Target margin design for real-time lung tumor tracking stereotactic body radiation therapy using CyberKnife Xsight Lung Tracking System

Results of Stereotactic Body Radiotherapy With CyberKnife-M6 for Primary and Metastatic Lung Cancer

Background

The aim of the study was to evaluate the efficacy of stereotactic body radiotherapy (SBRT) using the CyberKnife-M6 (CK-M6) with lung optimized treatment (LOT) module in patients with primary lung cancer and lung metastases.

Methods

Forty-two lesions from 35 patients were treated between 2019 and 2022. Four-dimensional computed tomography images were obtained when the patients were in a free breathing modality. Tracking modality was selected prospectively according to the visibility of the target. The median prescribed dose was 48 Gy in four fractions (fx) (28 - 55 Gy/1- 7 fx). The median age was 68 years (47 - 82 years), and 43% of cases were adenocarcinoma. The median lesion size was 15 mm (6 - 36 mm).

Results

Complete, partial and stable responses were obtained as 26%, 62%, and 9.5% at a median of 2 months (1 - 6 months), and 35.5%, 47.5% and 5% at the 12th month evaluation, respectively. Grade 3 and higher toxicity was not observed in any case. The mean and 2-year overall survival (OS) was 31.5 months and 54%, and the local recurrence-free survival (LRFS) was 29.6 months and 51%, respectively. In univariate analysis, target lesion type, complete response (CR), and higher esophagus maximum dose were favorable factors for OS and LRFS (P < 0.05). The CR at 12th month evaluation remained significant in multivariate analysis in terms of OS (hazard ratio = 8.602, 95% confidence interval: 1.05 - 70.01; P = 0.044).

Conclusions

A mean LRFS of 29.6 months and OS of 31.5 months were obtained in patients with primary and metastatic lung cancer. With a median treatment time of 25 min, motion-managed strategy with CK-M6-LOT-based SBRT is an effective, safe, and comfortable treatment method for lung cancer.

Population-based asymmetric margins for moving targets in real-time tumor tracking

BACKGROUND

Both geometric and dosimetric components are commonly considered when determining the margin for planning target volume (PTV). As dose distribution is shaped by controlling beam aperture in peripheral dose prescription and dose-escalated simultaneously integrated boost techniques, adjusting the margin by incorporating the variable dosimetric component into the PTV margin is inappropriate; therefore, geometric components should be accurately estimated for margin calculations.

PURPOSE

We introduced an asymmetric margin-calculation theory using the guide to the expression of uncertainty in measurement (GUM) and intra-fractional motion. The margins in fiducial marker-based real-time tumor tracking (RTTT) for lung, liver, and pancreatic cancers were calculated and were then evaluated using Monte Carlo (MC) simulations.

METHODS

A total of 74 705, 73 235, and 164 968 sets of intra- and inter-fractional positional data were analyzed for 48 lung, 48 liver, and 25 pancreatic cancer patients, respectively, in RTTT clinical trials. The 2.5th and 97.5th percentiles of the positional error were considered representative values of each fraction of the disease site. The population-based statistics of the probability distributions of these representative positional errors (PD-RPEs) were calculated in six directions. A margin covering 95% of the population was calculated using the proposed formula. The content rate in which the clinical target volume (CTV) was included in the PTV was calculated through MC simulations using the PD-RPEs.

RESULTS

The margins required for RTTT were at most 6.2, 4.6, and 3.9 mm for lung, liver, and pancreatic cancer, respectively. MC simulations revealed that the median content rates using the proposed margins satisfied 95% for lung and liver cancers and 93% for pancreatic cancer, closer to the expected rates than the margins according to van Herk's formula.

CONCLUSIONS

Our proposed formula based on the GUM and motion probability distributions (MPD) accurately calculated the practical margin size for fiducial marker-based RTTT. This was verified through MC simulations.

Predicting successful clinical candidates for fiducial-free lung tumor tracking with a deep learning binary classification model

OBJECTIVES

The CyberKnife system is a robotic radiosurgery platform that allows the delivery of lung SBRT treatments using fiducial-free soft-tissue tracking. However, not all lung cancer patients are eligible for lung tumor tracking. Tumor size, density, and location impact the ability to successfully detect and track a lung lesion in 2D orthogonal X-ray images. The standard workflow to identify successful candidates for lung tumor tracking is called Lung Optimized Treatment (LOT) simulation, and involves multiple steps from CT acquisition to the execution of the simulation plan on CyberKnife. The aim of the study is to develop a deep learning classification model to predict which patients can be successfully treated with lung tumor tracking, thus circumventing the LOT simulation process.

METHODS

Target tracking is achieved by matching orthogonal X-ray images with a library of digital radiographs reconstructed from the simulation CT scan (DRRs). We developed a deep learning model to create a binary classification of lung lesions as being trackable or untrackable based on tumor template DRR extracted from the CyberKnife system, and tested five different network architectures. The study included a total of 271 images (230 trackable, 41 untrackable) from 129 patients with one or multiple lung lesions. Eighty percent of the images were used for training, 10% for validation, and the remaining 10% for testing.

RESULTS

For all five convolutional neural networks, the binary classification accuracy reached 100% after training, both in the validation and the test set, without any false classifications.

