The application of the sinusoidal model to lung cancer patient respiratory motion

The impact of audiovisual breathing guidance on respiratory-triggered cardiac magnetic resonance cine imaging

BACKGROUND

In patients who have difficulty holding their breath, a free breathing (FB) respiratory-triggered (RT) bSSFP cine technique may be used. However, this technique may have inferior image quality and a longer scan time than breath-hold (BH) bSSFP cine acquisitions. This study examined the effect of an audiovisual breathing guidance (BG) system on RT bSSFP cine image quality, scan time, and ventricular measurements.

METHODS

This study evaluated a BG system that provides audiovisual instructions and feedback on the timing of inspiration and expiration to the patient during image acquisition using input from the respiratory bellows to guide them toward a regular breathing pattern with extended end-expiration. In this single-center prospective study in patients undergoing a clinical cardiac magnetic resonance examination, a ventricular short-axis stack of bSSFP cine images was acquired using 3 techniques in each patient: 1) FB and RT (FBRT), 2) BG system and RT (BGRT), and 3) BH. The 3 acquisitions were compared for image quality metrics (endocardial edge definition, motion artifact, and blood-to-myocardial contrast) scored on a Likert scale, scan time, and ventricular volumes and mass.

RESULTS

Thirty-two patients (19 females; median age 21 years, IQR 18-32) completed the study protocol. For scan time, BGRT was faster than FBRT (163 s vs. 345 s, p < 0.001). Endocardial edge definition, motion artifact, and blood-to-myocardial contrast were all better for BGRT than FBRT (p < 0.001). Left ventricular (LV) end-systolic volume (ESV) was smaller (3%, p = 0.02) and LV ejection fraction (EF) was larger (0.5%, p = 0.003) with BGRT than with FBRT. There was no significant difference in LV end-diastolic volume (EDV), LV mass, right ventricular (RV) EDV, RV ESV, and RV EF. Scan times were shorter for BGRT compared to BH. Endocardial edge definition and blood-to-myocardial contrast were better for BH than BGRT. Compared to BH, the LV EDV, LV ESV, RV EDV, and RV ESV were mildly smaller (all differences <7%) for BGRT.

CONCLUSIONS

The addition of a BG system to RT bSSFP cine acquisitions decreased the scan time and improved image quality. Further exploration of this BG approach is warranted in more diverse populations and with other free breathing sequences.

Effectiveness of visual biofeedback-guided respiratory-correlated 4D-MRI for radiotherapy guidance on the MR-linac

PURPOSE

Respiratory-correlated 4D-MRI may provide motion characteristics for radiotherapy but is susceptible to irregular breathing. This study investigated the effectiveness of visual biofeedback (VBF) guidance for breathing regularization during 4D-MRI acquisitions on an MR-linac.

METHODS

A simultaneous multislice-accelerated 4D-MRI sequence was interleaved with a one-dimensional respiratory navigator (1D-RNAV) in 10 healthy volunteers on a 1.5T Unity MR-linac (Elekta AB, Stockholm, Sweden). Volunteer-specific breathing amplitudes and periods were derived from the 1D-RNAV signal obtained during unguided 4D-MRI acquisitions. These were used for the guidance waveform, while the 1D-RNAV positions were overlayed as VBF. VBF effectiveness was quantified by calculating the change in coefficient of variation (CV diff $$ {\mathrm{CV}}^{\mathrm{diff}} $$ ) for the breathing amplitude and period, the position SD of end-exhale, end-inhale and midposition locations, and the agreement between the 1D-RNAV signals and guidance waveforms. The 4D-MRI quality was assessed by quantifying amounts of missing data.

RESULTS

VBF had an average latency of 520 ± 2 ms. VBF reduced median breathing variations by 18% to 35% (amplitude) and 29% to 57% (period). Median position SD reductions ranged from -3% to 35% (end-exhale), 29% to 38% (end-inhale), and 25% to 37% (midposition). Average differences between guidance waveforms and 1D-RNAV signals were 0.0 s (period) and +1.7 mm (amplitude). VBF also decreased the median amount of missing data by 11% and 29%.

CONCLUSION

A VBF system was successfully implemented, and all volunteers were able to adapt to the guidance waveform. VBF during 4D-MRI acquisitions drastically reduced breathing variability but had limited effect on missing data in respiratory-correlated 4D-MRI.

A novel external/internal tumor tracking approach to compensate for respiratory motion baseline drifts

Objective.Real-time respiratory tumor tracking as implemented in a robotic treatment unit is based on continuous optical measurement of the position of external markers and a correlation model between them and internal target positions, which are established with X-ray imaging of the tumor, or fiducials placed in or around the tumor. Correlation models are created with fifteen simultaneously measured external/internal marker position pairs divided over the respiratory cycle. Every 45-150 s, the correlation model is updated by replacing the three first acquired data pairs with three new pairs. Tracking simulations for >120.000 computer-generated respiratory tracks demonstrated that this tracking approach resulted in relevant inaccuracies in internal target position predictions, especially in case of presence of respiratory motion baseline drifts.Approach.To better cope with drifts, we introduced a novel correlation model with an explicit time dependence, and we proposed to replace the currently applied linear-motion tracking (LMT) by mixed-model tracking (MMT). In MMT, the linear correlation model is extended with an explicit time dependence in case of a detected baseline drift. MMT prediction accuracies were then established for the same >120.000 computer-generated patients as used for LMT.Main results.For 150 s update intervals, MMT outperformed LMT in internal target position prediction accuracy for 93.7 ∣ 97.2% of patients with 0.25 ∣ 0.5 mm min^-1linear respiratory motion baseline drifts with similar numbers of X-ray images and similar treatment times. For the upper 25% of patients, mean 3D internal target position prediction errors reduced by 0.7 ∣ 1.8 mm, while near maximum reductions (upper 10% of patients) were 0.9 ∣ 2.0 mm.Significance.For equal numbers of acquired X-ray images, MMT greatly improved tracking accuracy compared to LMT, especially in the presence of baseline drifts. Even with almost 50% less acquired X-ray images, MMT still outperformed LMT in internal target position prediction accuracy.

Results from the AAPM Task Group 324 respiratory motion management in radiation oncology survey

PURPOSE

To quantify the clinical practice of respiratory motion management in radiation oncology.

METHODS

A respiratory motion management survey was designed and conducted based on clinician survey guidelines. The survey was administered to American Association of Physicists in Medicine (AAPM) members on 17 August 2020 and closed on 13 September 2020.

RESULTS

A total of 527 respondents completed the entire survey and 651 respondents completed part of the survey, with the partially completed surveys included in the analysis. Overall, 84% of survey respondents used deep inspiration breath hold for left-sided breast cancer. Overall, 83% of respondents perceived respiratory motion management for thoracic and abdominal cancer radiotherapy patients to be either very important or required. Overall, 95% of respondents used respiratory motion management for thoracic and abdominal sites, with 36% of respondents using respiratory motion management for at least 90% of thoracic and abdominal patients. The majority (60%) of respondents used the internal target volume method to treat thoracic and abdominal cancer patients, with 25% using breath hold or abdominal compression and 13% using gating or tracking.

CONCLUSIONS

A respiratory motion management survey has been completed by AAPM members. Respiratory motion management is generally considered very important or required and is widely used for breast, thoracic, and abdominal cancer treatments.

Dosimetric impact of phase shifts on Radixact Synchrony tracking system with patient-specific breathing patterns

Purpose

The Synchrony tracking system of Radixact is capable of real‐time tumor tracking by building a correlation model between external light‐emitting diodes on the patient's chest and an internal marker. A phase shift between the chest wall and a lung tumor has been reported. Hence, this study focused on evaluating the accuracy of the tracking system, especially under a patient‐specific breathing pattern with respiratory phase shifts.

Methods

A phantom containing fiducial markers was placed on a moving platform. The intrinsic delivery accuracy was verified with a patient‐specific breathing pattern. Three patient‐specific breathing patterns were then implemented, for which phase shifts, φ, were introduced. Phase shifts with +0.3 s and +1 s were tested for dosimetric aspects, whereas ±0.3, ±0.6, and ±0.8 s shifts were used for tracking accuracy. The resultant dose distributions were analyzed by γ comparison. Dose profiles in the superior‐inferior and lateral directions were compared. Logfiles of the tracking information were extracted from the system and compared with the input breathing pattern. The root mean square (RMS) difference was used to quantify the consistency.

