Reduction of cone-beam CT artifacts in a robotic CBCT device using saddle trajectories with integrated infrared tracking.

BACKGROUND: Cone beam computed tomography (CBCT) is widely used in many medical fields. However, conventional CBCT circular scans suffer from cone beam (CB) artifacts that limit the quality and reliability of the reconstructed images due to incomplete data. PURPOSE: Saddle trajectories in theory might be able to improve the CBCT image quality by providing a larger region with complete data. Therefore, we investigated the feasibility and performance of saddle trajectory CBCT scans and compared them to circular trajectory scans. METHODS: We performed circular and saddle trajectory scans using a novel robotic CBCT scanner (Mobile ImagingRing (IRm); medPhoton, Salzburg, Austria). For the saddle trajectory, the gantry executed yaw motion up to ± 10 ∘ $\pm 10^{\circ }$ using motorized wheels driving on the floor. An infrared (IR) tracking device with reflective markers was used for online geometric calibration correction (mainly floor unevenness). All images were reconstructed using penalized least-squares minimization with the conjugate gradient algorithm from RTK with 0.5 × 0.5 × 0.5 mm 3 $0.5 \times 0.5\times 0.5 \text{ mm}^3$ voxel size. A disk phantom and an Alderson phantom were scanned to assess the image quality. Results were correlated with the local incompleteness value represented by tan ( ψ ) $\tan (\psi)$ , which was calculated at each voxel as a function of the source trajectory and the voxel's 3D coordinates. We assessed the magnitude of CB artifacts using the full width half maximum (FWHM) of each disk profile in the axial center of the reconstructed images. Spatial resolution was also quantified by the modulation transfer function at 10% (MTF10). RESULTS: When using the saddle trajectory, the region without CB artifacts was increased from 43 to 190 mm in the SI direction compared to the circular trajectory. This region coincided with low values for tan ( ψ ) $\tan (\psi)$ . When tan ( ψ ) $\tan (\psi)$ was larger than 0.02, we found there was a linear relationship between the FWHM and tan ( ψ ) $\tan (\psi)$ . For the saddle, IR tracking allowed the increase of MTF10 from 0.37 to 0.98 lp/mm. CONCLUSIONS: We achieved saddle trajectory CBCT scans with a novel CBCT system combined with IR tracking. The results show that the saddle trajectory provides a larger region with reliable reconstruction compared to the circular trajectory. The proposed method can be used to evaluate other non-circular trajectories.

Comparison of deep learning networks for fully automated head and neck tumor delineation on multi-centric PET/CT images.

OBJECTIVES: Deep learning-based auto-segmentation of head and neck cancer (HNC) tumors is expected to have better reproducibility than manual delineation. Positron emission tomography (PET) and computed tomography (CT) are commonly used in tumor segmentation. However, current methods still face challenges in handling whole-body scans where a manual selection of a bounding box may be required. Moreover, different institutions might still apply different guidelines for tumor delineation. This study aimed at exploring the auto-localization and segmentation of HNC tumors from entire PET/CT scans and investigating the transferability of trained baseline models to external real world cohorts. METHODS: We employed 2D Retina Unet to find HNC tumors from whole-body PET/CT and utilized a regular Unet to segment the union of the tumor and involved lymph nodes. In comparison, 2D/3D Retina Unets were also implemented to localize and segment the same target in an end-to-end manner. The segmentation performance was evaluated via Dice similarity coefficient (DSC) and Hausdorff distance 95th percentile (HD95). Delineated PET/CT scans from the HECKTOR challenge were used to train the baseline models by 5-fold cross-validation. Another 271 delineated PET/CTs from three different institutions (MAASTRO, CRO, BERLIN) were used for external testing. Finally, facility-specific transfer learning was applied to investigate the improvement of segmentation performance against baseline models. RESULTS: Encouraging localization results were observed, achieving a maximum omnidirectional tumor center difference lower than 6.8 cm for external testing. The three baseline models yielded similar averaged cross-validation (CV) results with a DSC in a range of 0.71-0.75, while the averaged CV HD95 was 8.6, 10.7 and 9.8 mm for the regular Unet, 2D and 3D Retina Unets, respectively. More than a 10% drop in DSC and a 40% increase in HD95 were observed if the baseline models were tested on the three external cohorts directly. After the facility-specific training, an improvement in external testing was observed for all models. The regular Unet had the best DSC (0.70) for the MAASTRO cohort, and the best HD95 (7.8 and 7.9 mm) in the MAASTRO and CRO cohorts. The 2D Retina Unet had the best DSC (0.76 and 0.67) for the CRO and BERLIN cohorts, and the best HD95 (12.4 mm) for the BERLIN cohort. CONCLUSION: The regular Unet outperformed the other two baseline models in CV and most external testing cohorts. Facility-specific transfer learning can potentially improve HNC segmentation performance for individual institutions, where the 2D Retina Unets could achieve comparable or even better results than the regular Unet.

Simultaneous object detection and segmentation for patient-specific markerless lung tumor tracking in simulated radiographs with deep learning.

