Effect of simultaneous integrated boost concepts on photoneutron and distant out-of-field doses in VMAT for prostate cancer.

BACKGROUND: A simultaneous integrated boost (SIB) may result in increased out-of-field (DOOF) and photoneutron (HPN) doses in volumetric modulated arc therapy (VMAT) for prostate cancer (PCA). This work therefore aimed to compare DOOF and HPN in flattened (FLAT) and flattening filter-free (FFF) 6­MV and 10-MV VMAT treatment plans with and without SIB. METHODS: Eight groups of 30 VMAT plans for PCA with 6 MV or 10 MV, with or without FF and with uniform (2 Gy) or SIB target dose (2.5/3.0 Gy) prescriptions (CONV, SIB), were generated. All 240 plans were delivered on a slab-phantom and compared with respect to measured DOOF and HPN in 61.8 cm distance from the isocenter. The 6­ and 10-MV flattened VMAT plans with conventional fractionation (6- and 10-MV FLAT CONV) served as standard reference groups. Doses were analyzed as a function of delivered monitor units (MU) and weighted equivalent square field size Aeq. Pearson's correlation coefficients between the presented quantities were determined. RESULTS: The SIB plans resulted in decreased HPN over an entire prostate RT treatment course (10-MV SIB vs. CONV -38.2%). Omission of the flattening filter yielded less HPN (10-MV CONV -17.2%; 10-MV SIB -22.5%). The SIB decreased DOOF likewise by 39% for all given scenarios, while the FFF mode reduced DOOF on average by 60%. A strong Pearson correlation was found between MU and HPN (r > 0.9) as well as DOOF (0.7 < r < 0.9). CONCLUSION: For a complete treatment, SIB reduces both photoneutron and OOF doses to almost the same extent as FFF deliveries. It is recommended to apply moderately hypofractionated 6­MV SIB FFF-VMAT when considering photoneutron or OOF doses.

A novel end-to-end test for combined dosimetric and geometric treatment verification using a 3D-printed phantom.

The dosimetric and geometric accuracy are important components to ensure safe patient treatment in radiation therapy. Therefore, these components must be checked during quality control. This work presents a possible solution for the determination of the geometric isocenter deviation in the entire treatment chain. Additionally, the dose measurement of the established end-to-end test workflow measured in the same procedure as the geometric deviation is described. An in-house designed end-to-end test phantom went through the entire procedure of a standard patient treatment and the dosimetric and geometric accuracy were determined. At 3 linear accelerators (linac), the phantom was positioned either with cone beam computed tomography or with surface guidance. In this position, a Winston-Lutz test was performed and the deviations of the gantry, collimator and couch isocenter measurements to the phantom position were determined. Additionally, a dose measurement in the phantom was performed and compared to the dose predicted in the treatment planning system. To validate the results obtained with the in-house designed phantom, comparative measurements with commercial phantoms were performed. According to the performed end-to-end test, 2 out of the 3 linacs showed isocenter variations larger than 1 mm for collimator and gantry rotations and larger than 2 mm for couch rotations. With an isocenter variation of less than 1 mm for collimator and gantry rotations, 1 linac fulfilled the tolerance for stereotactic treatments without couch rotation. With couch rotation, an isocenter variation of less than 2 mm was detected at this linac, which fulfilled the tolerance for IMRT treatments. The mean dose deviation between measurement and treatment planning system was 1.82% ± 1.03%. The results acquired with the UMM phantom did not show statistically significant deviations to those acquired with relevant other commercial phantoms. The novel end-to-end test procedure allows for a combined dosimetric and geometric treatment evaluation. Besides the commonly performed dose end-to-end test the geometric isocenter deviation within a patient treatment workflow was evaluated and categorized for IMRT or SBRT.

Ultrasound-based repositioning and real-time monitoring for abdominal SBRT in DIBH.