Results

When the φ value was as large as 1 s, a severe inconsistency was observed. The target was significantly underdosed, down to 89% of the originally planned dose. γ analysis revealed that the failed portion was concentrated in the target region. The RMS of the tracking difference was close to 1 mm when φ was ±0.3 s and approximately 4 mm when φ was ±0.8 s. Tracking errors increased with an increase in the degree of phase shifts.

Conclusion

Phase shifts between the patient chest wall and the internal target may hamper treatment delivery and jeopardize treatment using Synchrony Tracking. Hence, a larger planning target volume (PTV) may be necessary if a large phase shift is observed in a patient, especially when an external surrogate shows a lag in motion when compared with the tumor.

Correlation of Optical Surface Respiratory Motion Signal and Internal Lung and Liver Tumor Motion: A Retrospective Single-Center Observational Study

Purpose: Surface-guided radiation therapy (SGRT) application has limitations. This study aimed to explore the relationship between patient characteristics and their external/internal correlation to qualitatively assess the external/internal correlation in a particular patient. Methods: Liver and lung cancer patients treated with radiotherapy in our institution were retrospectively analyzed. The external/internal correlation were calculated with Spearman correlation coefficient (SCC) and SCC after support vector regression (SVR) fitting (SCC_svr). The relationship between the external/internal correlation and magnitudes of motion of the tumor and external marker (A_i, A_e), tumor volume V_t, patient age, gender, and tumor location were explored. Results: The external/internal motions of liver and lung cancer patients were strongly correlated in the S-I direction, with mean SCC_svr values of 0.913 and 0.813. The correlation coefficients between the external/internal correlations and the patients’ characteristics (A_i, A_e, V_t, and age) were all smaller than 0.5; A_i, A_e and liver tumor volumes were positively correlated with the strength of the external/internal correlation, while lung tumor volumes and patient age were negative. The external/internal correlations in males and females were roughly equal, and the external/internal correlations in patients with peripheral lung cancers were stronger than those in patients with central lung cancers. Conclusion: The external/internal correlation shows great individual differences. The effects of A_i, A_e, V_t, and age are weakly to moderately correlated. Our results suggest the necessity of individualized assessment of patient's external/internal motion correlation prior to the application of SGRT technique for breath motion monitoring.

How should we model and evaluate breathing interplay effects in IMPT?

Breathing interplay effects in Intensity Modulated Proton Therapy (IMPT) arise from the interaction between target motion and the scanning beam. Assessing the detrimental effect of interplay and the clinical robustness of several mitigation techniques requires statistical evaluation procedures that take into account the variability of breathing during dose delivery. In this study, we present such a statistical method to model intra-fraction respiratory motion based on breathing signals and assess clinical relevant aspects related to the practical evaluation of interplay in IMPT such as how to model irregular breathing, how small breathing changes affect the final dose distribution, and what is the statistical power (number of different scenarios) required for trustworthy quantification of interplay effects. First, two data-driven methodologies to generate artificial patient-specific breathing signals are compared: a simple sinusoidal model, and a precise probabilistic deep learning model generating very realistic samples of patient breathing. Second, we investigate the highly fluctuating relationship between interplay doses and breathing parameters, showing that small changes in breathing period result in large local variations in the dose. Our results indicate that using a limited number of samples to calculate interplay statistics introduces a bigger error than using simple sinusoidal models based on patient parameters or disregarding breathing hysteresis during the evaluation. We illustrate the power of the presented statistical method by analyzing interplay robustness of 4DCT and Internal Target Volume (ITV) treatment plans for a 8 lung cancer patients, showing that, unlike 4DCT plans, even 33 fraction ITV plans systematically fail to fulfill robustness requirements.

A semi-supervised autoencoder framework for joint generation and classification of breathing

BACKGROUND AND OBJECTIVE

One of the main problems with biomedical signals is the limited amount of patient-specific data and the significant amount of time needed to record the sufficient number of samples needed for diagnostic and treatment purposes. In this study, we present a framework to simultaneously generate and classify biomedical time series based on a modified Adversarial Autoencoder (AAE) algorithm and one-dimensional convolutions. Our work is based on breathing time series, with specific motivation to capture breathing motion during radiotherapy lung cancer treatments.

METHODS

First, we explore the potential in using the Variational Autoencoder (VAE) and AAE algorithms to model breathing signals from individual patients. We then extend the AAE algorithm to allow joint semi-supervised classification and generation of different types of signals within a single framework. To simplify the modeling task, we introduce a pre-processing and post-processing compressing algorithm that transforms the multi-dimensional time series into vectors containing time and position values, which are transformed back into time series through an additional neural network.

RESULTS

The resulting models are able to generate realistic and varied samples of breathing. By incorporating 4% and 12% of the labeled samples during training, our model outperforms other purely discriminative networks in classifying breathing baseline shift irregularities from a dataset completely different from the training set, achieving an average macro F1-score of 94.91% and 96.54%, respectively.

CONCLUSION

To our knowledge, the presented framework is the first approach that unifies generation and classification within a single model for this type of biomedical data, enabling both computer aided diagnosis and augmentation of labeled samples within a single framework.

[Extraction of respiratory signals from chest tomosynthesis]

OBJECTIVE

To analyze the respiratory motion of the scanned object during acquisition of digital chest tomosynthesis (CTS) using a linear model.

OBJECTIVE

Respiratory signals were generated by extracting the motion of the diaphragm from the projection radiographs. The diaphragm trajectory obtained by dynamic programming (DP) was modeled and fitted, and according to the fitting of the data, the base motion curve and respiratory signal curve of the diaphragm were separated. Multipurpose chest phantom data, simulated digital Xcat phantom data and the datasets of 3 clinical patients were used to validate the performance of the proposed method.

OBJECTIVE

The motion trajectory of the diaphragm extracted from multipurpose chest phantom simulation data was linear. The respiratory signals could be effectively extracted from the 3 datasets of clinical patients in different respiratory states. The correlation coefficient between the respiratory signal extracted in Xcat simulation experiment and the original design was 0.9797.

OBJECTIVE

The linear model can effectively obtain the respiratory motion information of patients in real time, thus enabling the physicians to make clinical decisions on a rescan.

Application of discrete cosine transform to assess the effect of tumor motion variations on the definition of ITV in lung and liver SBRT

PURPOSE

To use Discrete Cosine Transform to include tumor motion variations on ITV definition of SBRT patients.

METHODS

Data from 66 patients was collected. 2D planar fluoroscopy images (FI) were available for 54 patients. Daily CBCT projections (CBCTp) from 29 patients were employed to measure interfraction amplitude variability. Systematic amplitude variations were obtained from 17 patients with data from both FI and CBCTp. Tumor motion curves obtained from FI were characterized with a Cosine model (CM), based on cosine functions to the power of 2, 4 or 6, and DCT. Performance of both models was evaluated by means of R² coefficient and by comparing their results on Internal Target Volume (ITV) margins against those calculated from original tumor motion curves. Amplitude variations from CBCTp, as well as estimations of baseline shift variations were added to the DCT model to account for their effect on ITV margins.

RESULTS

DCT replicated tumor motion curves with a mean R² values for all patients of 0.86, 0.91 and 0.96 for the lateral (LAT), anterior-posterior (AP) and cranio-caudal (CC) directions respectively. CM yielded worst results, with R² values of 0.64, 0.61 and 0.74 in the three directions. Interfraction amplitude variation increased ITV margins by a 9%, while baseline shift variability implied a 40% and 80-100% increase for normalized values of baseline shift of 0.2 and 0.4 respectively.

CONCLUSIONS

Probability distribution functions of tumor positions can be successfully characterized with DCT. This permits to include tumor motion variablilities obtained from patient population into patient specific ITVs.

