Out of field scatter from electron applicator in modern linear accelerators.

BACKGROUND: Electron out-of-field scatter is generally not given importance mainly in electron fields. However, this is important when applicator down and boost treatments are given usually at an angle from the central axis. The electron scatter dose is found to be far away from the central axis which could be easily ignored. PURPOSE: This study aims to investigate the out-of-field radiation doses from electron applicators and their effects on clinical treatment. By identifying the parameters that contribute to out-of-field doses and to explore potential strategies for reducing these doses in order to improve patient outcomes from modern machines. METHODS: Measurements were performed in water phantom using electron diode for modern Elekta and Varian machines. Dose profiles were acquired at surface and dmax with 0° and 90° collimation angle. Various gantry angles were also studied for some data with IC Profiler. The profiles were normalized with respect to the central axis dose. RESULTS: The scatter dose peaks were found at a distance between 11 and 28 cm from the central axis on all machines. However, the peak shifts to 15 cm at 90° collimator when beam is tilted. The position and intensity of the dose varies with depth, collimator, and gantry angles for both Elekta and Varian machines. Due to clearance issues more gantry angles were studied for Elekta applicator compared to Varian. In general, Varian TrueBeam has a lower scatter that Elekta Infinity. The 90° collimator angle has a higher scatter compared to zero degree for both machines. CONCLUSIONS: There are clinically significant peripheral doses around 3% of the central axis dose from the electron applicator. Elekta has a slightly higher scatter (3%) than Varian (2%) that peaks at 25 cm which is clinically important but often overlooked.

Utilizing a novel hybrid brachytherapy technique FINITO (Freehand Interstitial Needles in addition to Tandem and Ovoid) for locally advanced cervical cancer.

PURPOSE: We aimed to assess the clinical feasibility and advantages of using a novel hybrid brachytherapy technique by placing Freehand Interstitial Needles in addition to the Tandem and Ovoid applicator (FINITO) for the treatment of locally advanced cervical cancer (LAC). METHODS AND MATERIALS: A retrospective analysis was performed on two cohorts of patients with LACC treated at our institution: 29 patients in the FINITO group and 17 patients in the control group using T&O only approach. Clinical outcomes of interest included local control (LC), progression-free survival (PFS), overall survival (OS), and rates of acute and late toxicities. Kaplan-Meier methodology was used to estimate OS, PFS, and LC. Wilcoxon signed-rank test was used to compare the median values for dosimetry parameters. A p-value of ≤ 0.05 was considered statistically significant. All statistical analyses were performed using RStudio. RESULTS: At a median of 2 years there was no difference in rates of OS, PFS or LC between the FINITO and the control group of patients. The 2-year OS, PFS, and LC for the FINITO group were 59% (95% CI 34%-100%), 58% (95% CI 38%-89%), and 84% (95% CI 69%-100%), respectively. Late toxicities were significantly lower in the FINITO group for both gastrointestinal and genitourinary symptoms (p = 0.001 and 0.01, respectively) as compared to the T&O group. CONCLUSION: Based on the equivalent LC rate and lower toxicity profile, our FINITO technique appears to be an excellent alternative to the standard intracavitary brachytherapy in patients with advanced disease, especially in resource-limited settings.

An integrated and fast imaging quality assurance phantom for a 0.35 T magnetic resonance imaging linear accelerator.

Purpose: Periodic imaging quality assurance (QA) of magnetic resonance imaging linear accelerator (MRL) is critical. The feasibility of a new MRL imaging phantom used for QA in the low field was evaluated with automated image analysis of various parameters for accuracy and reproducibility. Methods and materials: The new MRL imaging phantom was scanned across every 30 degrees of the gantry, having the on/off state of the linac in a low-field MRL system using three magnetic resonance imaging sequences: true fast imaging with steady-state precession (TrueFISP), T1 weighted (T1W), and T2 weighted (T2W). The DICOM files were used to calculate the imaging parameters: geometric distortion, uniformity, resolution, signal-to-noise ratio (SNR), and laser alignment. The point spread function (PSF) and edge spread function (ESF) were also calculated for resolution analysis. Results: The phantom data showed a small standard deviation - and high consistency for each imaging parameter. The highest variability in data was observed with the true fast imaging sequence at the calibration angle, which was expected because of low resolution and short scan time (25 sec). The mean magnitude of the largest distortion measured within 200 mm diameter with TrueFISP was 0.31 ± 0.05 mm. The PSF, ESF, signal uniformity, and SNR measurements remained consistent. Laser alignment traditional offsets and angular deviation remained consistent. Conclusions: The new MRL imaging phantom is reliable, reproducible, time effective, and easy to use for a 0.35 T MRL system. The results promise a more streamlined, time-saving, and error-free QA process for low-field MRL adapted in our clinical setting.