BACKGROUND: Real-time tumor tracking is one motion management method to address motion-induced uncertainty. To date, fiducial markers are often required to reliably track lung tumors with X-ray imaging, which carries risks of complications and leads to prolonged treatment time. A markerless tracking approach is thus desirable. Deep learning-based approaches have shown promise for markerless tracking, but systematic evaluation and procedures to investigate applicability in individual cases are missing. Moreover, few efforts have been made to provide bounding box prediction and mask segmentation simultaneously, which could allow either rigid or deformable multi-leaf collimator tracking. PURPOSE: The purpose of this study was to implement a deep learning-based markerless lung tumor tracking model exploiting patient-specific training which outputs both a bounding box and a mask segmentation simultaneously. We also aimed to compare the two kinds of predictions and to implement a specific procedure to understand the feasibility of markerless tracking on individual cases. METHODS: We first trained a Retina U-Net baseline model on digitally reconstructed radiographs (DRRs) generated from a public dataset containing 875 CT scans and corresponding lung nodule annotations. Afterwards, we used an independent cohort of 97 lung patients to develop a patient-specific refinement procedure. In order to determine the optimal hyperparameters for automatic patient-specific training, we selected 13 patients for validation where the baseline model predicted a bounding box on planning CT (PCT)-DRR with intersection over union (IoU) with the ground-truth higher than 0.7. The final test set contained the remaining 84 patients with varying PCT-DRR IoU. For each testing patient, the baseline model was refined on the PCT-DRR to generate a patient-specific model, which was then tested on a separate 10-phase 4DCT-DRR to mimic the intrafraction motion during treatment. A template matching algorithm served as benchmark model. The testing results were evaluated by four metrics: the center of mass (COM) error and the Dice similarity coefficient (DSC) for segmentation masks, and the center of box (COB) error and the DSC for bounding box detections. Performance was compared to the benchmark model including statistical testing for significance. RESULTS: A PCT-DRR IoU value of 0.2 was shown to be the threshold dividing inconsistent (68%) and consistent (100%) success (defined as mean bounding box DSC > 0.6) of PS models on 4DCT-DRRs. Thirty-seven out of the eighty-four testing cases had a PCT-DRR IoU above 0.2. For these 37 cases, the mean COM error was 2.6 mm, the mean segmentation DSC was 0.78, the mean COB error was 2.7 mm, and the mean box DSC was 0.83. Including the validation cases, the model was applicable to 50 out of 97 patients when using the PCT-DRR IoU threshold of 0.2. The inference time per frame was 170 ms. The model outperformed the benchmark model on all metrics, and the comparison was significant (p < 0.001) over the 37 PCT-DRR IoU > 0.2 cases, but not over the undifferentiated 84 testing cases. CONCLUSIONS: The implemented patient-specific refinement approach based on a pre-trained baseline model was shown to be applicable to markerless tumor tracking in simulated radiographs for lung cases.

Intra-frame motion deterioration effects and deep-learning-based compensation in MR-guided radiotherapy.

BACKGROUND: Current commercially available hybrid magnetic resonance linear accelerators (MR-Linac) use 2D+t cine MR imaging to provide intra-fractional motion monitoring. However, given the limited temporal resolution of cine MR imaging, target intra-frame motion deterioration effects, resulting in effective time latency and motion artifacts in the image domain, can be appreciable, especially in the case of fast breathing. PURPOSE: The aim of this work is to investigate intra-frame motion deterioration effects in MR-guided radiotherapy (MRgRT) by simulating the motion-corrupted image acquisition, and to explore the feasibility of deep-learning-based compensation approaches, relying on the intra-frame motion information which is spatially and temporally encoded in the raw data (k-space). METHODS: An intra-frame motion model was defined to simulate motion-corrupted MR images, with 4D anthropomorphic digital phantoms being exploited to provide ground truth 2D+t cine MR sequences. A total number of 10 digital phantoms were generated for lung cancer patients, with randomly selected eight patients for training or validation and the remaining two for testing. The simulation code served as the data generator, and a dedicated motion pattern perturbation scheme was proposed to build the intra-frame motion database, where three degrees of freedom were designed to guarantee the diversity of intra-frame motion trajectories, enabling a thorough exploration in the domain of the potential anatomical structure positions. U-Nets with three types of loss functions: L1 or L2 loss defined in image or Fourier domain, referred to as NNImgLoss-L1 , NNFloss-L1 and NNL2-Loss were trained to extract information from the motion-corrupted image and used to estimate the ground truth final-position image, corresponding to the end of the acquisition. Images before and after compensation were evaluated in terms of (i) image mean-squared error (MSE) and mean absolute error (MAE), and (ii) accuracy of gross tumor volume (GTV) contouring, based on optical-flow image registration. RESULTS: Image degradation caused by intra-frame motion was observed: for a linearly and fully acquired Cartesian readout k-space trajectory, intra-frame motion resulted in an imaging latency of approximately 50% of the acquisition time; in comparison, the motion artifacts exhibited only a negligible contribution to the overall geometric errors. All three compensation models led to a decrease in image MSE/MAE and GTV position offset compared to the motion-corrupted image. In the investigated testing dataset for GTV contouring, the average dice similarity coefficients (DSC) improved from 88% to 96%, and the 95th percentile Hausdorff distance (HD95 ) dropped from 4.8 mm to 2.1 mm. Different models showed slight performance variations across different intra-frame motion amplitude categories: NNImgLoss-L1 excelled for small/medium amplitudes, whereas NNFloss-L1 demonstrated higher DSC median values at larger amplitudes. The saliency maps of the motion-corrupted image highlighted the major contribution of the later acquired k-space data, as well as the edges of the moving anatomical structures at their final positions, during the model inference stage. CONCLUSIONS: Our results demonstrate the deep-learning-based approaches have the potential to compensate for intra-frame motion by utilizing the later acquired data to drive the convergence of the earlier acquired k-space components.