CONCLUSIONS

A deep learning model can distinguish features of trackable and untrackable lesions in DRR images, and can predict successful candidates for fiducial-free lung tumor tracking.

Assessing the impact of intrafraction motion correction on PTV margins and target and OAR dosimetry for single-fraction free-breathing lung stereotactic body radiation therapy

The objective of this research is to investigate intrafraction motion correction on planning target volume (PTV) margin requirements and target and organ-at-risk (OAR) dosimetry in single-fraction lung stereotactic body radiation therapy (SBRT). Sixteen patients (15 with upper lobe lesions, 1 with a middle lobe lesion) were treated with single-fraction lung SBRT. Cone-beam computed tomography (CBCT) images were acquired before the treatment, between the arcs, and after the delivery of the treatment fraction. Shifts from the reference images were recorded in anterior-posterior (AP), superior-inferior (SI), and lateral (LAT) dimensions. The deviations from the reference image were calculated for 3 clinical scenarios: not applying intratreatment couch shifts and not correcting for pretreatment deviations < 3 mm ( scenario 1), not applying intratreatment couch shifts and correcting for pretreatment deviations < 3 mm ( scenario 2), and applying all pre- and intratreatment couch shifts (scenario 3). PTV margins were determined using the van Herk formalism for each scenario and maximum and average deviations were assessed. The clinical scenarios were modelled in the treatment planning system based on each patient dataset to assess target and OAR dosimetry. Calculated lower-bound PTV margins in the AP, SI, and LAT dimensions were [4.6, 3.5, 2.3] mm in scenario 1, [4.6, 2.4, 2.2] mm in scenario 2, and [1.7, 1.2, 1.0] mm in scenario 3. The margins are lower bounds because they do not include contributions from nonmotion related errors. Average and maximum intrafraction deviations were larger in the AP dimension compared to the SI and LAT dimensions for all scenarios. A unidimensional movement (several mm) in the negative AP dimension was observed in clinical scenarios 1 and 2 but not scenario 3. Average intrafraction deviation vectors were 1.2, 1.1, and 0.3 mm for scenarios 1, 2, and 3, respectively. Modelled clinical scenarios revealed that using scenario 3 yields significantly fewer treatment plan objective failures compared to scenarios 1 and 2 using a Wilcoxon signed-rank test. Intratreatment motion correction between each arc may enable reductions PTV margin requirements. It may also compensate for unidimensional negative AP movement, and improve target and OAR dosimetry.

The impact of the ClearRT™ upgrade on target motion tracking accuracy in Radixact® Synchrony® lung treatments

Background

The objective was to investigate the change in segmentation error of Radixact^® Synchrony^® lung treatment after its kV imaging system was upgraded from Generation 1 to Generation 2 in the ClearRT™ installation.

Materials and methods

Radixact^® Lung Synchrony^® plans were created for the Model 18023 Xsight^® Lung Tracking "XLT" Phantom combined with different lung target inserts with densities of 0.280, 0.500, 0.943 and 1.093 g/cc. After Radixact^® Synchrony^® treatment delivery using the Generation 1 and Generation 2 kV systems according to each plan, the tracking performance of the two kV systems on each density insert was compared by calculating the root mean square (RMS) error (δ_RMS) between the Synchrony-predicted motion in the log file and the known phantom motion and by calculating δ95%, the maximum error within a 95% probability threshold.

Results

The δ_RMS and δ95% of Radixact^® Synchrony^® treatment for Gen1 kV systems deteriorated as the density of the target insert decreased, from 1.673 ± 0.064 mm and 3.049 ± 0.089 mm, respectively, for the 1.093 g/cc insert to 8.355 ± 5.873 mm and 15.297 ± 10.470 mm, respectively, for the 0.280 g/cc insert. In contrast, no such trend was observed in the δ_RMS or δ95% of Synchrony^® treatment using the Gen2 kV system. The δ_RMS and δ95%, respectively, fluctuated slightly from 1.586 to 1.687 mm and from 2.874 to 2.971 mm when different target inserts were tracked by the Gen2 kV system.

Conclusion

With improved image contrast in kV radiographs, the Gen2 kV imaging system can enhance the ability to track targets accurately in Radixact^® Lung Synchrony^® treatment and reduce the segmentation error. Our study showed that lung targets with density values as low as 0.280 cc/g could be tracked correctly in Synchrony treatment with the Gen2 kV imaging system.

CyberKnife Xsight versus fiducial-based target-tracking: a novel 3D dosimetric comparison in a dynamic phantom

Background

The CyberKnife Xsight lung-tracking system (XLTS) provides an alternative to fiducial-based target-tracking systems (FTTS) for non-small-cell lung cancer (NSCLC) patients without invasive fiducial insertion procedures. This study provides a method for 3D independent dosimetric verification of the accuracy of the FTTS compared to the XLTS without relying on log-files generated by the CyberKnife system.

Methods

A respiratory motion trace was taken from a 4D-CT of a real lung cancer patient and applied to a modified QUASAR™ respiratory motion phantom. A novel approach to 3D dosimetry was developed using Gafchromic EBT3 film, allowing the 3D dose distribution delivered to the moving phantom to be reconstructed. Treatments were planned using the recommended margins for one and three fiducial markers and XLTS 2-view, 1-view and 0-view target-tracking modalities. The dose delivery accuracy was analysed by comparing the reconstructed dose distributions to the planned dose distributions using gamma index analysis.