Deep learning-based protoacoustic signal denoising for proton range verification.

Proton therapy is a type of radiation therapy that can provide better dose distribution compared to photon therapy by delivering most of the energy at the end of range, which is called the Bragg peak (BP). The protoacoustic technique was developed to determine the BP locationsin vivo, but it requires a large dose delivery to the tissue to obtain a high number of signal averaging (NSA) to achieve a sufficient signal-to-noise ratio (SNR), which is not suitable for clinical use. A novel deep learning-based technique has been proposed to denoise acoustic signals and reduce BP range uncertainty with much lower doses. Three accelerometers were placed on the distal surface of a cylindrical polyethylene (PE) phantom to collect protoacoustic signals. In total, 512 raw signals were collected at each device. Device-specific stack autoencoder (SAE) denoising models were trained to denoise the noise-containing input signals, which were generated by averaging only 1, 2, 4, 8, 16, or 24 raw signals (low NSA signals), while the clean signals were obtained by averaging 192 raw signals (high NSA). Both supervised and unsupervised training strategies were employed, and the evaluation of the models was based on mean squared error (MSE), SNR, and BP range uncertainty. Overall, the supervised SAEs outperformed the unsupervised SAEs in BP range verification. For the high accuracy detector, it achieved a BP range uncertainty of 0.20 ± 3.44 mm by averaging over 8 raw signals, while for the other two low accuracy detectors, they achieved the BP uncertainty of 1.44 ± 6.45 mm and -0.23 ± 4.88 mm by averaging 16 raw signals, respectively. This deep learning-based denoising method has shown promising results in enhancing the SNR of protoacoustic measurements and improving the accuracy in BP range verification. It greatly reduces the dose and time for potential clinical applications.

A Novel Workflow with a Customizable 3D Printed Vaginal Template and a Direction Modulated Brachytherapy (DMBT) Tandem Applicator for Adaptive Interstitial Brachytherapy of the Cervix.

A novel clinical workflow utilizing a direction modulated brachytherapy (DMBT) tandem applicator in combination with a patient-specific, 3D printed vaginal needle-track template for an advanced image-guided adaptive interstitial brachytherapy of the cervix. The proposed workflow has three main steps: (1) pre-treatment MRI, (2) an initial optimization of the needle positions based on the DMBT tandem positioning and patient anatomy, and a subsequent inverse optimization using the combined DMBT tandem and needles, and (3) rapid 3D printing. We retrospectively re-planned five patient cases for two scenarios; one plan with the DMBT tandem (T) and ovoids (O) with the original needle (ND) positions (DMBT + O + ND) and another with the DMBT T&O and spatially reoptimized needles (OptN) positions (DMBT + O + OptN). All retrospectively reoptimized plans have been compared to the original plan (OP) as well. The accuracy of 3D printing was verified through the image registration between the planning CT and the CT of the 3D-printed template. The average difference in D2cc for the bladder, rectum, and sigmoid between the OPs and DMBT + O + OptNs were -8.03 ± 4.04%, -18.67 ± 5.07%, and -26.53 ± 4.85%, respectively. In addition, these average differences between the DMBT + O + ND and DMBT + O + OptNs were -2.55 ± 1.87%, -10.70 ± 3.45%, and -22.03 ± 6.01%, respectively. The benefits could be significant for the patients in terms of target coverage and normal tissue sparing and increase the optimality over free-hand needle positioning.

Intensity Modulated High Dose Rate (HDR) Brachytherapy Using Patient Specific 3D Metal Printed Applicators: Proof of Concept.