Real-time motion management in MRI-guided radiotherapy: Current status and AI-enabled prospects.

MRI-guided radiotherapy (MRIgRT) is a highly complex treatment modality, allowing adaptation to anatomical changes occurring from one treatment day to the other (inter-fractional), but also to motion occurring during a treatment fraction (intra-fractional). In this vision paper, we describe the different steps of intra-fractional motion management during MRIgRT, from imaging to beam adaptation, and the solutions currently available both clinically and at a research level. Furthermore, considering the latest developments in the literature, a workflow is foreseen in which motion-induced over- and/or under-dosage is compensated in 3D, with minimal impact to the radiotherapy treatment time. Considering the time constraints of real-time adaptation, a particular focus is put on artificial intelligence (AI) solutions as a fast and accurate alternative to conventional algorithms.

Experimental comparison of linear regression and LSTM motion prediction models for MLC-tracking on an MRI-linac.

BACKGROUND: Magnetic resonance imaging (MRI)-guided radiotherapy with multileaf collimator (MLC)-tracking is a promising technique for intra-fractional motion management, achieving high dose conformality without prolonging treatment times. To improve beam-target alignment, the geometric error due to system latency should be reduced by using temporal prediction. PURPOSE: To experimentally compare linear regression (LR) and long-short-term memory (LSTM) motion prediction models for MLC-tracking on an MRI-linac using multiple patient-derived traces with different complexities. METHODS: Experiments were performed on a prototype 1.0 T MRI-linac capable of MLC-tracking. A motion phantom was programmed to move a target in superior-inferior (SI) direction according to eight lung cancer patient respiratory motion traces. Target centroid positions were localized from sagittal 2D cine MRIs acquired at 4 Hz using a template matching algorithm. The centroid positions were input to one of four motion prediction models. We used (1) a LSTM network which had been optimized in a previous study on patient data from another cohort (offline LSTM). We also used (2) the same LSTM model as a starting point for continuous re-optimization of its weights during the experiment based on recent motion (offline+online LSTM). Furthermore, we implemented (3) a continuously updated LR model, which was solely based on recent motion (online LR). Finally, we used (4) the last available target centroid without any changes as a baseline (no-predictor). The predictions of the models were used to shift the MLC aperture in real-time. An electronic portal imaging device (EPID) was used to visualize the target and MLC aperture during the experiments. Based on the EPID frames, the root-mean-square error (RMSE) between the target and the MLC aperture positions was used to assess the performance of the different motion predictors. Each combination of motion trace and prediction model was repeated twice to test stability, for a total of 64 experiments. RESULTS: The end-to-end latency of the system was measured to be (389 ± 15) ms and was successfully mitigated by both LR and LSTM models. The offline+online LSTM was found to outperform the other models for all investigated motion traces. It obtained a median RMSE over all traces of (2.8 ± 1.3) mm, compared to the (3.2 ± 1.9) mm of the offline LSTM, the (3.3 ± 1.4) mm of the online LR and the (4.4 ± 2.4) mm when using the no-predictor. According to statistical tests, differences were significant (p-value <0.05) among all models in a pair-wise comparison, but for the offline LSTM and online LR pair. The offline+online LSTM was found to be more reproducible than the offline LSTM and the online LR with a maximum deviation in RMSE between two measurements of 10%. CONCLUSIONS: This study represents the first experimental comparison of different prediction models for MRI-guided MLC-tracking using several patient-derived respiratory motion traces. We have shown that among the investigated models, continuously re-optimized LSTM networks are the most promising to account for the end-to-end system latency in MRI-guided radiotherapy with MLC-tracking.

Continuous time-resolved estimated synthetic 4D-CTs for dose reconstruction of lung tumor treatments at a 0.35 T MR-linac.