A hierarchical model of abdominal configuration changes extracted from golden angle radial magnetic resonance imaging

Abdominal organs are subject to a variety of physiological forces that superimpose their effects to influence local motion and configuration. These forces not only include breathing, but can also arise from cyclic antral contractions and a range of slow configuration changes. To elucidate each individual motion pattern as well as their combined effects, a hierarchical motion model was built for characterization of these 3 motion modes (characterized as deformation maps between states) using golden angle radial MR signals. Breathing motions are characterized first. Antral contraction states are then reconstructed after breathing motion-induced deformation are corrected; slow configuration change states are further extracted from breathing motion-corrected image reconstructions. The hierarchical model is established based on these multimodal states, which can be either individually shown or combined to demonstrate any arbitrary composited motion patterns. The model was evaluated using 20 MR scans acquired from 9 subjects. Poor reproducibility of breathing motions both within as well as between scan sessions was observed, with an average intra-subject difference of 1.6 cycles min-1 for average breathing frequencies of 12.0 cycles min-1. Antral contraction frequency distributions were more stable than breathing, but also presented poor reproducibility between scans with an average difference of 0.3 cycles min-1 for average frequencies of 3.2 cycles min-1. The magnitudes of motions beyond breathing were found to be significant, with 14.4 and 33.8 mm maximal motions measured from antral contraction and slow configuration changes, respectively. Hierarchical motion models have potential in multiple applications in radiotherapy, including improving the accuracy of dose delivery estimation, providing guidance for margin creation, and supporting advanced decisions and strategies for immobilization, treatment monitoring and gating.

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.

An analytical model for the upper bound estimation of respiratory motion-induced dose uncertainty in spot-scanning proton beam therapy

PURPOSE

We developed an analytical model of a spot-scanning beam delivery system to estimate the upper bound of respiratory motion-induced dose uncertainty for a given treatment plan.

METHODS

The effective delivery time for each spot position in the treatment plan was calculated on the basis of the parameters of the delivery system. The upper bound of the dose uncertainty was then calculated as a function of the effective delivery time. Two-dimensional (2D) measurements with a detector array on a one-dimensional moving platform were obtained to validate the model.

RESULTS

We performed 351 two-dimensional measurements on a moving platform for different delivery sequences of a single-layer uniform pattern and patient treatment field. The measured dose uncertainty was a strong function of the effective delivery time: The shortest effective delivery time resulted in a maximum absolute dose error of >90%, while the longest ones resulted in a maximum absolute dose error of 4.9% for a single layer and 9.7% for a patient field with heterogeneity. The relationship of the effective delivery time and the measured dose uncertainty followed the analytical formula.

CONCLUSIONS

With our analytical model, the upper bound of the dose uncertainty due to motion can be estimated in spot-scanning proton therapy without four-dimensional dynamic dose calculation.

Nearest Neighbor Method to Estimate Internal Target for Real-Time Tumor Tracking

PURPOSE

This work proposed a nearest neighbor estimation method to track the respiration-induced tumor motion.

METHODS

Based on the simultaneously collected motion traces of external surrogate and internal target during the modeling phase prior to treatment, we first obtain the nearest neighbors of the current surrogate in external space. Subsequently, the concurrent targets in internal space are determined and used to estimate the current target position. The method was validated on 71 cases that were from 3 open access databases. In addition, to evaluate the method's estimation and prediction accuracy, we compared the method with other works.

RESULTS

Except for 2 cases, the nearest neighbor estimation achieved the root-mean-square error of <3 mm. The comparison indicated that the method had better estimation accuracy than polynomial model and good prediction performance.

DISCUSSION

The 2 exceptive cases were further analyzed for failure causes. We inferred that one was because of the lack of estimating new target in our method, and the other one was because of the mistake during data collection. Accordingly, the potential solutions were suggested. Besides, the method's estimation for surrogate outliers, effects of modeling length, calibration, and extension were discussed.

CONCLUSION

The results demonstrated nearest neighbor estimation's effectiveness. Except for this, the method imposes no restrictions on the modality of the pretreatment target images and does not assume a specific correspondence function between the surrogate and the target. With only 1 critical parameter, this nearest neighbor estimation method is easy to implement in clinical setting and thus has potential for broad applications.

rConverse: Moment by Moment Conversation Detection Using a Mobile Respiration Sensor

Monitoring of in-person conversations has largely been done using acoustic sensors. In this paper, we propose a new method to detect moment-by-moment conversation episodes by analyzing breathing patterns captured by a mobile respiration sensor. Since breathing is affected by physical and cognitive activities, we develop a comprehensive method for cleaning, screening, and analyzing noisy respiration data captured in the field environment at individual breath cycle level. Using training data collected from a speech dynamics lab study with 12 participants, we show that our algorithm can identify each respiration cycle with 96.34% accuracy even in presence of walking. We present a Conditional Random Field, Context-Free Grammar (CRF-CFG) based conversation model, called rConverse, to classify respiration cycles into speech or non-speech, and subsequently infer conversation episodes. Our model achieves 82.7% accuracy for speech/non-speech classification and it identifies conversation episodes with 95.9% accuracy on lab data using a leave-one-subject-out cross-validation. Finally, the system is validated against audio ground-truth in a field study with 32 participants. rConverse identifies conversation episodes with 71.7% accuracy on 254 hours of field data. For comparison, the accuracy from a high-quality audio-recorder on the same data is 71.9%.

Dosimetric Analysis of Microscopic Disease in SBRT for Lung Cancers

Objective:

The objective of this study is to theoretically and experimentally evaluate the dosimetry in the microscopic disease regions surrounding the tumor under stereotactic body radiation therapy of lung cancer.

Methods:

For simplicity, the tumor was considered moving along 1 dimension with a periodic function. The probability distribution function of the tumor position was generated according to the motion pattern and was used to estimate the delivered dose in the microscopic disease region. An experimental measurement was conducted to validate both the estimated dose with a probability function and the calculated dose from 4-dimensional computed tomography data using a dynamic thorax phantom. Four tumor motion patterns were simulated with cos⁴(x) and sin(x), each with 2 different amplitudes: 10 mm and 5 mm. A 7-field conformal plan was created for treatment delivery. Both films (EBT2) and optically stimulated luminescence detectors were inserted in and around the target of the phantom to measure the delivered doses. Dose differences were evaluated using gamma analysis with 3%/3 mm.

Results:

The average gamma index between measured doses using film and calculated doses using average intensity projection simulation computed tomography was 80.8% ± 0.9%. In contrast, between measured doses using film and calculated doses accumulated from 10 sets of 4-dimensional computed tomography data, it was 98.7% ± 0.6%. The measured doses using optically stimulated luminescence detectors matched very well (within 5% of the measurement uncertainty) with the theoretically calculated doses using probability distribution function at the corresponding position. Respiratory movement caused inadvertent irradiation exposure, with 70% to 80% of the dose line wrapped around the 10 mm region outside the target.

Conclusion:

The use of static dose calculation in the treatment planning system could substantially underestimate the actual delivered dose in the microscopic disease region for a moving target. The margin for microscopic disease may be substantially reduced or even eliminated for lung stereotactic body radiation therapy.

Sin-quadratic model for chest tomosynthesis respiratory signal analysis and its application in four dimensional chest tomosynthesis reconstruction

Chest tomosynthesis (CTS) is a newly developed imaging technique which provides pseudo-3D volume anatomical information of thorax from limited-angle projections and contains much less of superimposed anatomy than the chest X-ray radiography. One of the relatively common problems in CTS is the patient respiratory motion during image acquisition, which negatively impacts the detectability. In this work, we propose a sin-quadratic model to analyze the respiratory motion during CTS scan, which is a real time method where the respiratory signal is generated by extracting the motion of diaphragm from projection radiographs. According to the estimated respiratory signal, the CTS projections were then amplitude-based sorted into four to eight phases, and an iterative reconstruction strategy with total variation regularization was adopted to reconstruct the CTS images at each phase. Simulated digital XCAT phantom data and three sets of patient data were adopted for the experiments to validate the performance of the sin-quadratic model and its application in four dimensional (4D) CTS reconstruction. Results of the XCAT phantom simulation study show that the correlation coefficient between the extracted respiratory signal and the originally designed respiratory signal is 0.9964, which suggests that the proposed model could exactly extract the respiratory signal from CTS projections. The 4D CTS reconstructions of both the phantom data and the patient data show clear reduction of motion-induced blur.

Usefulness of a new online patient-specific quality assurance system for respiratory-gated radiotherapy

PURPOSE

The accuracy of gated irradiation may decrease when treatment is performed with short "beam-on" times. Also, the dose is subject to variation between treatment sessions if the respiratory rate is irregular. We therefore evaluated the impact of the differences between gated and non-gated treatment on doses using a new online quality assurance (QA) system for respiratory-gated radiotherapy.