Purpose: In high-dose-rate (HDR) brachytherapy, an anisotropic dose distribution may be desirable for achieving a higher therapeutic index, particularly when the anatomy imposes challenges. Several methods to deliver intensity-modulated brachytherapy (IMBT) have been proposed in the literature, however practical implementation is lacking due to issues of increased delivery times and complicated delivery mechanisms. This study presents the novel approach of designing a patient-specific inner shape of an applicator with 3D metal printing for IMBT using an inverse plan optimization model. Methods: The 3D printed patient-specific HDR applicator has an external shape that resembles the conventional brachytherapy applicator. However, at each dwell position of the HDR source, the shielding walls in the interior are divided into six equiangular sections with varying thicknesses. We developed a mathematical model to simultaneously optimize the shielding thicknesses and dwell times according to the patient's anatomical information to achieve the best possible target coverage. The model, which is a bi-convex optimization problem, is solved using alternating minimization. Finally, the applicator design parameters were input into 3D modeling software and saved in a 3D printable file. The applicator has been tested with both a digital phantom and a simulated clinical cervical cancer patient. Results: The proposed approach showed substantial improvements in the target coverage over the conventional method. For the phantom case, 99.18% of the target was covered by the prescribed dose using the proposed method, compared to only 58.32% coverage achieved by the conventional method. For the clinical case, the proposed method increased the coverage of the target from 56.21% to 99.92%. In each case, both methods satisfied the treatment constraints for neighboring OARs. Conclusion: The study simulates the concept of the IMBT with inverse planning using the 3D printed applicator design. The non-isotropic dose map can be produced with optimized shielding patterns and tailored to individual patient's anatomy, to plan a more conformal plan.

Investigation of Isotoxic Dose Escalation and Plan Quality with TDABC Analysis on a 0.35 T MR-Linac (MRL) System in Ablative 5-Fraction Stereotactic Magnetic Resonance-Guided Radiation Therapy (MRgRT) for Primary Pancreatic Cancer.

This study investigates plan quality generated by an MR-Linac (MRL) treatment planning system (TPS) for 5-fraction stereotactic body radiation therapy (SBRT) of primary pancreatic cancer (PCa). In addition, an isotoxic dose escalation was investigated with the MRL TPS based on stereotactic MR-guided adaptive radiation therapy (SMART) trial constraints. A clinical workflow was developed for adaptive and non-adaptive treatments with the MRL, on which a time-driven activity-based costing (TDABC) analysis was performed to quantify clinical efficacy. Fifteen PCa patients previously treated with a conventional Linac were retrospectively re-planned for this study. Three plans were generated for each patient using the original prescription dose (PD) and organ at risk (OAR) constraints (Plan 1), following SMART trial's OAR constraints but with the original PD (Plan 2), starting with Plan 2, following an isotoxic dose escalation strategy where the dose was escalated until any one of the SMART trial's OAR constraints reached its limit (Plan 3). Conformity index (CI) and the ratio of the 50% isodose volume to PTV (R50%) conformity metrics were calculated for all 45 MRL plans, in addition to standard dose-volume indices. Forty-five MRL plans were created which met their respective dosimetric criteria described above. For Plan 1, the MRL TPS successfully achieved equivalent or lower OAR doses while maintaining the prescribed PTV coverage for the 15 plans. A maximum dose to the small bowel was reduced on average by 4.97 Gy (range: 1.11-10.58 Gy). For Plan 2, the MRL TPS successfully met all SMART trial OAR constraints while maintaining equivalent PTV coverage. For Plan 3, the MRL TPS was able to escalate the prescription dose from the original 25-33 Gy by, on average, 36 Gy (range: 15-70 Gy), and dose to the PTV was successfully escalated to at least 50 Gy for all 15 plans. These achievements were made possible, in part, due to the omission of the ITV afforded by the MRL's real-time target tracking technology and sharper dose penumbra due to its unique dual-focus MLC design. The 0.35T MRL TPS can generate plans that are equivalent to conventional Linac-based plans for SBRT of PCa. Through analyzing Plan 2 and 3 strategies, and due to the real-time target localization capabilities of the MRL system, increased OAR sparing and/or target dose escalation are possible.

Review of Machine Learning in Lung Ultrasound in COVID-19 Pandemic.