Objective.To experimentally validate a method to create continuous time-resolved estimated synthetic 4D-computed tomography datasets (tresCTs) based on orthogonal cine MRI data for lung cancer treatments at a magnetic resonance imaging (MRI) guided linear accelerator (MR-linac).Approach.A breathing porcine lung phantom was scanned at a CT scanner and 0.35 T MR-linac. Orthogonal cine MRI series (sagittal/coronal orientation) at 7.3 Hz, intersecting tumor-mimicking gelatin nodules, were deformably registered to mid-exhale 3D-CT and 3D-MRI datasets. The time-resolved deformation vector fields were extrapolated to 3D and applied to a reference synthetic 3D-CT image (sCTref), while accounting for breathing phase-dependent lung density variations, to create 82 s long tresCTs at 3.65 Hz. Ten tresCTs were created for ten tracked nodules with different motion patterns in two lungs. For each dataset, a treatment plan was created on the mid-exhale phase of a measured ground truth (GT) respiratory-correlated 4D-CT dataset with the tracked nodule as gross tumor volume (GTV). Each plan was recalculated on the GT 4D-CT, randomly sampled tresCT, and static sCTrefimages. Dose distributions for corresponding breathing phases were compared in gamma (2%/2 mm) and dose-volume histogram (DVH) parameter analyses.Main results.The mean gamma pass rate between all tresCT and GT 4D-CT dose distributions was 98.6%. The mean absolute relative deviations of the tresCT with respect to GT DVH parameters were 1.9%, 1.0%, and 1.4% for the GTVD98%,D50%, andD2%, respectively, 1.0% for the remaining nodulesD50%, and 1.5% for the lungV20Gy. The gamma pass rate for the tresCTs was significantly larger (p< 0.01), and the GTVD50%deviations with respect to the GT were significantly smaller (p< 0.01) than for the sCTref.Significance.The results suggest that tresCTs could be valuable for time-resolved reconstruction and intrafractional accumulation of the dose to the GTV for lung cancer patients treated at MR-linacs in the future.

Deep learning based automatic segmentation of organs-at-risk for 0.35 T MRgRT of lung tumors.

BACKGROUND AND PURPOSE: Magnetic resonance imaging guided radiotherapy (MRgRT) offers treatment plan adaptation to the anatomy of the day. In the current MRgRT workflow, this requires the time consuming and repetitive task of manual delineation of organs-at-risk (OARs), which is also prone to inter- and intra-observer variability. Therefore, deep learning autosegmentation (DLAS) is becoming increasingly attractive. No investigation of its application to OARs in thoracic magnetic resonance images (MRIs) from MRgRT has been done so far. This study aimed to fill this gap. MATERIALS AND METHODS: 122 planning MRIs from patients treated at a 0.35 T MR-Linac were retrospectively collected. Using an 80/19/23 (training/validation/test) split, individual 3D U-Nets for segmentation of the left lung, right lung, heart, aorta, spinal canal and esophagus were trained. These were compared to the clinically used contours based on Dice similarity coefficient (DSC) and Hausdorff distance (HD). They were also graded on their clinical usability by a radiation oncologist. RESULTS: Median DSC was 0.96, 0.96, 0.94, 0.90, 0.88 and 0.78 for left lung, right lung, heart, aorta, spinal canal and esophagus, respectively. Median 95th percentile values of the HD were 3.9, 5.3, 5.8, 3.0, 2.6 and 3.5 mm, respectively. The physician preferred the network generated contours over the clinical contours, deeming 85 out of 129 to not require any correction, 25 immediately usable for treatment planning, 15 requiring minor and 4 requiring major corrections. CONCLUSIONS: We trained 3D U-Nets on clinical MRI planning data which produced accurate delineations in the thoracic region. DLAS contours were preferred over the clinical contours.

Multi-stage image registration based on list-mode proton radiographies for small animal proton irradiation: A simulation study.

We present a multi-stage and multi-resolution deformable image registration framework for image-guidance at a small animal proton irradiation platform. The framework is based on list-mode proton radiographies acquired at different angles, which are used to deform a 3D treatment planning CT relying on normalized mutual information (NMI) or root mean square error (RMSE) in the projection domain. We utilized a mouse X-ray micro-CT expressed in relative stopping power (RSP), and obtained Monte Carlo simulations of proton images in list-mode for three different treatment sites (brain, head and neck, lung). Rigid transformations and controlled artificial deformation were applied to mimic position misalignments, weight loss and breathing changes. Results were evaluated based on the residual RMSE of RSP in the image domain including the comparison of extracted local features, i.e. between the reference micro-CT and the one transformed taking into account the calculated deformation. The residual RMSE of the RSP showed that the accuracy of the registration framework is promising for compensating rigid (>97% accuracy) and non-rigid (∼95% accuracy) transformations with respect to a conventional 3D-3D registration. Results showed that the registration accuracy is degraded when considering the realistic detector performance and NMI as a metric, whereas the RMSE in projection domain is rather insensitive. This work demonstrates the pre-clinical feasibility of the registration framework on different treatment sites and its use for small animal imaging with a realistic detector. Further computational optimization of the framework is required to enable the use of this tool for online estimation of the deformation.

On the robustness of multilateration of ionoacoustic signals for localization of the Bragg peak at pre-clinical proton beam energies in water.