METHODS

We generated dose estimation models to associate dose and pulse information using a 0.6 cc Farmer chamber and our QA system. During gated irradiation with each of seven regular and irregular respiratory patterns, with the Farmer chamber readings as references, we evaluated our QA system's accuracy. We then used the QA system to assess the impact of respiratory patterns on dose distribution for three lung and three liver radiotherapy plans. Gated and non-gated plans were generated and compared.

RESULTS

There was agreement within 1.7% between the ionization chamber and our system for several regular and irregular motion patterns. For dose distributions with measured errors, there were larger differences between gated and non-gated treatment for high-dose regions within the planned treatment volume (PTV). Compared with a non-gated plan, PTV D_95% for a gated plan decreased by -1.5% to -2.6%. Doses to organs at risk were similar with both plans.

CONCLUSIONS

Our simple system estimated the radiation dose to the patient using only pulse information from the linac, even during irregular respiration. The quality of gated irradiation for each patient can be verified fraction by fraction.

Accuracy of the dose-shift approximation in estimating the delivered dose in SBRT of lung tumors considering setup errors and breathing motions

INTRODUCTION

Geometrical uncertainties can result in a delivered dose to the tumor different from that estimated in the static treatment plan. The purpose of this project was to investigate the accuracy of the dose calculated to the clinical target volume (CTV) with the dose-shift approximation, in stereotactic body radiation therapy (SBRT) of lung tumors considering setup errors and breathing motion. The dose-shift method was compared with a beam-shift method with dose recalculation.

MATERIAL AND METHODS

Included were 10 patients (10 tumors) selected to represent a variety of SBRT-treated lung tumors in terms of tumor location, CTV volume, and tumor density. An in-house developed toolkit within a treatment planning system allowed the shift of either the dose matrix or a shift of the beam isocenter with dose recalculation, to simulate setup errors and breathing motion. Setup shifts of different magnitudes (up to 10 mm) and directions as well as breathing with different peak-to-peak amplitudes (up to 10:5:5 mm) were modeled. The resulting dose-volume histograms (DVHs) were recorded and dose statistics were extracted.

RESULTS

Generally, both the dose-shift and beam-shift methods resulted in calculated doses lower than the static planned dose, although the minimum (D_98%) dose exceeded the prescribed dose in all cases, for setup shifts up to 5 mm. The dose-shift method also generally underestimated the dose compared with the beam-shift method. For clinically realistic systematic displacements of less than 5 mm, the results demonstrated that in the minimum dose region within the CTV, the dose-shift method was accurate to 2% (root-mean-square error). Breathing motion only marginally degraded the dose distributions.

CONCLUSIONS

Averaged over the patients and shift directions, the dose-shift approximation was determined to be accurate to approximately 2% (RMS) within the CTV, for clinically relevant geometrical uncertainties for SBRT of lung tumors.

Feasibility of using single photon counting X-ray for lung tumor position estimation based on 4D-CT

PURPOSE

In stereotactic body radiation therapy of lung tumors, reliable position estimation of the tumor is necessary in order to minimize normal tissue complication rate. While kV X-ray imaging is frequently used, continuous application during radiotherapy sessions is often not possible due to concerns about the additional dose. Thus, ultra low-dose (ULD) kV X-ray imaging based on a single photon counting detector is suggested. This paper addresses the lower limit of photons to locate the tumor reliably with an accuracy in the range of state-of-the-art methods, i.e. a few millimeters.

METHOD

18 patient cases with four dimensional CT (4D-CT), which serves as a-priori information, are included in the study. ULD cone beam projections are simulated from the 4D-CTs including Poisson noise. The projections from the breathing phases which correspond to different tumor positions are compared to the ULD projection by means of Poisson log-likelihood (PML) and correlation coefficient (CC), and template matching under these metrics.

RESULTS

The results indicate that in full thorax imaging five photons per pixel suffice for a standard deviation in tumor positions of less than half a breathing phase. Around 50 photons per pixel are needed to achieve this accuracy with the field of view restricted to the tumor region. Compared to CC, PML tends to perform better for low photon counts and shifts in patient setup. Template matching only improves the position estimation in high photon counts. The quality of the reconstruction is independent of the projection angle.

CONCLUSIONS

The accuracy of the proposed ULD single photon counting system is in the range of a few millimeters and therefore comparable to state-of-the-art tumor tracking methods. At the same time, a reduction in photons per pixel by three to four orders of magnitude relative to commercial systems with flatpanel detectors can be achieved. This enables continuous kV image-based position estimation during all fractions since the additional dose to the patient is negligible.

Technical note: real-time web-based wireless visual guidance system for radiotherapy

Describe a Web-based wireless visual guidance system that mitigates issues associated with hard-wired audio-visual aided patient interactive motion management systems that are cumbersome to use in routine clinical practice. Web-based wireless visual display duplicates an existing visual display of a respiratory-motion management system for visual guidance. The visual display of the existing system is sent to legacy Web clients over a private wireless network, thereby allowing a wireless setting for real-time visual guidance. In this study, active breathing coordinator (ABC) trace was used as an input for visual display, which captured and transmitted to Web clients. Virtual reality goggles require two (left and right eye view) images for visual display. We investigated the performance of Web-based wireless visual guidance by quantifying (1) the network latency of visual displays between an ABC computer display and Web clients of a laptop, an iPad mini 2 and an iPhone 6, and (2) the frame rate of visual display on the Web clients in frames per second (fps). The network latency of visual display between the ABC computer and Web clients was about 100 ms and the frame rate was 14.0 fps (laptop), 9.2 fps (iPad mini 2) and 11.2 fps (iPhone 6). In addition, visual display for virtual reality goggles was successfully shown on the iPhone 6 with 100 ms and 11.2 fps. A high network security was maintained by utilizing the private network configuration. This study demonstrated that a Web-based wireless visual guidance can be a promising technique for clinical motion management systems, which require real-time visual display of their outputs. Based on the results of this study, our approach has the potential to reduce clutter associated with wired-systems, reduce space requirements, and extend the use of medical devices from static usage to interactive and dynamic usage in a radiotherapy treatment vault.

Audiovisual biofeedback guided breath-hold improves lung tumor position reproducibility and volume consistency

Purpose

Respiratory variation can increase the variability of tumor position and volume, accounting for larger treatment margins and longer treatment times. Audiovisual biofeedback as a breath-hold technique could be used to improve the reproducibility of lung tumor positions at inhalation and exhalation for the radiation therapy of mobile lung tumors. This study aimed to assess the impact of audiovisual biofeedback breath-hold (AVBH) on interfraction lung tumor position reproducibility and volume consistency for respiratory-gated lung cancer radiation therapy.

Methods

Lung tumor position and volume were investigated in 9 patients with lung cancer who underwent a breath-hold training session with AVBH before 2 magnetic resonance imaging (MRI) sessions. During the first MRI session (before treatment), inhalation and exhalation breath-hold 3-dimensional MRI scans with conventional breath-hold (CBH) using audio instructions alone and AVBH were acquired. The second MRI session (midtreatment) was repeated within 6 weeks after the first session. Gross tumor volumes (GTVs) were contoured on each dataset. CBH and AVBH were compared in terms of tumor position reproducibility as assessed by GTV centroid position and position range (defined as the distance of GTV centroid position between inhalation and exhalation) and tumor volume consistency as assessed by GTV between inhalation and exhalation.

Results

Compared with CBH, AVBH improved the reproducibility of interfraction GTV centroid position by 46% (P = .009) from 8.8 mm to 4.8 mm and GTV position range by 69% (P = .052) from 7.4 mm to 2.3 mm. Compared with CBH, AVBH also improved the consistency of intrafraction GTVs by 70% (P = .023) from 7.8 cm³ to 2.5 cm³.

Conclusions

This study demonstrated that audiovisual biofeedback can be used to improve the reproducibility and consistency of breath-hold lung tumor position and volume, respectively. These results may provide a pathway to achieve more accurate lung cancer radiation treatment in addition to improving various medical imaging and treatments by using breath-hold procedures.

Motion and volumetric change as demonstrated by 4DCT: The effects of abdominal compression on the GTV, lungs, and heart in lung cancer patients

PURPOSE

Lung tumors move during respiration, complicating radiation therapy. The abdominal compression plate (ACP) is thought to reduce respiratory motion. This study quantifies ACP efficacy on respiratory-induced motion by using 4-dimensional computed tomography to evaluate volume and displacement changes of the heart, lungs, and tumor with and without ACP.