Ultrasound imaging of the lung has played an important role in managing patients with COVID-19-associated pneumonia and acute respiratory distress syndrome (ARDS). During the COVID-19 pandemic, lung ultrasound (LUS) or point-of-care ultrasound (POCUS) has been a popular diagnostic tool due to its unique imaging capability and logistical advantages over chest X-ray and CT. Pneumonia/ARDS is associated with the sonographic appearances of pleural line irregularities and B-line artefacts, which are caused by interstitial thickening and inflammation, and increase in number with severity. Artificial intelligence (AI), particularly machine learning, is increasingly used as a critical tool that assists clinicians in LUS image reading and COVID-19 decision making. We conducted a systematic review from academic databases (PubMed and Google Scholar) and preprints on arXiv or TechRxiv of the state-of-the-art machine learning technologies for LUS images in COVID-19 diagnosis. Openly accessible LUS datasets are listed. Various machine learning architectures have been employed to evaluate LUS and showed high performance. This paper will summarize the current development of AI for COVID-19 management and the outlook for emerging trends of combining AI-based LUS with robotics, telehealth, and other techniques.

Flat-panel imager energy-dependent proton radiography for a proton pencil-beam scanning system.

In proton-based radiotherapy, proton radiography could allow for direct measurement of the water-equivalent path length (WEPL) in tissue, which can then be used to determine relative stopping power (RSP). Additionally, proton radiographs allow for imaging in the beam's-eye-view. In this work, a proton radiography technique using a flat-panel imager and a pencil-beam scanning (PBS) system is demonstrated in phantom. Proton PBS plans were delivered on a Varian ProBeam system to a flat-panel imager. Each proton plan consisted of energy layers separated by 4.8 MeV, and a field size of 25 cm × 25 cm. All measured data is binned into a layer-by-layer delivery in post processing. To build a calibration curve correlating detector response to WEPL, the plans were delivered to slabs of solid water of increasing thickness. Pixel-by-pixel detector response in the time/energy domain is then fit as a function of WEPL. Tissue equivalent phantoms are imaged for evaluation of WEPL accuracy. A spatial resolution phantom and a head phantom are also imaged. For all experiments, the detector was run with an effective pixel size of 0.4 mm × 0.4 mm. The proposed method reconstructed RSP with mean errors of 2.65%, -0.14%, and 0.61% for lung, soft tissue, and bone, respectively. In a 40 mm thick spatial resolution phantom, a 2 mm deep pinhole with 1 mm diameter can be seen. The accuracy and spatial resolution of the method show that it could be implemented for patient position verification. Further development could lead to patient-specific verification of RSP to be used for treatment guidance.

Analytical Low-Dose CBCT Reconstruction Using Non-local Total Variation Regularization for Image Guided Radiation Therapy.

Purpose: Conventional iterative low-dose CBCT reconstruction techniques are slow and tend to over-smooth edges through uniform weighting of the image penalty gradient. In this study, we present a non-iterative analytical low-dose CBCT reconstruction technique by restoring the noisy low-dose CBCT projection with the non-local total variation (NLTV) method. Methods: We modeled the low-dose CBCT reconstruction as recovering high quality, high-dose CBCT x-ray projections (100 kVp, 1.6 mAs) from low-dose, noisy CBCT x-ray projections (100 kVp, 0.1 mAs). The restoration of CBCT projections was performed using the NLTV regularization method. In NLTV, the x-ray image is optimized by minimizing an energy function that penalizes gray-level difference between pair of pixels between noisy x-ray projection and denoising x-ray projection. After the noisy projection is restored by NLTV regularization, the standard FDK method was applied to generate the final reconstruction output. Results: Significant noise reduction was achieved comparing to original, noisy inputs while maintaining the image quality comparable to the high-dose CBCT projections. The experimental validations show the proposed NLTV algorithm can robustly restore the noise level of x-ray projection images while significantly improving the overall image quality. The improvement in normalized mean square error (NMSE) and peak signal-to-noise ratio (PSNR) measured from the non-local total variation-gradient projection (NLTV-GPSR) algorithm is noticeable compared to that of uncorrected low-dose CBCT images. Moreover, the difference of CNRs from the gains from the proposed algorithm is noticeable and comparable to high-dose CBCT. Conclusion: The proposed method successfully restores noise degraded, low-dose CBCT projections to high-dose projection quality. Such an outcome is a considerable improvement to the reconstruction result compared to the FDK-based method. In addition, a significant reduction in reconstruction time makes the proposed algorithm more attractive. This demonstrates the potential use of the proposed algorithm for clinical practice in radiotherapy.