Objectives.The energy deposited in a medium by a pulsed proton beam results in the emission of thermoacoustic waves, also called ionoacoustics (IA). The proton beam stopping position (Bragg peak) can be retrieved from a time-of-flight analysis (ToF) of IA signals acquired at different sensor locations (multilateration). This work aimed to assess the robustness of multilateration methods in proton beams at pre-clinical energies for the development of a small animal irradiator.Approach.The accuracy of multilateration performed using different algorithms; namely, time of arrival and time difference of arrival, was investigatedin-silicofor ideal point sources in the presence of realistic uncertainties on the ToF estimation and ionoacoustic signals generated by a 20 MeV pulsed proton beam stopped in a homogeneous water phantom. The localisation accuracy was further investigated experimentally based on two different measurements with pulsed monoenergetic proton beams at energies of 20 and 22 MeV.Main results.It was found that the localisation accuracy mainly depends on the position of the acoustic detectors relative to the proton beam due to spatial variation of the error on the ToF estimation. By optimally positioning the sensors to reduce the ToF error, the Bragg peak could be locatedin-silicowith an accuracy better than 90µm (2% error). Localisation errors going up to 1 mm were observed experimentally due to inaccurate knowledge of the sensor positions and noisy ionoacoustic signals.Significance.This study gives a first overview of the implementation of different multilateration methods for ionoacoustics-based Bragg peak localisation in two- and three-dimensions at pre-clinical energies. Different sources of uncertainty were investigated, and their impact on the localisation accuracy was quantifiedin-silicoand experimentally.

Evaluation of real-time tumor contour prediction using LSTM networks for MR-guided radiotherapy.

BACKGROUND AND PURPOSE: Magnetic resonance imaging guided radiotherapy (MRgRT) with deformable multileaf collimator (MLC) tracking would allow to tackle both rigid displacement and tumor deformation without prolonging treatment. However, the system latency must be accounted for by predicting future tumor contours in real-time. We compared the performance of three artificial intelligence (AI) algorithms based on long short-term memory (LSTM) modules for the prediction of 2D-contours 500ms into the future. MATERIALS AND METHODS: Models were trained (52 patients, 3.1h of motion), validated (18 patients, 0.6h) and tested (18 patients, 1.1h) with cine MRs from patients treated at one institution. Additionally, we used three patients (2.9h) treated at another institution as second testing set. We implemented 1) a classical LSTM network (LSTM-shift) predicting tumor centroid positions in superior-inferior and anterior-posterior direction which are used to shift the last observed tumor contour. The LSTM-shift model was optimized both in an offline and online fashion. We also implemented 2) a convolutional LSTM model (ConvLSTM) to directly predict future tumor contours and 3) a convolutional LSTM combined with spatial transformer layers (ConvLSTM-STL) to predict displacement fields used to warp the last tumor contour. RESULTS: The online LSTM-shift model was found to perform slightly better than the offline LSTM-shift and significantly better than the ConvLSTM and ConvLSTM-STL. It achieved a 50% Hausdorff distance of 1.2mm and 1.0mm for the two testing sets, respectively. Larger motion ranges were found to lead to more substantial performance differences across the models. CONCLUSION: LSTM networks predicting future centroids and shifting the last tumor contour are the most suitable for tumor contour prediction. The obtained accuracy would allow to reduce residual tracking errors during MRgRT with deformable MLC-tracking.

Assessment of intrafractional prostate motion and its dosimetric impact in MRI-guided online adaptive radiotherapy with gating.

PURPOSE: This study aimed to evaluate the intrafractional prostate motion captured during gated magnetic resonance imaging (MRI)-guided online adaptive radiotherapy for prostate cancer and analyze its impact on the delivered dose as well as the effect of gating. METHODS: Sagittal 2D cine-MRI scans were acquired at 4 Hz during treatment at a ViewRay MRIdian (ViewRay Inc., Oakwood Village, OH, USA) MR linac. Prostate shifts in anterior-posterior (AP) and superior-inferior (SI) directions were extracted separately. Using the static dose cloud approximation, the planned fractional dose was shifted according to the 2D gated motion (residual motion in gating window) to estimate the delivered dose by superimposing and averaging the shifted dose volumes. The dose of a hypothetical non-gated delivery was reconstructed similarly using the non-gated motion. For the clinical target volume (CTV), rectum, and bladder, dose-volume histogram parameters of the planned and reconstructed doses were compared. RESULTS: In total, 174 fractions (15.7 h of cine-MRI) from 10 patients were evaluated. The average (±1 σ) non-gated prostate motion was 0.6 ± 1.0 mm in the AP and 0.0 ± 0.6 mm in the SI direction with respect to the centroid position of the gating boundary. 95% of the shifts were within [-3.5, 2.7] mm in the AP and [-2.9, 3.2] mm in the SI direction. For the gated treatment and averaged over all fractions, CTV D98% decreased by less than 2% for all patients. The rectum and the bladder D2% increased by less than 3% and 0.5%, respectively. Doses reconstructed for gated and non-gated delivery were similar for most fractions. CONCLUSION: A pipeline for extraction of prostate motion during gated MRI-guided radiotherapy based on 2D cine-MRI was implemented. The 2D motion data enabled an approximate estimation of the delivered dose. For the majority of fractions, the benefit of gating was negligible, and clinical dosimetric constraints were met, indicating safety of the currently adopted gated MRI-guided treatment workflow.