METHODS AND MATERIALS

Lung cancer patients (n = 17) received 4-dimensional computed tomography simulations (10 computed tomography scans from 0% to 90% breathing phases) with and without ACP under maximally tolerated diaphragmatic pressure. Gross tumor volume (GTV), heart, and lungs were contoured in treatment planning software for each phase. Structures were exported for analysis. For each phase, with and without ACP, tumor and organ absolute centroid range of motion and volume were calculated.

RESULTS

ACP did not significantly affect GTV, heart, or lung motion on the sample as a whole, but instead demonstrated patient-specific results. ACP reduced GTV motion in 3 (17.6%; 3 upper lobe tumors) by 2.9 mm (P < .01), increased motion in 5 (29.4%; 3 upper lobe tumors, 1 middle lobe, 1 lower lobe) by 1.9 mm (P < .03), and did not significantly change 9. Of the 3 patients exhibiting significantly decreased GTV motion, GTV, heart, and lung range of motion was 7.4 mm, 11.8 mm, and 11.9 mm, respectively, without compression and 4.5 mm, 8.4 mm, and 10.9 mm, respectively, with compression. Averaged across the sample, ACP did not exhibit any axis-specific effect.

CONCLUSIONS

ACP efficacy was patient-specific, possibly because of pre-existing factors including chronic obstructive pulmonary disease severity, chest wall elasticity, tumor location, and patient comfort. Tumor lobe location does not predetermine compression efficacy; therefore, patients should be simulated with and without ACP, regardless of tumor location. GTV motion seems most important in determining suitability for compression. Alternative motion control should be considered in patients not benefited by compression. In patients who benefited, ACP may enhance tumor coverage while minimizing toxicity. Larger scale studies are necessary for definitive treatment recommendations.

Modeling the respiratory motion of solitary pulmonary nodules and determining the impact of respiratory motion on their detection in SPECT imaging

The objectives of this investigation were to model the respiratory motion of solitary pulmonary nodules (SPN) and then use this model to determine the impact of respiratory motion on the localization and detection of small SPN in SPECT imaging for four reconstruction strategies. The respiratory motion of SPN was based on that of normal anatomic structures in the lungs determined from breath-held CT images of a volunteer acquired at two different stages of respiration. End-expiration (EE) and time-averaged (Frame Av) non-uniform-B-spline cardiac torso (NCAT) digital-anthropomorphic phantoms were created using this information for respiratory motion within the lungs. SPN were represented as 1 cm diameter spheres which underwent linear motion during respiration between the EE and end-inspiration (EI) time points. The SIMIND Monte Carlo program was used to produce SPECT projection data simulating Tc-99m depreotide (NeoTect) imaging. The projections were reconstructed using 1) no correction (NC), 2) attenuation correction (AC), 3) resolution compensation (RC), and 4) attenuation correction, scatter correction, and resolution compensation (AC_SC_RC). A human-observer localization receiver operating characteristics (LROC) study was then performed to determine the difference in localization and detection accuracy with and without the presence of respiratory motion. The LROC comparison determined that respiratory motion degrades tumor detection for all four reconstruction strategies, thus correction for SPN motion would be expected to improve detection accuracy. The inclusion of RC in reconstruction improved detection accuracy for both EE and Frame Av over NC and AC. Also the magnitude of the impact of motion was least for AC_SC_RC.

Lungtech, a phase II EORTC trial of SBRT for centrally located lung tumours - a clinical physics perspective

Background

The EORTC has launched a phase II trial to assess safety and efficacy of SBRT for centrally located NSCLC: The EORTC 22113-08113—Lungtech trial. Due to neighbouring critical structures, these tumours remain challenging to treat. To guarantee accordance to protocol and treatment safety, an RTQA procedure has been implemented within the frame of the EORTC RTQA levels. These levels are here expanded to include innovative tools beyond protocol compliance verification: the actual dose delivered to each patient will be estimated and linked to trial outcomes to enable better understanding of dose related response and toxicity.

Method

For trial participation, institutions must provide a completed facility questionnaire and beam output audit results. To insure ability to comply with protocol specifications a benchmark case is sent to all centres. After approval, institutions are allowed to recruit patients. Nonetheless, each treatment plan will be prospectively reviewed insuring trial compliance consistency over time. As new features, patient’s CBCT images and applied positioning corrections will be saved for dose recalculation on patient’s daily geometry. To assess RTQA along the treatment chain, institutions will be visited once during the time of the trial. Over the course of this visit, end-to-end tests will be performed using the 008ACIRS-breathing platform with two separate bodies. The first body carries EBT3 films and an ionization chamber. The other body newly developed for PET- CT evaluation is fillable with a solution of high activity. 3D or 4D PET-CT and 4D-CT scanning techniques will be evaluated to assess the impact of motion artefacts on target volume accuracy. Finally, a dosimetric evaluation in static and dynamic conditions will be performed.

Discussion

Previous data on mediastinal toxicity are scarce and source of cautiousness for setting-up SBRT treatments for centrally located NSCLC. Thanks to the combination of documented patient related outcomes and CBCT based dose recalculation we expect to provide improved models for dose response and dose related toxicity.

Conclusion

We have developed a comprehensive RTQA model for trials involving modern radiotherapy. These procedures could also serve as examples of extended RTQA for future radiotherapy trials involving quantitative use of PET and tumour motion.

Robustness of sweeping-window arc therapy treatment sequences against intrafractional tumor motion

PURPOSE

Due to the potentially periodic collimator dynamic in volumetric modulated arc therapy (VMAT) dose deliveries with the sweeping-window arc therapy (SWAT) technique, additional manifestations of dosimetric deviations in the presence of intrafractional motion may occur. With a fast multileaf collimator (MLC), and a flattening filter free dose delivery, treatment times close to 60 s per fraction are clinical reality. For these treatment sequences, the human breathing period can be close to the collimator sweeping period. Compared to a random arrangement of the segments, this will cause a further degradation of the dose homogeneity.

METHODS

Fifty VMAT sequences of potentially moving target volumes were delivered on a two dimensional ionization chamber array. In order to detect interplay effects along all three coordinate axes, time resolved measurements were performed twice--with the detector aligned in vertical (V) or horizontal (H) orientation. All dose matrices were then moved within a simulation software by a time-dependent motion vector. The minimum relative equivalent uniform dose EUDr,m for all breathing starting phases was determined for each amplitude and period. Furthermore, an estimation of periods with minimum EUD was performed. Additionally, LINAC logfiles were recorded during plan delivery. The MLC, jaw, gantry angle, and monitor unit settings were continuously saved and used to calculate the correlation coefficient between the target motion and the dose weighed collimator motion component for each direction (CC, LR, AP) separately.

RESULTS

The resulting EUDr,m were EUDr,m(CCV) = (98.3 ± 0.6)%, EUDr,m(CCH) = (98.6 ± 0.5)%, EUDr,m(APV) = (97.7 ± 0.9)%, and EUDr,m(LRH) = (97.8 ± 0.9)%. The overall minimum relative EUD observed for 360(∘) arc midventilation treatments was 94.6%. The treatment plan with the shortest period and a minimum relative EUD of less than 97% was found at T = 6.1 s. For a partial 120(∘) arc, an EUDr,m = 92.0% was found. In all cases, a correlation coefficient above 0.5 corresponded to a minimum in EUD.

CONCLUSIONS

With the advent of fast VMAT delivery techniques, nonrobust treatment sequences for human breathing patterns can be generated. These sequences are characterized by a large correlation coefficient between a target motion component and the corresponding collimator dynamic. By iteratively decreasing the maximum allowed dose rate, a low correlation coefficient and consequentially a robust treatment sequence are ensured.

Audiovisual Biofeedback Improves Cine-Magnetic Resonance Imaging Measured Lung Tumor Motion Consistency

PURPOSE

To assess the impact of an audiovisual (AV) biofeedback on intra- and interfraction tumor motion for lung cancer patients.