Results

For the 3%/2 mm gamma criterion, gamma passing rates up to 99.37% were observed for the static deliveries. The 3-fiducial and 1-fiducial-based deliveries exhibited passing rates of 93.74% and 97.82%, respectively, in the absence of target rotation. When target rotation was considered, the passing rate for 1-fiducial tracking degraded to 91.24%. The passing rates observed for XLTS 2-view, 1-view and 0-view target-tracking were 92.78%, 96.22% and 76.08%, respectively.

Conclusions

Except for the XLTS 0-view, the dosimetric accuracy of the XLTS was comparable to the FTTS under equivalent treatment conditions. This study gives us further confidence in the CyberKnife XLTS and FTTS systems.

Target margin design through analyzing a large cohort of clinical log data in the cyberknife system

Purpose

Calculating the adequate target margin for real‐time tumor tracking using the Cyberknife system is a challenging issue since different sources of error exist. In this study, the clinical log data of the Cyberknife system were analyzed to adequately quantify the planned target volume (PTV) margins of tumors located in the lung and abdomen regions.

Methods

In this study, 45 patients treated with the Cyberknife module were examined. In this context, adequate PTV margins were estimated based on the Van Herk formulation and the uncertainty estimation method by considering the impact of errors and uncertainties. To investigate the impact of errors and uncertainties on the estimated PTV margins, a statistical analysis was also performed.

Results

Our study demonstrates five different sources of errors, including segmentation, deformation, correlation, prediction, and targeting errors, which were identified as the main sources of error in the Cyberknife system. Furthermore, the clinical evaluation of the current study reveals that the two different formalisms provided almost identical PTV margin estimates. Additionally, 4–5 mm and 5 mm margins on average could provide adequate PTV margins at lung and abdomen tumors in all three directions, respectively. Overall, it was found that concerning the PTV margins, the impact of correlation and prediction errors is very high, while the impact of robotics error is low.

Conclusions

The current study can address two limitations in previous researches, namely insufficient sample sites and a smaller number of patients. A comparison of the present results concerning the lung and abdomen areas with other studies reveals that the proposed strategy could provide a better reference in selection the PTV margins. To our knowledge, this study is one of the first attempts to estimate the PTV margins in the lung and abdomen regions for a large cohort of patients treated using the Cyberknife system.

Commissioning and routine quality assurance of the radixact synchrony system

PURPOSE

The Radixact Synchrony system allows for target motion correction when tracking either fiducials in/around the target or a dense lesion in the lung. As such evaluation testing and Quality Assurance (QA) tests are required.

METHODS

To allow for QA procedures to be performed with a range of available phantoms evaluation of the dosimetric delivery accuracy was performed for a range of motions, phantoms and motion platforms. A CIRS 1D motion platform and Accuray Tomotherapy "cheese" phantom was utilized to perform absolute dose and EBT3 film measurements. A HexaMotion platform and Delta⁴ phantom were utilized to quantify the effects of 1D and 3D motions. Inter-device comparison was performed with the ArcCHECK and Delta⁴ phantoms and GafChromic film, five patient plans were delivered to each phantom when static and with two different motion types both with and without Synchrony motion correction.

RESULTS

A range of QA tests are described. A phantom was designed to allow for daily verification of system functionality. This test allows for detection of either fiducials or a dense silicone target with a stationary phantom. Monthly testing procedures are described that allow the user to verify the dosimetric improvement when utilizing synchrony delivery motion compensation vs. uncorrected motions. These can be performed utilizing a 1D motion stage with an ion-chamber and GafChromic film to allow for a 2D dosimetric validation. Alternatively, a 3D motion platform can be utilized where available. Monthly and annual imaging tests are described. Finally, annual test procedures designed to verify the coincidence of the imaging system and treatment isocenter are described. Evaluation of the Synchrony system using a range of QA devices shows consistently high dosimetric accuracy with similar trends in passing criteria found with GafChromic film, ArcCHECK and Delta⁴ phantoms for density based respiratory model compensation.

CONCLUSION

These results highlight the large improvements in the dose distribution when motion is accounted for with the Synchrony system as measured with a range of phantoms and motion platforms that the majority of users will have available. The testing methods and QA procedures described provide guidance for new users of the Radixact Synchrony system as they implement their own quality assurance programs for this system, until such time as an AAPM task group report is made available. QA procedures including kV imaging quality metrics and imaging dose parameters, dose deposition accuracy, target detection coincidence and target position detection accuracy are described. This article is protected by copyright. All rights reserved.

Lung tumor segmentation in 4D CT images using motion convolutional neural networks

PURPOSE

Manual delineation on all breathing phases of lung cancer 4D CT image datasets can be challenging, exhaustive, and prone to subjective errors because of both the large number of images in the datasets and variations in the spatial location of tumors secondary to respiratory motion. The purpose of this work is to present a new deep learning-based framework for fast and accurate segmentation of lung tumors on 4D CT image sets.

