was retrospectively analyzed for 972 fractions from 83 patients over the same time period. The amount of time spent in the beam‐hold state was combined with the monthly measured relative dark current dose rate to estimate the dark current dose contribution. Results: The relative dark current dose rate compared to the beam‐on dose rate was 0.12% ± 0.027%. In a near worst‐case estimation, the dark current dose contribution accounted for 0.90% ± 0.67% of the prescription dose across all fractions (3.61% maximum). Gantry and MLC motion between segments accounted for 87% of the dark current contribution, with the remaining 13% attributable to gating during segment delivery. The largest dark current contributions were associated with plans delivering a small dose per treatment segment. Conclusions: The dark current associated with new clinical treatment units should be considered prior to treatment delivery to ensure it will not lead to dosimetric inaccuracies. For the MRIdian linac system investigated in this work, the contribution from dark current remained relatively low, though users should be cognizant of the larger potential dosimetric contribution for plans with small doses per segment. [ABSTRACT FROM AUTHOR]

Purpose: With advance magnetic resonance (MR)‐guided online adaptive radiotherapy (MRgoART) relying on calculation‐based intensity‐modulated radiation therapy (IMRT) quality assurance (QA), accurate and sensitive QA of the multileaf collimator (MLC) becomes an increasingly essential component for routine machine QA. As such, it is important to assure compliance with the AAPM TG142 guidelines to supplement calculation‐based QA methods for an online adaptive radiotherapy program. We have developed and implemented an efficient and highly sensitive QA procedure using an ionization chamber profiler (ICP) array to enable real‐time characterization of the positional accuracy of a double‐focused and double‐stacked MLC on a clinical MR‐guided radiotherapy (MRgRT) system and to supplement calculation‐based QA for an MRgoART program. Methods: An in‐house MR‐compatible jig was used to position the ICP (detector resolution 5 mm on X/Y axis) at an extended SDD of 108.4 cm to enable each MLC leaf (8.3 mm leaf width at isocenter) to be uniquely determined by two neighboring ion chambers. The MRgRT linac system utilizes a novel jawless, double‐focused, and double‐stacked MLC design such that the upper bank (MLC1) and lower bank (MLC2) are offset by half a leaf width. Positional accuracy was characterized by three methods: single bank half‐beam block (HBB) at central axis, forward slash diagonal (FSD), and backslash diagonal (BSD) at off‐axis. Measurements were performed for each bank in which each leaf occludes half of a detector. A corresponding reference field with the MLC retracted from occlusion was measured. The sensitivities of HBB, FSD, and BSD were evaluated by introducing 0.5–2.5 mm of known errors in 0.5 mm increments, in both positive and negative directions. The relationship between detector response and MLC error was established. Over a 6‐month longitudinal assessment, we have evaluated MLC performance with weekly QA of HBB among cardinal angles, and monthly QA of FSD and BSD. Results: A strong correlation was found between detector response of percentage dose difference and MLC positional error introduced (N = 350 introduced errors) for both HBB and FSD/BSD with coefficient of determination of 0.999 and 0.977, respectively. The relationship between detector response to MLC positional change was found to be 20.65%/mm for HBB and 11.14%/mm for FSD and BSD. At baseline, the mean MLC positional accuracy averaged across all leaves was 0.06 ± 0.27 mm (HBB), 0.04 ± 0.52 mm (FSD), −0.06 ± 0.51 mm (BSD). The mean MLC positional accuracy relative to baseline over the 6‐month assessment was found to be highly reproducible at 0.00 ± 0.12 mm (HBB; N = 28 weeks), −0.02 ± 0.19 mm (FSD; N = 6 months), −0.03 ± 0.19 mm (BSD; N = 6 months). Conclusions: Positional accuracy of a novel jawless, double‐focused, double‐stacked MLC has been characterized and monitored over 6 months with an efficient, highly sensitive, and robust method using an ICP array. This routine QA method supplements calculation‐based IMRT QA for an online adaptive radiotherapy program. Longitudinal assessment demonstrated no‐drift in the MLC calibration. A highly reproducible jig setup allowed the validation of MLC positional accuracy to be within TG142 criteria of ±1 mm for 99% of measurements (i.e., 100% HBB, 95% BSD, 95% FSD) over the 6‐month assessment. [ABSTRACT FROM AUTHOR]

Purpose: To describe and characterize daily machine quality assurance (QA) for an MR‐guided radiotherapy (MRgRT) linac system, in addition to reporting a longitudinal assessment of the dosimetric and mechanical stability over a 7‐month period of clinical operation. Methods: Quality assurance procedures were developed to evaluate MR imaging/radiation isocenter, imaging and patient handling system, and linear accelerator stability. A longitudinal assessment was characterized for safety interlocks, laser and imaging isocenter coincidence, imaging and radiation (RT) isocentricity, radiation dose rate and output, couch motion, and MLC positioning. A cylindrical water phantom and an MR‐compatible A1SL detector were utilized. MR and RT isocentricity and MLC positional accuracy was quantified through dose measured with a 0.40 cm2 x 0.83 cm2 field at each cardinal angle. The relationship between detector response to MR/RT isocentricity and MLC positioning was established through introducing known errors in phantom position. Results: Correlation was found between detector response and introduced positional error (N = 27) with coefficients of determination of 0.9996 (IEC‐X), 0.9967 (IEC‐Y), 0.9968 (IEC‐Z) in each respective shift direction. The relationship between dose (DoseMR/RT+MLC) and the vector magnitude of MLC and MR/RT positional error (Errormag) was calculated to be a nonlinear response and resembled a quadratic function: DoseMR/RT+MLC[%] = −0.0253 Errormag [mm]2 − 0.0195 Errormag [mm]. For the temporal assessment (N = 7 months), safety interlocks were functional. Laser coincidence to MR was within ±2.0 mm (99.6%) and ±1.0 mm (86.8%) over the 7‐month assessment. IGRT position–reposition shifts were within ±2.0 mm (99.4%) and ±1.0 mm (92.4%). Output was within ±3% (99.4%). Mean MLC and MR/RT isocenter accuracy was 1.6 mm, averaged across cardinal angles for the 7‐month period. Conclusions: The linac and IGRT accuracy of an MR‐guided radiotherapy system has been validated and monitored over seven months for daily QA. Longitudinal assessment demonstrated a drift in dose rate, but temporal assessment of output, MLC position, and isocentricity has been stable. [ABSTRACT FROM AUTHOR]

The aim of this study is to investigate the role of Delta Radiomics analysis in the prediction of one-year local control (1yLC) in patients affected by locally advanced pancreatic cancer (LAPC) and treated using Magnetic Resonance guided Radiotherapy (MRgRT). A total of 35 patients from two institutions were enrolled: A 0.35 Tesla T2*/T1 MR image was acquired for each case during simulation and on each treatment fraction. Physical dose was converted in biologically effective dose (BED) to compensate for different radiotherapy schemes. Delta Radiomics analysis was performed considering the gross tumour volume (GTV) delineated on MR images acquired at BED of 20, 40, and 60 Gy. The performance of the delta features in predicting 1yLC was investigated in terms of Wilcoxon Mann–Whitney test and area under receiver operating characteristic (ROC) curve (AUC). The most significant feature in predicting 1yLC was the variation of cluster shade calculated at BED = 40 Gy, with a p-value of 0.005 and an AUC of 0.78 (0.61–0.94). Delta Radiomics analysis on low-field MR images might play a promising role in 1yLC prediction for LAPC patients: further studies including an external validation dataset and a larger cohort of patients are recommended to confirm the validity of this preliminary experience. [ABSTRACT FROM AUTHOR]