AIM: Ultrasound-based repositioning and real-time-monitoring aim at the improvement of the precision of SBRT in deep inspiration breath-hold (DIBH). Accuracy of ultrasound-based daily repositioning was estimated by comparison with DIBH-cone-beam-CT. Intrafraction motion during beam-delivery was assessed by ultrasound-real-time-monitoring. PATIENTS/METHODS: Residual error after ultrasound-based interfractional repositioning (85 fractions, 16 SBRT-series; 14 patients) was assessed by marker-based (7 series) or liver-contour-based (9 series) matching in DIBH-CBCT. During beam-delivery, the percentage of 3D misalignment vector below 2 mm, between 2 and 5 mm, 5-7 mm and over 7 mm was estimated. Percentage of relevant target-displacements was analyzed as a function of DIBH-duration. RESULTS: Residual error after ultrasound-based positioning was 0.4 ±â€¯3.3 mm in LR (left-right), 0.2 ±â€¯4.3 mm in CC (cranio-caudal) and 1.0 ±â€¯3.0 mm in AP (anterior-posterior) directions (vector magnitude 5.4 ±â€¯3.3 mm, MV ±â€¯SD). Over 544 DIBHs, target displacement was 1.3 ±â€¯0.5 mm, 0.7 ±â€¯0.3 mm, 1.6 ±â€¯0.6 mm for CC, LR and AP directions, respectively (3D-vector 2.5 ±â€¯0.7 mm). 3D misalignment vector length was below 2 mm in 49.8%, between 2 and 7 mm in 46.3%, and over 7 mm in 3.9% of the beam-delivery-time. During the first 5 s of the DIBH, 3D-misalignment vector length was always below 10 mm. Percentage of target displacements over 10 mm was 0.2%, 0.5% and 0.8% for 10 s, 15 s and 20 s DIBH-duration. CONCLUSIONS: Ultrasound-based interfractional repositioning is an accurate method for daily localization of abdominal DIBH-SBRT targets. Residual motion is <7 mm in 96% of the beam-delivery-time. Deviations >10 mm occur rarely and can be avoided by gating the beam at a predefined threshold. Ideal DIBH-duration should not exceed 15 s.

Hybrid 3D ranging and velocity tracking system combining multi-view cameras and simple LiDAR.

Scanning our surroundings has become one of the key challenges in automation. Effective and efficient position, distance and velocity sensing is key to accurate decision making in automated applications from robotics to driverless cars. Light detection and ranging (LiDAR) has become a key tool in these 3D sensing applications, where the time-of-flight (TOF) of photons is used to recover distance information. These systems typically rely on scanning of a laser spot to recover position information. Here we demonstrate a hybrid LiDAR approach which combines a multi-view camera system for position and distance information, and a simple (scanless) LiDAR system for velocity tracking and depth accuracy. We show that we are able to combine data from the two component systems to provide a compound image of a scene with position, depth and velocity data at more than 1 frame per second with depth accuracy of 2.5 cm or better. This hybrid approach avoids the bulk and expense of scanning systems while adding velocity information. We hope that this approach will offer a simpler, more robust alternative to 3D scanning systems for autonomous vehicles.

Image reconstruction from photon sparse data.

We report an algorithm for reconstructing images when the average number of photons recorded per pixel is of order unity, i.e. photon-sparse data. The image optimisation algorithm minimises a cost function incorporating both a Poissonian log-likelihood term based on the deviation of the reconstructed image from the measured data and a regularization-term based upon the sum of the moduli of the second spatial derivatives of the reconstructed image pixel intensities. The balance between these two terms is set by a bootstrapping technique where the target value of the log-likelihood term is deduced from a smoothed version of the original data. When compared to the original data, the processed images exhibit lower residuals with respect to the true object. We use photon-sparse data from two different experimental systems, one system based on a single-photon, avalanche photo-diode array and the other system on a time-gated, intensified camera. However, this same processing technique could most likely be applied to any low photon-number image irrespective of how the data is collected.