METHODS

The proposed DL framework leverages motion region convolutional neural network (R-CNN). Through integration of global and local motion estimation network architectures, the network can learn both major and minor changes caused by tumor motion. Our network design first extracts tumor motion information by feeding 4D CT images with consecutive phases into an integrated backbone network architecture, locating volume-of-interest (VOIs) via a regional proposal network and removing irrelevant information via a regional convolutional neural network. Extracted motion information is then advanced into the subsequent global and local motion head network architecture to predict corresponding deformation vector fields (DVFs) and further adjust tumor VOIs. Binary masks of tumors are then segmented within adjusted VOIs via a mask head. A self-attention strategy is incorporated in the mask head network to remove any noisy features that might impact segmentation performance. We performed two sets of experiments. In the first experiment, a five-fold cross-validation on 20 4D CT datasets, each consisting of 10 breathing phases (i.e., 200 3D image volumes in total). The network performance was also evaluated on an additional unseen 200 3D images volumes from 20 hold-out 4D CT datasets. In the second experiment, we trained another model with 40 patients' 4D CT datasets from experiment 1 and evaluated on additional unseen nine patients' 4D CT datasets. The Dice similarity coefficient (DSC), center of mass distance (CMD), 95th percentile Hausdorff distance (HD95 ), mean surface distance (MSD), and volume difference (VD) between the manual and segmented tumor contour were computed to evaluate tumor detection and segmentation accuracy. The performance of our method was quantitatively evaluated against four different methods (VoxelMorph, U-Net, network without global and local networks, and network without attention gate strategy) across all evaluation metrics through a paired t-test.

RESULTS

The proposed fully automated DL method yielded good overall agreement with the ground truth for contoured tumor volume and segmentation accuracy. Our model yielded significantly better values of evaluation metrics (p < 0.05) than all four competing methods in both experiments. On hold-out datasets of experiment 1 and 2, our method yielded DSC of 0.86 and 0.90 compared to 0.82 and 0.87, 0.75 and 0.83, 081 and 0.89, and 0.81 and 0.89 yielded by VoxelMorph, U-Net, network without global and local networks, and networks without attention gate strategy. Tumor VD between ground truth and our method was the smallest with the value of 0.50 compared to 0.99, 1.01, 0.92, and 0.93 for between ground truth and VoxelMorph, U-Net, network without global and local networks, and networks without attention gate strategy, respectively.

CONCLUSIONS

Our proposed DL framework of tumor segmentation on lung cancer 4D CT datasets demonstrates a significant promise for fully automated delineation. The promising results of this work provide impetus for its integration into the 4D CT treatment planning workflow to improve the accuracy and efficiency of lung radiotherapy.

Towards Non-Invasive Lung Tumor Tracking Based on Patient Specific Model of Respiratory System

The goal of this paper is to calculate a complex internal respiratory and tumoral movements by measuring respiratory air flows and thorax movements. In this context, we present a new lung tumor tracking approach based on a patient-specific biomechanical model of the respiratory system, which takes into account the physiology of respiratory motion to simulate the real non-reproducible motion. The behavior of the lungs, is directly driven by the simulated actions of the breathing muscles, i.e. the diaphragm and the intercostal muscles (the rib cage). In this paper, the lung model is monitored and controlled by a personalized lung pressure/volume relationship during a whole respiratory cycle. The lung pressure and rib kinematics are patient specific and obtained by surrogate measurement. The rib displacement corresponding to the transformation which was computed by finite helical axis method from the end of exhalation (EE) to the end of inhalation (EI). The lung pressure is calculated by an optimization framework based on inverse finite element analysis, by minimizing the lung volume errors, between the respiratory volume (respiratory airflow exchange) and the simulated volume (calculated by biomechanical simulation). We have evaluated the model accuracy on five public datasets. We have also evaluated the lung tumor motion identified in 4D CT scan images and compared it with the trajectory that was obtained by finite element simulation. The effects of rib kinematics on lung tumor trajectory were investigated. Over all phases of respiration, our developed model is able to predict the lung tumor motion with an average landmark error of [Formula: see text]. The results demonstrate the effectiveness of our physics-based model. We believe that this model can be potentially used in 4D dose computation, removal of breathing motion artifacts in positron emission tomography (PET) or gamma prompt image reconstruction.

Integrating CVH and LVH metrics into an optimization strategy for the selection of Iris collimator for Cyberknife Xsight lung tracking treatment

PURPOSE

We conducted this study to construct a target coverage-volume histogram (CVH) and leakage-volume histogram (LVH) metrics and optimization strategy for the selection of the Iris collimator in Cyberknife Xsight lung tracking treatment through a retrospective analysis of target structures and clinical data.

METHODS AND MATERIALS

CVH and LVH metrics were retrospectively analyzed for 37 lung cancer patients. CVH and LVH were the same as dose-volume histogram (DVH), but with a coverage and leakage replacing dose. For each patient, Iris collimator was optimized and selected based on CVH and LVH metrics. The CVH and LVH metrics were then compared to ascertain differences in 95% (C95) or 90% (C90) of the target coverage thresholds. The planning target volume (PTV) C95 and C90 coverage, absolute mean leakage value, leakage/coverage ratio, selected collimator diameter (Φ), Φ/length of the long axis of PTV (A_max ), and Φ/length of the short axis (A_min ) of PTV were compared. The correlation of the absolute mean leakage value, leakage/coverage ratio, Φ/A_min and Φ/A_max were evaluated.