Impact of secondary particles on the magnetic field generated by a proton pencil beam: a finite-element analysis based on Geant4-DNA simulations.

PURPOSE: To investigate the static magnetic field generated by a proton pencil beam as a candidate for range verification by means of Monte Carlo simulations, thereby improving upon existing analytical calculations. We focus on the impact of statistical current fluctuations and secondary protons and electrons. METHODS: We considered a pulsed beam (10 µ ${\umu}$ s pulse duration) during the duty cycle with a peak beam current of 0.2 µ $\umu$ A and an initial energy of 100 MeV. We ran Geant4-DNA Monte Carlo simulations of a proton pencil beam in water and extracted independent particle phase spaces. We calculated longitudinal and radial current density of protons and electrons, serving as an input for a magnetic field estimation based on a finite element analysis in a cylindrical geometry. We made sure to allow for non-solenoidal current densities as is the case of a stopping proton beam. RESULTS: The rising proton charge density toward the range is not perturbed by energy straggling and only lowered through nuclear reactions by up to 15%, leading to an approximately constant longitudinal current. Their relative low density however (at most 0.37 protons/mm3 for the 0.2  µ ${\umu}$ A current and a beam cross-section of 2.5 mm), gives rise to considerable current density fluctuations. The radial proton current resulting from lateral scattering and being two orders of magnitude weaker than the longitudinal current is subject to even stronger fluctuations. Secondary electrons with energies above 10 eV, that far outnumber the primary protons, reduce the primary proton current by only 10% due to their largely isotropic flow. A small fraction of electrons (<1%), undergoing head-on collisions, constitutes the relevant electron current. In the far-field, both contributions to the magnetic field strength (longitudinal and lateral) are independent of the beam spot size. We also find that the nuclear reaction-related losses cause a shift of 1.3 mm to the magnetic field profile relative to the actual range, which is further enlarged to 2.4 mm by the electron current (at a distance of ρ = 50 $\rho =50$  mm away from the central beam axis). For ρ > 45 $\rho >45$  mm, the shift increases linearly. While the current density variations cause significant magnetic field uncertainty close to the central beam axis with a relative standard deviation (RSD) close to 100%, they average out at a distance of 10 cm, where the RSD of the total magnetic field drops below 2%. CONCLUSIONS: With the small influence of the secondary electrons together with the low RSD, our analysis encourages an experimental detection of the magnetic field through sensitive instrumentation, such as optical magnetometry or SQUIDs.

18F-FET PET radiomics-based survival prediction in glioblastoma patients receiving radio(chemo)therapy.

BACKGROUND: Quantitative image analysis based on radiomic feature extraction is an emerging field for survival prediction in oncological patients. 18F-Fluorethyltyrosine positron emission tomography (18F-FET PET) provides important diagnostic and grading information for brain tumors, but data on its use in survival prediction is scarce. In this study, we aim at investigating survival prediction based on multiple radiomic features in glioblastoma patients undergoing radio(chemo)therapy. METHODS: A dataset of 37 patients with glioblastoma (WHO grade 4) receiving radio(chemo)therapy was analyzed. Radiomic features were extracted from pre-treatment 18F-FET PET images, following intensity rebinning with a fixed bin width. Principal component analysis (PCA) was applied for variable selection, aiming at the identification of the most relevant features in survival prediction. Random forest classification and prediction algorithms were optimized on an initial set of 25 patients. Testing of the implemented algorithms was carried out in different scenarios, which included additional 12 patients whose images were acquired with a different scanner to check the reproducibility in prediction results. RESULTS: First order intensity variations and shape features were predominant in the selection of most important radiomic signatures for survival prediction in the available dataset. The major axis length of the 18F-FET-PET volume at tumor to background ratio (TBR) 1.4 and 1.6 correlated significantly with reduced probability of survival. Additional radiomic features were identified as potential survival predictors in the PTV region, showing 76% accuracy in independent testing for both classification and regression. CONCLUSIONS: 18F-FET PET prior to radiation provides relevant information for survival prediction in glioblastoma patients. Based on our preliminary analysis, radiomic features in the PTV can be considered a robust dataset for survival prediction.

DeepClassPathway: Molecular pathway aware classification using explainable deep learning.