METHODS AND MATERIALS

Lung tumor motion was investigated in 9 lung cancer patients who underwent a breathing training session with AV biofeedback before 2 3T magnetic resonance imaging (MRI) sessions. The breathing training session was performed to allow patients to become familiar with AV biofeedback, which uses a guiding wave customized for each patient according to a reference breathing pattern. In the first MRI session (pretreatment), 2-dimensional cine-MR images with (1) free breathing (FB) and (2) AV biofeedback were obtained, and the second MRI session was repeated within 3-6 weeks (mid-treatment). Lung tumors were directly measured from cine-MR images using an auto-segmentation technique; the centroid and outlier motions of the lung tumors were measured from the segmented tumors. Free breathing and AV biofeedback were compared using several metrics: intra- and interfraction tumor motion consistency in displacement and period, and the outlier motion ratio.

RESULTS

Compared with FB, AV biofeedback improved intrafraction tumor motion consistency by 34% in displacement (P=.019) and by 73% in period (P<.001). Compared with FB, AV biofeedback improved interfraction tumor motion consistency by 42% in displacement (P<.046) and by 74% in period (P=.005). Compared with FB, AV biofeedback reduced the outlier motion ratio by 21% (P<.001).

CONCLUSIONS

These results demonstrated that AV biofeedback significantly improved intra- and interfraction lung tumor motion consistency for lung cancer patients. These results demonstrate that AV biofeedback can facilitate consistent tumor motion, which is advantageous toward achieving more accurate medical imaging and radiation therapy procedures.

Impact of temporal probability in 4D dose calculation for lung tumors

The purpose of this study was to evaluate the dosimetric uncertainty in 4D dose calculation using three temporal probability distributions: uniform distribution, sinusoidal distribution, and patient‐specific distribution derived from the patient respiratory trace. Temporal probability, defined as the fraction of time a patient spends in each respiratory amplitude, was evaluated in nine lung cancer patients. Four‐dimensional computed tomography (4D CT), along with deformable image registration, was used to compute 4D dose incorporating the patient's respiratory motion. First, the dose of each of 10 phase CTs was computed using the same planning parameters as those used in 3D treatment planning based on the breath‐hold CT. Next, deformable image registration was used to deform the dose of each phase CT to the breath‐hold CT using the deformation map between the phase CT and the breath‐hold CT. Finally, the 4D dose was computed by summing the deformed phase doses using their corresponding temporal probabilities. In this study, 4D dose calculated from the patient‐specific temporal probability distribution was used as the ground truth. The dosimetric evaluation matrix included: 1) 3D gamma analysis, 2) mean tumor dose (MTD), 3) mean lung dose (MLD), and 4) lung V20. For seven out of nine patients, both uniform and sinusoidal temporal probability dose distributions were found to have an average gamma passing rate >95% for both the lung and PTV regions. Compared with 4D dose calculated using the patient respiratory trace, doses using uniform and sinusoidal distribution showed a percentage difference on average of −0.1%±0.6% and −0.2%±0.4% in MTD, −0.2%±1.9% and −0.2%±1.3% in MLD, 0.09%±2.8% and −0.07%±1.8% in lung V20, −0.1%±2.0% and 0.08%±1.34% in lung V10, 0.47%±1.8% and 0.19%±1.3% in lung V5, respectively. We concluded that four‐dimensional dose computed using either a uniform or sinusoidal temporal probability distribution can approximate four‐dimensional dose computed using the patient‐specific respiratory trace.

PACS number: 87.55.D‐

Impact of scanning parameters and breathing patterns on image quality and accuracy of tumor motion reconstruction in 4D CBCT: a phantom study

Four-dimensional, cone-beam CT (4D CBCT) substantially reduces respiration-induced motion blurring artifacts in three-dimension (3D) CBCT. However, the image quality of 4D CBCT is significantly degraded which may affect its accuracy in localizing a mobile tumor for high-precision, image-guided radiation therapy (IGRT). The purpose of this study was to investigate the impact of scanning parameters hereinafter collectively referred to as scanning sequence) and breathing patterns on the image quality and the accuracy of computed tumor trajectory for a commercial 4D CBCT system, in preparation for its clinical implementation. We simulated a series of periodic and aperiodic sinusoidal breathing patterns with a respiratory motion phantom. The aperiodic pattern was created by varying the period or amplitude of individual sinusoidal breathing cycles. 4D CBCT scans of the phantom were acquired with a manufacturer-supplied scanning sequence (4D-S-slow) and two in-house modified scanning sequences (4D-M-slow and 4D-M-fast). While 4D-S-slow used small field of view (FOV), partial rotation (200°), and no imaging filter, 4D-M-slow and 4D-M-fast used medium FOV, full rotation, and the F1 filter. The scanning speed was doubled in 4D-M-fast (100°/min gantry rotation). The image quality of the 4D CBCT scans was evaluated using contrast-to-noise ratio (CNR), signal-to-noise ratio (SNR), and motion blurring ratio (MBR). The trajectory of the moving target was reconstructed by registering each phase of the 4D CBCT with a reference CT. The root-mean-squared-error (RMSE) analysis was used to quantify its accuracy. Significant decrease in CNR and SNR from 3D CBCT to 4D CBCT was observed. The 4D-S-slow and 4D-M-fast scans had comparable image quality, while the 4D-M-slow scans had better performance due to doubled projections. Both CNR and SNR decreased slightly as the breathing period increased, while no dependence on the amplitude was observed. The difference of both CNR and SNR between periodic and aperiodic breathing patterns was insignificant (p > 0.48). At end-exhale phases, the motion blurring was negligible for both periodic and aperiodic breathing patterns; at mid-inhale phase, the motion blurring increased as the period, the amplitude or the amount of cycle-to-cycle variation on amplitude increased. Overall, the accuracy of localizing the moving target in 4D CBCT was within 2 mm under all studied cases. No difference in the RMSEs was noticed among the three scanning sequences. The 4D-M-fast scans, free of volume truncation artifacts, exhibited comparable image quality and accuracy in tumor motion reconstruction as the 4D-S-slow scans with reduced imaging dose (0.60 cGy vs. 0.99 cGy) due to the use of faster gantry rotation and the F1 filter, suggesting its suitability for clinical use.

Respiratory motion management using audio-visual biofeedback for respiratory-gated radiotherapy of synchrotron-based pulsed heavy-ion beam delivery

PURPOSE

To efficiently deliver respiratory-gated radiation during synchrotron-based pulsed heavy-ion radiotherapy, a novel respiratory guidance method combining a personalized audio-visual biofeedback (BFB) system, breath hold (BH), and synchrotron-based gating was designed to help patients synchronize their respiratory patterns with synchrotron pulses and to overcome typical limitations such as low efficiency, residual motion, and discomfort.

METHODS

In-house software was developed to acquire body surface marker positions and display BFB, gating signals, and real-time beam profiles on a LED screen. Patients were prompted to perform short BHs or short deep breath holds (SDBH) with the aid of BFB following a personalized standard BH/SDBH (stBH/stSDBH) guiding curve or their own representative BH/SDBH (reBH/reSDBH) guiding curve. A practical simulation was performed for a group of 15 volunteers to evaluate the feasibility and effectiveness of this method. Effective dose rates (EDRs), mean absolute errors between the guiding curves and the measured curves, and mean absolute deviations of the measured curves were obtained within 10%-50% duty cycles (DCs) that were synchronized with the synchrotron's flat-top phase.

RESULTS

All maneuvers for an individual volunteer took approximately half an hour, and no one experienced discomfort during the maneuvers. Using the respiratory guidance methods, the magnitude of residual motion was almost ten times less than during nongated irradiation, and increases in the average effective dose rate by factors of 2.39-4.65, 2.39-4.59, 1.73-3.50, and 1.73-3.55 for the stBH, reBH, stSDBH, and reSDBH guiding maneuvers, respectively, were observed in contrast with conventional free breathing-based gated irradiation, depending on the respiratory-gated duty cycle settings.

CONCLUSIONS

The proposed respiratory guidance method with personalized BFB was confirmed to be feasible in a group of volunteers. Increased effective dose rate and improved overall treatment precision were observed compared to conventional free breathing-based, respiratory-gated irradiation. Because breathing guidance curves could be established based on the respective average respiratory period and amplitude for each patient, it may be easier for patients to cooperate using this technique.