RESULTS

For each patient, the PTV C95 coverage (70.45 vs 63.19) and C90 coverage (77.25 vs 69.96) were higher in the C95 coverage threshold group compared to the C90 coverage threshold group. The leakage/coverage ratio (0.56 vs 0.69) and absolute mean leakage value (0.56 vs 0.61) were lower in C90 coverage threshold group than in C95 coverage threshold group. The Spearmen correlation test showed the Φ/A_min were significantly correlated with leakage/coverage ratio and absolute mean leakage value. Upon analysis of the selected collimator diameters, the mean value of Φ/A_min of the optimized collimator diameters was found to be 1.10.

CONCLUSION

The CVH and LVH analysis is able to quantitatively evaluate the tradeoff between target coverage and normal tissue sparing.

Deep learning-based real-time volumetric imaging for lung stereotactic body radiation therapy: a proof of concept study

Due to the inter- and intra- variation of respiratory motion, it is highly desired to provide real-time volumetric images during the treatment delivery of lung stereotactic body radiation therapy (SBRT) for accurate and active motion management. In this proof-of-concept study, we propose a novel generative adversarial network integrated with perceptual supervision to derive instantaneous volumetric images from a single 2D projection. Our proposed network, named TransNet, consists of three modules, i.e. encoding, transformation and decoding modules. Rather than only using image distance loss between the generated 3D images and the ground truth 3D CT images to supervise the network, perceptual loss in feature space is integrated into loss function to force the TransNet to yield accurate lung boundary. Adversarial supervision is also used to improve the realism of generated 3D images. We conducted a simulation study on 20 patient cases, who had received lung SBRT treatments in our institution and undergone 4D-CT simulation, and evaluated the efficacy and robustness of our method for four different projection angles, i.e. 0°, 30°, 60° and 90°. For each 3D CT image set of a breathing phase, we simulated its 2D projections at these angles. For each projection angle, a patient's 3D CT images of 9 phases and the corresponding 2D projection data were used to train our network for that specific patient, with the remaining phase used for testing. The mean absolute error of the 3D images obtained by our method are 99.3 ± 14.1 HU. The peak signal-to-noise ratio and structural similarity index metric within the tumor region of interest are 15.4 ± 2.5 dB and 0.839 ± 0.090, respectively. The center of mass distance between the manual tumor contours on the 3D images obtained by our method and the manual tumor contours on the corresponding 3D phase CT images are within 2.6 mm, with a mean value of 1.26 mm averaged over all the cases. Our method has also been validated in a simulated challenging scenario with increased respiratory motion amplitude and tumor shrinkage, and achieved acceptable results. Our experimental results demonstrate the feasibility and efficacy of our 2D-to-3D method for lung cancer patients, which provides a potential solution for in-treatment real-time on-board volumetric imaging for tumor tracking and dose delivery verification to ensure the effectiveness of lung SBRT treatment.

Evaluation of OAR dose sparing and plan robustness of beam-specific PTV in lung cancer IMRT treatment

PURPOSE

Margins are employed in radiotherapy treatment planning to mitigate the dosimetric effects of geometric uncertainties for the clinical target volume (CTV). Here, we proposed a margin concept that takes into consideration the beam direction, thereby generating a beam-specific planning target volume (BSPTV) on a beam entrance view. The total merged BSPTV was considered a target for optimization. We investigated the impact of this novel approach for lung intensity-modulated radiotherapy (IMRT) treatment, and compared the treatment plans generated using BSPTV with general PTV.

METHODS AND MATERIALS

We generated the BSPTV by expanding the CTV perpendicularly to the incident beam direction using the 2D version of van Herk's margin concept. The BSPTV and general PTV margin were analyzed using digital phantom simulation. Fifteen lung cancer patients were used in the planning study. First, all patient targets were performed with the CTV projection area analysis to select the suitable beam angles. Then, BSPTV was generated according to the selected beam angles. IMRT plans were optimized with the general PTV and BSPTV as the target volumes, respectively. The dosimetry metrics were calculated and evaluated between these two plans. The plan robustness of both plans for setup uncertainties was evaluated using worst-case analysis.

RESULTS

Both general PTV and BSPTV plans satisfied the CTV coverage. In addition, the BSPTV plans improved the sparing of high doses to target-surrounding lung tissues compared to the general PTV plans. Both D_mean of Ring PTV and Ring BSPTV were significantly lower in BSPTV plans (38.89 Gy and 39.43 Gy) compared to the general PTV plans (40.27 Gy and 40.68 Gy). The V20, V5, and mean lung dose of the affected lung were significant lower in BSPTV plans (16.20%, 28.75% and 8.93 Gy) compared to general PTV plans (16.69%, 29.22% and 9.18 Gy). In uncertainty scenarios, about 80% of target coverage was achieved for both general PTV and BSPTV plans.

CONCLUSIONS

The results suggested that plan robustness can be guaranteed in both the BSPTV and general PTV plans. However, the BSPTV plan spared normal tissues, such as the lungs, significantly better compared to the general PTV plans.