OBJECTIVE: HPV-associated head and neck cancer is correlated with favorable prognosis; however, its underlying biology is not fully understood. We propose an explainable convolutional neural network (CNN) classifier, DeepClassPathway, that predicts HPV-status and allows patient-specific identification of molecular pathways driving classifier decisions. METHODS: The CNN was trained to classify HPV-status on transcriptome data from 264 (13% HPV-positive) and tested on 85 (25% HPV-positive) head and neck squamous carcinoma patients after transformation into 2D-treemaps representing molecular pathways. Grad-CAM saliency was used to quantify pathways contribution to individual CNN decisions. Model stability was assessed by shuffling pathways within 2D-images. RESULTS: The classification performance of the CNN-ensembles achieved ROC-AUC/PR-AUC of 0.96/0.90 for all treemap variants. Quantification of the averaged pathway saliency heatmaps consistently identified KRAS, spermatogenesis, bile acid metabolism, and inflammation signaling pathways as the four most informative for classifying HPV-positive patients and MYC targets, epithelial-mesenchymal transition, and protein secretion pathways for HPV-negative patients. CONCLUSION: We have developed and applied an explainable CNN classification approach to transcriptome data from an oncology cohort with typical sample size that allows classification while accounting for the importance of molecular pathways in individual-level decisions.

Deep learning based time-to-event analysis with PET, CT and joint PET/CT for head and neck cancer prognosis.

OBJECTIVES: Recent studies have shown that deep learning based on pre-treatment positron emission tomography (PET) or computed tomography (CT) is promising for distant metastasis (DM) and overall survival (OS) prognosis in head and neck cancer (HNC). However, lesion segmentation is typically required, resulting in a predictive power susceptible to variations in primary and lymph node gross tumor volume (GTV) segmentation. This study aimed at achieving prognosis without GTV segmentation, and extending single modality prognosis to joint PET/CT to allow investigating the predictive performance of combined- compared to single-modality inputs. METHODS: We employed a 3D-Resnet combined with a time-to-event outcome model to incorporate censoring information. We focused on the prognosis of DM and OS for HNC patients. For each clinical endpoint, five models with PET and/or CT images as input were compared: PET-GTV, PET-only, CT-GTV, CT-only, and PET/CT-GTV models, where -GTV indicates that the corresponding images were masked using the GTV contour. Publicly available delineated CT and PET scans from 4 different Canadian hospitals (293) and the MAASTRO clinic (74) were used for training by 3-fold cross-validation (CV). For independent testing, we used 110 patients from a collaborating institution. The predictive performance was evaluated via Harrell's Concordance Index (HCI) and Kaplan-Meier curves. RESULTS: In a 5-year time-to-event analysis, all models could produce CV HCIs with median values around 0.8 for DM and 0.7 for OS. The best performance was obtained with the PET-only model, achieving a median testing HCI of 0.82 for DM and 0.69 for OS. Compared with the PET/CT-GTV model, the PET-only still had advantages of up to 0.07 in terms of testing HCI. The Kaplan-Meier curves and corresponding log-rank test results also demonstrated significant stratification capability of our models for the testing cohort. CONCLUSION: Deep learning-based DM and OS time-to-event models showed predictive capability and could provide indications for personalized RT. The best predictive performance achieved by the PET-only model suggested GTV segmentation might be less relevant for PET-based prognosis.

Single-isocenter stereotactic radiosurgery for multiple brain metastases: Impact of patient misalignments on target coverage in non-coplanar treatments.

Frameless single-isocenter non-coplanar stereotactic radiosurgery (SRS) for patients with multiple brain metastases is a treatment at high geometrical complexity. The goal of this study is to analyze the dosimetric impact of non-coplanar image guidance with stereoscopic X-ray imaging. Such an analysis is meant to provide insights on the adequacy of safety margins, and to evaluate the benefit of imaging at non-coplanar configurations. The ExacTrac® (ET) system (Brainlab AG, Munich, Germany) was used for stereoscopic X-ray imaging in frameless single-isocenter non-coplanar SRS for multiple brain metastases. Sub-millimeter precision was found for the ET-based pre-treatment setup, whereas a degradation was noted for non-coplanar treatment angles. Misalignments without intra-fractional positioning corrections were reconstructed in 6 degrees of freedom (DoF) to resemble the situation without non-coplanar image guidance. Dose recalculation in 20 SRS patients with applied positioning corrections did not reveal any significant differences in D98% for 75 planning target volumes (PTVs) and gross tumor volumes (GTVs). For recalculation without applied positioning corrections, significant differences (p<0.05) were reported in D98% for both PTVs and GTVs, with stronger effects for small PTV volumes. A worst-case analysis at increasing translational and rotational misalignment revealed that dosimetric changes are a complex function of the combination thereof. This study highlighted the important role of positioning correction with ET at non-coplanar configurations in frameless single-isocenter non-coplanar SRS for patients with multiple brain metastases. Uncorrected patient misalignments at non-coplanar couch angles were linked to a significant loss of PTV coverage, with effects varying according to the combination of single DoF and PTV geometrical properties.

Offline and online LSTM networks for respiratory motion prediction in MR-guided radiotherapy.