Digital reconstruction of high-quality daily 4D cone-beam CT images using prior knowledge of anatomy and respiratory motion

Conventional in-room cone-beam computed tomography (CBCT) lacks explicit representation of patient respiratory motion and usually has poor image quality and inaccurate CT numbers for target delineation and/or adaptive treatment planning. In-room four-dimensional (4D) CBCT image acquisition is still time consuming and suffers the same issue of poor image quality. To overcome this limitation, we developed a computational framework to digitally synthesize high-quality daily 4D CBCT images using the prior knowledge of motion and appearance learned from the planning 4D CT dataset. A patient-specific respiratory motion model was first constructed from the planning 4D CT images using principal component analysis of displacement vector fields across different respiratory phases. Subsequently, the respiratory motion model as well as the image content of the planning CT was spatially mapped onto the daily CBCT using deformable image registration. The synthesized 4D images possess explicit patient motion while maintaining the geometric accuracy of patient's anatomy at the time of treatment. We validated our model by quantitatively comparing the synthesized 4D CBCT against the 4D CT dataset acquired in the same day from protocol patients undergoing daily in-room CBCT setup and weekly 4D CT for treatment evaluation. Our preliminary results have demonstrated good agreement of contours in different motion phases between the synthesized and acquired scans. Various imaging artifacts were also suppressed and soft-tissue visibility was enhanced.

Comparing approaches to correct for respiratory motion in NH3 PET-CT cardiac perfusion imaging

Aim

Respiratory motion affects cardiac PET-computed tomography (CT) imaging by reducing attenuation correction (AC) accuracy and by introducing blur. The aim of this study was to compare three approaches for reducing motion-induced AC errors and evaluate the inclusion of respiratory motion correction.

Materials and methods

AC with a helical CT was compared with averaged cine and gated cine CT, as well as with a pseudo-gated CT, which was produced by applying PET-derived motion fields to the helical CT. Data-driven gating was used to produce respiratory-gated PET and CT images, and 60 NH₃ PET scans were attenuation corrected with each of the CTs. Respiratory motion correction was applied to the gated and pseudo-gated attenuation-corrected PET images.

Results

Anterior and lateral wall intensity measured in attenuation-corrected PET images generally increased when PET-CT alignment improved and decreased when alignment degraded. On average, all methods improved PET-CT liver and cardiac alignment, and increased anterior wall intensity by more than 10% in 36, 33 and 25 cases for the averaged, gated and pseudo-gated CTAC PET images, respectively. However, cases were found where alignment worsened and severe artefacts resulted. This occurred in more cases and to a greater extent for the averaged and gated CT, where the anterior wall intensity reduced by more than 10% in 21 and 24 cases, respectively, compared with six cases for the pseudo-gated CT. Application of respiratory motion correction increased the average anterior and inferior wall intensity, but only 13% of cases increased by more than 10%.

Conclusion

All methods improved average respiratory-induced AC errors; however, some severe artefacts were produced. The pseudo-gated CT was found to be the most robust method.

Efficacy evaluation of retrospectively applying the Varian normal breathing predictive filter for volume definition and artifact reduction in 4D CT lung patients

Phase-based sorting of four-dimensional computed tomography (4D CT) datasets is prone to image artifacts due to patient's breathing irregularities that occur during the image acquisition. The purpose of this study is to investigate the effect of the Varian normal breathing predictive filter (NBPF) as a retrospective phase-sorting parameter in 4D CT. Ten 4D CT lung cancer datasets were obtained. The volumes of all tumors present, as well as the total lung volume, were calculated on the maximum intensity projection (MIP) images as well as each individual phase image. The NBPF was varied retrospectively within the available range, and changes in volume and image quality were recorded. The patients' breathing trace was analysed and the magnitude and location of any breathing irregularities were correlated to the behavior of the NBPF. The NBPF was found to have a considerable effect on the quality of the images in MIP and single-phase datasets. When used appropriately, the NBPF is shown to have the ability to account for and correct image artifacts. However, when turned off (0%) or set above a critical level (approximately 40%), it resulted in erroneous volume reconstructions with variations in tumor volume up to 26.6%. Those phases associated with peak inspiration were found to be more susceptible to changes in the NBPF. The NBPF settings selected prior to exporting the breathing trace for patients evaluated using 4D CT directly affect the accuracy of the targeting and volume estimation of lung tumors. Recommendations are made to address potential errors in patient anatomy introduced by breathing irregularities, specifically deep breath or cough irregularities, by implementing the proper settings and use of this tool.

IMRT treatment planning on 4D geometries for the era of dynamic MLC tracking

The problem addressed here was to obtain optimal and deliverable dynamic multileaf collimator (MLC) leaf sequences from four-dimensional (4D) geometries for dynamic MLC tracking delivery. The envisaged scenario was where respiratory phase and position information of the target was available during treatment, from which the optimal treatment plan could be further adapted in real time. A tool for 4D treatment plan optimization was developed that integrates a commercially available treatment planning system and a general-purpose optimization system. The 4D planning method was applied to the 4D computed tomography planning scans of three lung cancer patients. The optimization variables were MLC leaf positions as a function of monitor units and respiratory phase. The objective function was the deformable dose-summed 4D treatment plan score. MLC leaf motion was constrained by the maximum leaf velocity between control points in terms of monitor units for tumor motion parallel to the leaf travel direction and between phases for tumor motion parallel to the leaf travel direction. For comparison and a starting point for the 4D optimization, three-dimensional (3D) optimization was performed on each of the phases. The output of the 4D IMRT planning process is a leaf sequence which is a function of both monitor unit and phase, which can be delivered to a patient whose breathing may vary between the imaging and treatment sessions. The 4D treatment plan score improved during 4D optimization by 34%, 4%, and 50% for Patients A, B, and C, respectively, indicating 4D optimization generated a better 4D treatment plan than the deformable sum of individually optimized phase plans. The dose-volume histograms for each phase remained similar, indicating robustness of the 4D treatment plan to respiratory variations expected during treatment delivery. In summary, 4D optimization for respiratory phase-dependent treatment planning with dynamic MLC motion tracking improved the 4D treatment plan score by 4–50% compared with 3D optimization. The 4D treatment plans had leaf sequences that varied from phase to phase to account for anatomic motion, but showed similar target dose distributions in each phase. The current method could in principle be generalized for use in offline replanning between fractions or for online 4D treatment planning based on 4D cone-beam CT images. Computation time remains a challenge.

Use of MRI to assess the prediction of heart motion with gross body motion in myocardial perfusion imaging by stereotracking of markers on the body surface

PURPOSE

The aim of this study is to determine using MRI in volunteers whether the rigid-body-motion (RBM) model can be approximately used to estimate the gross body-motion of the heart from that of external markers on patient's chest. Our target clinical application is to use a visual-tracking-system (VTS) which employs stereoimaging to estimate heart motion during SPECT/CT and PET∕CT myocardial perfusion imaging.

METHODS

To investigate body-motion separate from the respiration the authors had the volunteers hold their breath during the acquisition of a sequence of two sets of EKG-triggered MRI sagittal slices. The first set was acquired pre-motion, and the second postmotion. The motion of the heart within each breath-hold set of slices was estimated by registration to the semiautomatic 3D segmentation of the heart region in a baseline set acquired using the Navigator technique. The motion of the heart between the pre- and postmotion sets was then determined as the difference in the individual motions in comparison to the Navigator sets. An analysis of the combined motion of the individual markers on the chest was used to obtain an estimate of the six-degree-of-freedom RBM from the VTS system. The metric for judging agreement between the motion estimated by MRI and the VTS was the average error. This was defined as the average of the magnitudes of the differences in the vector displacements of all voxels in the heart region. Studies with the Data Spectrum Anthropomorphic Phantom and "No-Motion" studies in which the volunteer did not intentionally move were used to establish a baseline for agreement. With volunteer studies a t-test was employed to determine when statistically significant differences in Average Errors occurred compared to the No-motion studies.

RESULTS

For phantom acquisitions, the Average Error when the motion was just translation was 0.1 mm. With complex motions, which included a combination of rotations and translations, the Average Error increased to 3.6 mm. In the volunteers the Average Error averaged over all No-Motion acquisitions was 1.0 mm. For the case of translational motion, which might be expected to be RBM, the Average Error averaged over all volunteer studies increased to 2.6 mm, which was statistically different from the No-Motion studies. For the case of bends and twists of the torso, which would be expected to challenge the RBM model, the Average Error averaged over all such volunteer studies was 4.9 mm and was again statistically different. Investigations of motion of the arm including just bending at the elbow and leg motion resulted in Average Errors which were not statistically different from the No-Motion studies. However, when shoulder movement was included with arm motion the Average Error was near that of torso bends and twists, and statistically different.