The Reintroduction of Radiotherapy Into the Integrated Management of Kidney Cancer

The incidence of renal cell carcinoma (RCC) has been increasing, with a moderate subgroup of individuals who later develop metastatic disease. Historically, metastatic RCC has been managed with systemic therapy because RCC was believed to be radioresistant. Local therapies, such as stereotactic body radiation therapy, also known as stereotactic ablative radiotherapy, which utilize focused high-dose-rate radiation delivered over a limited number of treatments, have been successful in controlling local disease and, in some cases, extending survival in patients with intracranial and extracranial metastatic RCC. Stereotactic ablative radiotherapy is highly effective in treating intact disease when patients are not surgical candidates. Stereotactic ablative radiotherapy is well tolerated when used in conjunction with systemic therapy such as tyrosine kinase inhibitors and immune checkpoint inhibitors. These successes have prompted investigators to evaluate the efficacy of stereotactic body radiation therapy in novel settings such as neoadjuvant treatment of advanced RCC with tumor thrombus and oligometastatic/oligoprogressive disease states.

Optimized CyberKnife Lung Treatment: Effect of Fractionated Tracking Volume Change on Tracking Results

Objectives

To explore the impact of volume change in the fractionated tracking of stereotactic radiotherapy on the results of synchronous, respiratory tracking algorithm using CyberKnife.

Methods

A total of 38 lung tumor patients receiving stereotactic radiotherapy at our center from March 2018 to October 2019 were counted. Photoshop CS4 image processing software was used to obtain the pixels and the average value of brightness of the tracking volume in the image and calculate the grayscale within the contour of the tracking volume on the real-time X-ray image. At the same time, parameters of the synchronous respiratory tracking algorithm of the fractional CyberKnife were extracted for comparison between the volume of image-guided image tracking and the number of fractions during stereotactic radiotherapy. We also analyzed the relationship between fraction tumor location and characteristics and the calculated results of synchronous respiratory tracking by CyberKnife.

Results

There were no significant differences between the first four fractions (p > 0.05) for left lung lesions and no significant differences between the first five fractions for right lung lesions (p ≥ 0.05). For peripheral lung cancer, longer fractional treatment led to greater variation in grayscale (G-A: >4 fractions p < 0.05), while for central lung cancer, longer fractional treatment led to greater variation in parameters of the synchronous respiratory tracking algorithm (Uncertainty A and Uncertainty B: >4 fractions p < 0.05). There was a significant correlation between radiotherapy-graded tumor density and relevant parameters, and the correlation was strong (>0.7, p < 0.05).

Conclusion

With the increase of treatment fractions, the gray value in the patient tracking volume decreased. Patients of >4 fractions were advised to reevaluate with simulated CT and replan. For tumors with small diameter and low density, the imaging changes of volume should be closely followed during treatment. For left lung and central lung cancer, carefully select the synchronous tracking treatment with 2-view.

Real-time markerless tumour tracking with patient-specific deep learning using a personalised data generation strategy: proof of concept by phantom study

OBJECTIVE

For real-time markerless tumour tracking in stereotactic lung radiotherapy, we propose a different approach which uses patient-specific deep learning (DL) using a personalised data generation strategy, avoiding the need for collection of a large patient data set. We validated our strategy with digital phantom simulation and epoxy phantom studies.

METHODS

We developed lung tumour tracking for radiotherapy using a convolutional neural network trained for each phantom's lesion by using multiple digitally reconstructed radiographs (DRRs) generated from each phantom's treatment planning four-dimensional CT. We trained tumour-bone differentiation using large numbers of training DRRs generated with various projection geometries to simulate tumour motion. We solved the problem of using DRRs for training and X-ray images for tracking using the training DRRs with random contrast transformation and random noise addition.

RESULTS

We defined adequate tracking accuracy as the percentage frames satisfying <1 mm tracking error of the isocentre. In the simulation study, we achieved 100% tracking accuracy in 3 cm spherical and 1.5×2.25×3 cm ovoid masses. In the phantom study, we achieved 100 and 94.7% tracking accuracy in 3 cm and 2 cm spherical masses, respectively. This required 32.5 ms/frame (30.8 fps) real-time processing.

CONCLUSIONS

We proved the potential feasibility of a real-time markerless tumour tracking framework for stereotactic lung radiotherapy based on patient-specific DL with personalised data generation with digital phantom and epoxy phantom studies.

ADVANCES IN KNOWLEDGE

Using DL with personalised data generation is an efficient strategy for real-time lung tumour tracking.

Multiple-CT optimization: An adaptive optimization method to account for anatomical changes in intensity-modulated proton therapy for head and neck cancers

PURPOSE

We aimed to determine whether multiple-CT (MCT) optimization of intensity-modulated proton therapy (IMPT) could improve plan robustness to anatomical changes and therefore reduce the additional need for adaptive planning.

METHODS AND MATERIALS

Ten patients with head and neck cancer who underwent IMPT were included in this retrospective study. Each patient had primary planning CT (PCT), a first adaptive planning CT (ACT1), and a second adaptive planning CT (ACT2). Selective robust IMPT plans were generated using each CT data set (PCT, ACT1, and ACT2). Moreover, a MCT optimized plan was generated using the PCT and ACT1 data sets together. Dose distributions optimized using each of the four plans (PCT, ACT1, ACT2, and MCT plans) were re-calculated on ACT2 data. The doses to the target and to organs at risk were compared between optimization strategies.