Objective.Gated beam delivery is the current clinical practice for respiratory motion compensation in MR-guided radiotherapy, and further research is ongoing to implement tracking. To manage intra-fractional motion using multileaf collimator tracking the total system latency needs to be accounted for in real-time. In this study, long short-term memory (LSTM) networks were optimized for the prediction of superior-inferior tumor centroid positions extracted from clinically acquired 2D cine MRIs.Approach.We used 88 patients treated at the University Hospital of the LMU Munich for training and validation (70 patients, 13.1 h), and for testing (18 patients, 3.0 h). Three patients treated at Fondazione Policlinico Universitario Agostino Gemelli were used as a second testing set (1.5 h). The performance of the LSTMs in terms of root mean square error (RMSE) was compared to baseline linear regression (LR) models for forecasted time spans of 250 ms, 500 ms and 750 ms. Both the LSTM and the LR were trained with offline (offlineLSTM andofflineLR) and online schemes (offline+onlineLSTM andonlineLR), the latter to allow for continuous adaptation to recent respiratory patterns.Main results.We found theoffline+onlineLSTM to perform best for all investigated forecasts. Specifically, when predicting 500 ms ahead it achieved a mean RMSE of 1.20 mm and 1.00 mm, while the best performing LR model achieved a mean RMSE of 1.42 mm and 1.22 mm for the LMU and Gemelli testing set, respectively.Significance.This indicates that LSTM networks have potential as respiratory motion predictors and that continuous online re-optimization can enhance their performance.

X-ray CT adaptation based on a 2D-3D deformable image registration framework using simulated in-room proton radiographies.

The aim of this work is to investigate in-room proton radiographies to compensate realistic rigid and non-rigid transformations in clinical-like scenarios based on 2D-3D deformable image registration (DIR) framework towards future clinical implementation of adaptive radiation therapy (ART). Monte Carlo simulations of proton radiographies (pRads) based on clinical x-ray CT of a head and neck, and a brain tumor patients are simulated for two different detector configurations (i.e. integration-mode and list-mode detectors) including high and low proton statistics. A realistic deformation, derived from cone beam CT of the patient, is applied to the treatment planning CT. Rigid inaccuracies in patient positioning are also applied and the effect of small, medium and large fields of view (FOVs) is investigated. A stopping criterion, as desirable in realistic scenarios devoid of ground truth proton CT (pCT), is proposed and investigated. Results show that rigid and non-rigid transformations can be compensated based on a limited number of low dose pRads. The root mean square error with respect to the pCT shows that the 2D-3D DIR of the treatment planning CT based on 10 pRads from integration-mode data and 2 pRads from list-mode data is capable of achieving comparable accuracy (∼90% and >90%, respectively) to conventional 3D-3D DIR. The dice similarity coefficient over the segmented regions of interest also verifies the improvement in accuracy prior to and after 2D-3D DIR. No relevant changes in accuracy are found between high and low proton statistics except for 2 pRads from integration-mode data. The impact of FOV size is negligible. The convergence of the metric adopted for the stopping criterion indicates the optimal convergence of the 2D-3D DIR. This work represents a further step towards the potential implementation of ART in proton therapy. Further computational optimization is however required to enable extensive clinical validation.

Dosimetric impact of geometric distortions in an MRI-only proton therapy workflow for lung, liver and pancreas.

In a radiation therapy workflow based on Magnetic Resonance Imaging (MRI), dosimetric errors may arise due to geometric distortions introduced by MRI. The aim of this study was to quantify the dosimetric effect of system-dependent geometric distortions in an MRI-only workflow for proton therapy applied at extra-cranial sites. An approach was developed, in which computed tomography (CT) images were distorted using an MRI displacement map, which represented the MR distortions in a spoiled gradient-echo sequence due to gradient nonlinearities and static magnetic field inhomogeneities. A retrospective study was conducted on 4DCT/MRI digital phantoms and 18 4DCT clinical datasets of the thoraco-abdominal site. The treatment plans were designed and separately optimized for each beam in a beam specific Planning Target Volume on the distorted CT, and the final dose distribution was obtained as the average. The dose was then recalculated in undistorted CT using the same beam geometry and beam weights. The analysis was performed in terms of Dose Volume Histogram (DVH) parameters. No clinically relevant dosimetric impact was observed on organs at risk, whereas in the target structure, geometric distortions caused statistically significant variations in the planned dose DVH parameters and dose homogeneity index (DHI). The dosimetric variations in the target structure were smaller in abdominal cases (ΔD2%, ΔD98%, and ΔDmean all below 0.1% and ΔDHI below 0.003) compared to the lung cases. Indeed, lung patients with tumors isolated inside lung parenchyma exhibited higher dosimetric variations (ΔD2%≥0.3%, ΔD98%≥15.9%, ΔDmean≥3.3% and ΔDHI≥0.102) than lung patients with tumor close to soft tissue (ΔD2%≤0.4%, ΔD98%≤5.6%, ΔDmean≤0.9% and ΔDHI≤0.027) potentially due to higher density variations along the beam path. Results suggest the potential applicability of MRI-only proton therapy, provided that specific analysis is applied for isolated lung tumors.