CONCLUSIONS

Use of the RBM model with VTS predictions of heart motion during reconstruction should decrease the extent of artifacts for the types of patient motion studied. The impact of correction would be less for torso bends and twists, and arm motion which includes the shoulders.

Respiratory triggered 4D cone-beam computed tomography: a novel method to reduce imaging dose

PURPOSE

A novel method called respiratory triggered 4D cone-beam computed tomography (RT 4D CBCT) is described whereby imaging dose can be reduced without degrading image quality. RT 4D CBCT utilizes a respiratory signal to trigger projections such that only a single projection is assigned to a given respiratory bin for each breathing cycle. In contrast, commercial 4D CBCT does not actively use the respiratory signal to minimize image dose.

METHODS

To compare RT 4D CBCT with conventional 4D CBCT, 3600 CBCT projections of a thorax phantom were gathered and reconstructed to generate a ground truth CBCT dataset. Simulation pairs of conventional 4D CBCT acquisitions and RT 4D CBCT acquisitions were developed assuming a sinusoidal respiratory signal which governs the selection of projections from the pool of 3600 original projections. The RT 4D CBCT acquisition triggers a single projection when the respiratory signal enters a desired acquisition bin; the conventional acquisition does not use a respiratory trigger and projections are acquired at a constant frequency. Acquisition parameters studied were breathing period, acquisition time, and imager frequency. The performance of RT 4D CBCT using phase based and displacement based sorting was also studied. Image quality was quantified by calculating difference images of the test dataset from the ground truth dataset. Imaging dose was calculated by counting projections.

RESULTS

Using phase based sorting RT 4D CBCT results in 47% less imaging dose on average compared to conventional 4D CBCT. Image quality differences were less than 4% at worst. Using displacement based sorting RT 4D CBCT results in 57% less imaging dose on average, than conventional 4D CBCT methods; however, image quality was 26% worse with RT 4D CBCT.

CONCLUSIONS

Simulation studies have shown that RT 4D CBCT reduces imaging dose while maintaining comparable image quality for phase based 4D CBCT; image quality is degraded for displacement based RT 4D CBCT in its current implementation.

Experimental evaluations of the accuracy of 3D and 4D planning in robotic tracking stereotactic body radiotherapy for lung cancers

PURPOSE

Due to the complexity of 4D target tracking radiotherapy, the accuracy of this treatment strategy should be experimentally validated against established standard 3D technique. This work compared the accuracy of 3D and 4D dose calculations in respiration tracking stereotactic body radiotherapy (SBRT).

METHODS

Using the 4D planning module of the CyberKnife treatment planning system, treatment plans for a moving target and a static off-target cord structure were created on different four-dimensional computed tomography (4D-CT) datasets of a thorax phantom moving in different ranges. The 4D planning system used B-splines deformable image registrations (DIR) to accumulate dose distributions calculated on different breathing geometries, each corresponding to a static 3D-CT image of the 4D-CT dataset, onto a reference image to compose a 4D dose distribution. For each motion, 4D optimization was performed to generate a 4D treatment plan of the moving target. For comparison with standard 3D planning, each 4D plan was copied to the reference end-exhale images and a standard 3D dose calculation was followed. Treatment plans of the off-target structure were first obtained by standard 3D optimization on the end-exhale images. Subsequently, they were applied to recalculate the 4D dose distributions using DIRs. All dose distributions that were initially obtained using the ray-tracing algorithm with equivalent path-length heterogeneity correction (3D EPL and 4D EPL) were recalculated by a Monte Carlo algorithm (3D MC and 4D MC) to further investigate the effects of dose calculation algorithms. The calculated 3D EPL, 3D MC, 4D EPL, and 4D MC dose distributions were compared to measurements by Gafchromic EBT2 films in the axial and coronal planes of the moving target object, and the coronal plane for the static off-target object based on the γ metric at 5%/3mm criteria (γ5%/3mm). Treatment plans were considered acceptable if the percentage of pixels passing γ5%/3mm (Pγ<1) ≥ 90%.

RESULTS

The averaged Pγ<1 values of the 3D EPL, 3D MC, 4D EPL, and 4D MC dose calculation methods for the moving target plans are 95%, 95%, 94%, and 95% for reproducible motion, and 95%, 96%, 94%, and 93% for nonreproducible motion during actual treatment delivery. The overall measured target dose distributions are in better agreement with the 3DMC dose distributions than the 4DMC dose distributions. Conversely, measured dose distributions agree much better with the 4D EPL/MC than the 3D EPL/MC dose distributions in the static off-target structure, resulting in higher Pγ<1 values with 4D EPL/MC (91%) vs 3D EPL (24%) and 3D MC (25%). Systematic changes of target motion reduced the averaged Pγ<1 to 47% and 53% for 4D EPL and 4D MC dose calculations, and 22% for 3D EPL/MC dose calculations in the off-target films.

CONCLUSIONS

In robotic tracking SBRT, 4D treatment planning was found to yield better prediction of the dose distributions in the off-target structure, but not necessarily in the moving target, compared to standard 3D treatment planning, for reproducible and nonreproducible target motion. It is important to ensure on a patient-by-patient basis that the cumulative uncertainty associated with the 4D-CT artifacts, deformable image registration, and motion variability is significantly smaller than the cumulative uncertainty occurred in standard 3D planning in order to make 4D planning a justified option.

A novel technique for markerless, self-sorted 4D-CBCT: feasibility study

PURPOSE

Four-dimensional CBCT (4D-CBCT) imaging in the treatment room can provide verification of moving targets, facilitating the potential for margin reduction and consequent dose escalation. Reconstruction of 4D-CBCT images requires correlation of respiratory phase with projection acquisition, which is often achieved with external surrogate measures of respiration. However, external measures may not be a direct representation of the motion of the internal anatomy and it is therefore the aim of this work to develop a novel technique for markerless, self-sorted 4D-CBCT reconstruction.

METHODS

A novel 4D-CBCT reconstruction technique based on the principles of Fourier transform (FT) theory was investigated for markerless extraction of respiratory phase directly from projection data. In this FT technique, both phase information (FT-phase) and magnitude information (FT-magnitude) were separately implemented in order to discern projections corresponding to peak inspiration, which then facilitated the proceeding sort and bin processes involved in retrospective 4D image reconstruction. In order to quantitatively evaluate the accuracy of the Fourier methods, peak-inspiration projections identified each by FT-phase and FT-magnitude were compared to those manually identified by visual tracking of structures. The average phase difference as assigned by each method vs the manual technique was calculated per projection dataset. The percentage of projections that were assigned within 10% phase of each other was also computed. Both Fourier methods were tested on two phantom datasets, programmed to exhibit sinusoidal respiratory cycles of 2.0 cm in amplitude with respiratory cycle lengths of 3 and 6 s, respectively. Additionally, three sets of patient projections were studied. All of the data were previously acquired at slow-gantry speeds ranging between 0.6°/s and 0.7°/s over a 200° rotation. Ten phase bins with 10% phase windows were selected for 4D-CBCT reconstruction of one phantom and one patient case for visual and quantitative comparison. Line profiles were plotted for the 0% and 50% phase images as reconstructed by the manual technique and each of the Fourier methods.

RESULTS

As compared with the manual technique, the FT-phase method resulted in average phase differences of 1.8% for the phantom with the 3 s respiratory cycle, 3.9% for the phantom with the 6 s respiratory cycle, 2.9% for patient 1, 5.0% for patient 2, and 3.8% for patient 3. For the FT-magnitude method, these numbers were 2.1%, 4.0%, 2.9%, 5.3%, and 3.5%, respectively. The percentage of projections that were assigned within 10% phase by the FT-phase method as compared to the manual technique for the five datasets were 100.0%, 100.0%, 97.6%, 93.4%, and 94.1%, respectively, whereas for the FT-magnitude method these percentages were 98.1%, 92.3%, 98.7%, 87.3%, and 95.7%. Reconstructed 4D phase images for both the phantom and patient case were visually and quantitatively equivalent between each of the Fourier methods vs the manual technique.

CONCLUSIONS

A novel technique employing the basics of Fourier transform theory was investigated and demonstrated to be feasible in achieving markerless, self-sorted 4D-CBCT reconstruction.