RESULTS

MCT plans for all patients met all target dose and organs-at-risk criteria for all three CT data sets. Target dose and organs-at-risk dose for PCT and ACT1 plans re-calculated on ACT2 data set were compromised, indicating the need for adaptive planning on ACT2 if PCT or ACT1 plans were used. The D98% of CTV1 and CTV3 of MCT plan re-calculated on ACT2 were both above the coverage criteria. The CTV2 coverage of the MCT plan re-calculated on ACT2 was worse than ACT2 plan. The MCT plan re-calculated on ACT2 data set had lower chiasm, esophagus, and larynx doses than did PCT, ACT1, or ACT2 plans re-calculated on ACT2 data set.

CONCLUSIONS

MCT optimization can improve plan robustness toward anatomical change and may reduce the number of plan adaptation for head and neck cancers.

Dosimetric Effect of Intrafraction Tumor Motion in Lung Stereotactic Body Radiotherapy Using CyberKnife Static Tracking System

Purpose:

We investigated the dosimetric effect of intrafraction tumor motion in lung stereotactic body radiotherapy using the CyberKnife static tracking system.

Methods:

Four-dimensional computed tomography scans of a dynamic thorax phantom were acquired. Two motion ranges, 3 collimator sizes, and 4 treatment starting phases were investigated. Monte Carlo dose distributions were calculated on internal target volume with a treatment-specific setup margin for 6 Gy/1 fraction. Dosimetric effects of intrafractional tumor motion were assessed with Gafchromic films. γ (5%/3 mm), dose differences, and distance to agreement were analyzed.

Results:

With 30 mm collimator plans, the measured dose passed the criteria γ (5%/3 mm) in all tumor motion ranges. The γ passing rates of the plans using 20 mm or 20+35 mm collimators were much lower than that with 30 mm collimator, especially with the 30 mm tumor motion range. The measured dose of 10 mm tumor motion ranges all passed the 90% criteria of γ (5%/3 mm), the results being much better than those of 30 mm tumor motion ranges, which were below 80%. The results of same delivered plan but treated with different starting phases varies greatly.

Conclusion:

Xsight Lung Tracking technique should be used with caution in lung cancer stereotactic body radiation therapy because the temporal dose variations can be significant.

Variation in target volume and centroid position due to breath holding during four-dimensional computed tomography scanning: A phantom study

This study investigated the effects of respiratory motion, including unwanted breath holding, on the target volume and centroid position on four‐dimensional computed tomography (4DCT) imaging. Cine 4DCT images were reconstructed based on a time‐based sorting algorithm, and helical 4DCT images were reconstructed based on both the time‐based sorting algorithm and an amplitude‐based sorting algorithm. A spherical object 20 mm in diameter was moved according to several simulated respiratory motions, with a motion period of 4.0 s and maximum amplitude of 5 mm. The object was extracted automatically, and the target volume and centroid position in the craniocaudal direction were measured using a treatment planning system. When the respiratory motion included unwanted breath‐holding times shorter than the breathing cycle, the root mean square errors (RSME) between the reference and imaged target volumes were 18.8%, 14.0%, and 5.5% in time‐based images in cine mode, time‐based images in helical mode, and amplitude‐based images in helical mode, respectively. In helical mode, the RSME between the reference and imaged centroid position was reduced from 1.42 to 0.50 mm by changing the reconstruction method from time‐ to amplitude‐based sorting. When the respiratory motion included unwanted breath‐holding times equal to the breathing cycle, the RSME between the reference and imaged target volumes were 19.1%, 24.3%, and 15.6% in time‐based images in cine mode, time‐based images in helical mode, and amplitude‐based images in helical mode, respectively. In helical mode, the RSME between the reference and imaged centroid position was reduced from 1.61 to 0.83 mm by changing the reconstruction method from time‐ to amplitude‐based sorting. With respiratory motion including breath holding of shorter duration than the breathing cycle, the accuracies of the target volume and centroid position were improved by amplitude‐based sorting, particularly in helical 4DCT.

See, Think, and Act: Real-Time Adaptive Radiotherapy

The world is embracing the information age, with real-time data at hand to assist with many decisions. Similarly, in cancer radiotherapy we are inexorably moving toward using information in a smarter and faster fashion, to usher in the age of real-time adaptive radiotherapy. The three critical steps of real-time adaptive radiotherapy, aligned with driverless vehicle technology are a continuous see, think, and act loop. See: use imaging systems to probe the patient anatomy or physiology as it evolves with time. Think: use current and prior information to optimize the treatment using the available adaptive degrees of freedom. Act: deliver the real-time adapted treatment. This paper expands upon these three critical steps for real-time adaptive radiotherapy, provides a historical context, reviews the clinical rationale, and gives a future outlook for real-time adaptive radiotherapy.

Localization accuracy of robotic radiosurgery in 1-view tracking

PURPOSE

When a lung lesion is detected by only one couple of X-ray tube and image detector integrated with CyberKnife®, the fiducial-less tracking is limited to 1-view (34% of lung treatments at Centro Diagnostico Italiano). The aim of the study was mainly to determine the margin needed to take into account the localization uncertainty along the blind view (out-of-plane direction).

METHODS

36 patients treated in 2-view tracking modality (127 fractions in total) were included in the study. The actual tumor positions were determined retrospectively through logfile analysis and were projected onto 2D image planes. In the same plots the planned target positions based on biphasic breath-hold CT scans were represented preserving the metric with respect to the imaging center. The internal margin necessary to cover in out-of-plane direction the 95% of the target position distribution in the 95% of cases was calculated by home-made software in Matlab®. A validation test was preliminarily performed using XLT Phantom (CIRS) both in 2-view and 1-view scenarios.

RESULTS

The validation test proved the reliability of the method, in spite of some intrinsic limitations. Margins were estimated equal to 5 and 6 mm for targets in upper and lower lobe respectively. Biphasic breath-hold CT led to underestimate the target movement in the hypothetical out-of-plane direction. The inter-fractional variability of spine-target distance was an important source of uncertainty for 1-view treatments.

CONCLUSION

This graphic comparison method preserving metric could be employed in the clinical workflow of 1-view treatments to get patient-related information for customized margin definition.

Commissioning of a fluoroscopic-based real-time markerless tumor tracking system in a superconducting rotating gantry for carbon-ion pencil beam scanning treatment

PURPOSE

To perform the final quality assurance of our fluoroscopic-based markerless tumor tracking for gated carbon-ion pencil beam scanning (C-PBS) radiotherapy using a rotating gantry system, we evaluated the geometrical accuracy and tumor tracking accuracy using a moving chest phantom with simulated respiration.

METHODS

The positions of the dynamic flat panel detector (DFPD) and x-ray tube are subject to changes due to gantry sag. To compensate for this, we generated a geometrical calibration table (gantry flex map) in 15° gantry angle steps by the bundle adjustment method. We evaluated five metrics: (a) Geometrical calibration was evaluated by calculating chest phantom positional error using 2D/3D registration software for each 5° step of the gantry angle. (b) Moving phantom displacement accuracy was measured (±10 mm in 1-mm steps) with a laser sensor. (c) Tracking accuracy was evaluated with machine learning (ML) and multi-template matching (MTM) algorithms, which used fluoroscopic images and digitally reconstructed radiographic (DRR) images as training data. The chest phantom was continuously moved ±10 mm in a sinusoidal path with a moving cycle of 4 s and respiration was simulated with ±5 mm expansion/contraction with a cycle of 2 s. This was performed with the gantry angle set at 0°, 45°, 120°, and 240°. (d) Four types of interlock function were evaluated: tumor velocity, DFPD image brightness variation, tracking anomaly detection, and tracking positional inconsistency in between the two corresponding rays. (e) Gate on/off latency, gating control system latency, and beam irradiation latency were measured using a laser sensor and an oscilloscope.

RESULTS

By applying the gantry flex map, phantom positional accuracy was improved from 1.03 mm/0.33° to <0.45 mm/0.27° for all gantry angles. The moving phantom displacement error was 0.1 mm. Due to long computation time, the tracking accuracy achieved with ML was <0.49 mm (=95% confidence interval [CI]) for imaging rates of 15 and 7.5 fps; those at 30 fps were decreased to 1.84 mm (95% CI: 1.79 mm-1.92 mm). The tracking positional accuracy with MTM was <0.52 mm (=95% CI) for all gantry angles and imaging frame rates. The tumor velocity interlock signal delay time was 44.7 ms (=1.3 frame). DFPD image brightness interlock latency was 34 ms (=1.0 frame). The tracking positional error was improved from 2.27 ± 2.67 mm to 0.25 ± 0.24 mm by the tracking anomaly detection interlock function. Tracking positional inconsistency interlock signal was output within 5.0 ms. The gate on/off latency was <82.7 ± 7.6 ms. The gating control system latency was <3.1 ± 1.0 ms. The beam irradiation latency was <8.7 ± 1.2 ms.

CONCLUSIONS

Our markerless tracking system is now ready for clinical use. We hope to shorten the computation time needed by the ML algorithm at 30 fps in the future.

Effect of setup and inter-fraction anatomical changes on the accumulated dose in CT-guided breath-hold intensity modulated proton therapy of liver malignancies

PURPOSE

To evaluate the effect of setup uncertainties including uncertainties between different breath holds (BH) and inter-fractional anatomical changes under CT-guided BH with intensity-modulated proton therapy (IMPT) in patients with liver cancer.

METHODS AND MATERIALS

This retrospective study considered 17 patients with liver tumors who underwent feedback-guided BH (FGBH) IMRT treatment with daily CT-on-rail imaging. Planning CT images were acquired at simulation using FGBH, and FGBH CT-on-rail images were also acquired prior to each treatment. Selective robust IMPT plans were generated using planning CT and re-calculated on each daily CT-on-rail image. Subsequently, the fractional doses were deformed and accumulated onto the planning CT according to the deformable image registration between daily and planning CTs. The doses to the target and organs at risk (OARs) were compared between IMRT, planned IMPT, and accumulated IMPT doses.

RESULTS

For IMPT plans, the mean of D_98% of CTV for all 17 patients was slightly reduced from the planned dose of 68.90 ± 1.61 Gy to 66.48 ± 1.67 Gy for the accumulated dose. The target coverage could be further improved by adjusting planning techniques. The dose-volume histograms of both planned and accumulated IMPT doses showed better sparing of OARs than that of the IMRT.

CONCLUSIONS

IMPT with FGBH and CT-on-rail guidance is a robust treatment approach for liver tumor cases.