Audio-visual biofeedback does not improve the reliability of target delineation using maximum intensity projection in 4-dimensional computed tomography radiation therapy planning.

PURPOSE: To investigate whether coaching patients' breathing would improve the match between ITVMIP (internal target volume generated by contouring in the maximum intensity projection scan) and ITV10 (generated by combining the gross tumor volumes contoured in 10 phases of a 4-dimensional CT [4DCT] scan). METHODS AND MATERIALS: Eight patients with a thoracic tumor and 5 patients with an abdominal tumor were included in an institutional review board-approved prospective study. Patients underwent 3 4DCT scans with: (1) free breathing (FB); (2) coaching using audio-visual (AV) biofeedback via the Real-Time Position Management system; and (3) coaching via a spirometer system (Active Breathing Coordinator or ABC). One physician contoured all scans to generate the ITV10 and ITVMIP. The match between ITVMIP and ITV10 was quantitatively assessed with volume ratio, centroid distance, root mean squared distance, and overlap/Dice coefficient. We investigated whether coaching (AV or ABC) or uniform expansions (1, 2, 3, or 5 mm) of ITVMIP improved the match. RESULTS: Although both AV and ABC coaching techniques improved frequency reproducibility and ABC improved displacement regularity, neither improved the match between ITVMIP and ITV10 over FB. On average, ITVMIP underestimated ITV10 by 19%, 19%, and 21%, with centroid distance of 1.9, 2.3, and 1.7 mm and Dice coefficient of 0.87, 0.86, and 0.88 for FB, AV, and ABC, respectively. Separate analyses indicated a better match for lung cancers or tumors not adjacent to high-intensity tissues. Uniform expansions of ITVMIP did not correct for the mismatch between ITVMIP and ITV10. CONCLUSIONS: In this pilot study, audio-visual biofeedback did not improve the match between ITVMIP and ITV10. In general, ITVMIP should be limited to lung cancers, and modification of ITVMIP in each phase of the 4DCT data set is recommended.

Maintaining tumor targeting accuracy in real-time motion compensation systems for respiration-induced tumor motion.

PURPOSE: To determine how best to time respiratory surrogate-based tumor motion model updates by comparing a novel technique based on external measurements alone to three direct measurement methods. METHODS: Concurrently measured tumor and respiratory surrogate positions from 166 treatment fractions for lung or pancreas lesions were analyzed. Partial-least-squares regression models of tumor position from marker motion were created from the first six measurements in each dataset. Successive tumor localizations were obtained at a rate of once per minute on average. Model updates were timed according to four methods: never, respiratory surrogate-based (when metrics based on respiratory surrogate measurements exceeded confidence limits), error-based (when localization error ≥ 3 mm), and always (approximately once per minute). RESULTS: Radial tumor displacement prediction errors (mean ± standard deviation) for the four schema described above were 2.4 ± 1.2, 1.9 ± 0.9, 1.9 ± 0.8, and 1.7 ± 0.8 mm, respectively. The never-update error was significantly larger than errors of the other methods. Mean update counts over 20 min were 0, 4, 9, and 24, respectively. CONCLUSIONS: The same improvement in tumor localization accuracy could be achieved through any of the three update methods, but significantly fewer updates were required when the respiratory surrogate method was utilized. This study establishes the feasibility of timing image acquisitions for updating respiratory surrogate models without direct tumor localization.

Online monitoring and error detection of real-time tumor displacement prediction accuracy using control limits on respiratory surrogate statistics.

PURPOSE: To evaluate Hotelling's T(2) statistic and the input variable squared prediction error (Q((X))) for detecting large respiratory surrogate-based tumor displacement prediction errors without directly measuring the tumor's position. METHODS: Tumor and external marker positions from a database of 188 Cyberknife Synchrony™ lung, liver, and pancreas treatment fractions were analyzed. The first ten measurements of tumor position in each fraction were used to create fraction-specific models of tumor displacement using external surrogates as input; the models were used to predict tumor position from subsequent external marker measurements. A partial least squares (PLS) model with four scores was developed for each fraction to determine T(2) and Q((X)) confidence limits based on the first ten measurements in a fraction. The T(2) and Q((X)) statistics were then calculated for every set of external marker measurements. Correlations between model error and both T(2) and Q((X)) were determined. Receiver operating characteristic analysis was applied to evaluate sensitivities and specificities of T(2), Q((X)), and T(2)∪Q((X)) for predicting real-time tumor localization errors >3 mm over a range of T(2) and Q((X)) confidence limits. RESULTS: Sensitivity and specificity of detecting errors >3 mm varied with confidence limit selection. At 95% sensitivity, T(2)∪Q((X)) specificity was 15%, 2% higher than either T(2) or Q((X)) alone. The mean time to alarm for T(2)∪Q((X)) at 95% sensitivity was 5.3 min but varied with a standard deviation of 8.2 min. Results did not differ significantly by tumor site. CONCLUSIONS: The results of this study establish the feasibility of respiratory surrogate-based online monitoring of real-time respiration-induced tumor motion model accuracy for lung, liver, and pancreas tumors. The T(2) and Q((X)) statistics were able to indicate whether inferential model errors exceeded 3 mm with high sensitivity. Modest improvements in specificity were achieved by combining T(2) and Q((X)) results.

Mitigating errors in external respiratory surrogate-based models of tumor position.

PURPOSE: To investigate the effect of tumor site, measurement precision, tumor-surrogate correlation, training data selection, model design, and interpatient and interfraction variations on the accuracy of external marker-based models of tumor position. METHODS AND MATERIALS: Cyberknife Synchrony system log files comprising synchronously acquired positions of external markers and the tumor from 167 treatment fractions were analyzed. The accuracy of Synchrony, ordinary-least-squares regression, and partial-least-squares regression models for predicting the tumor position from the external markers was evaluated. The quantity and timing of the data used to build the predictive model were varied. The effects of tumor-surrogate correlation and the precision in both the tumor and the external surrogate position measurements were explored by adding noise to the data. RESULTS: The tumor position prediction errors increased during the duration of a fraction. Increasing the training data quantities did not always lead to more accurate models. Adding uncorrelated noise to the external marker-based inputs degraded the tumor-surrogate correlation models by 16% for partial-least-squares and 57% for ordinary-least-squares. External marker and tumor position measurement errors led to tumor position prediction changes 0.3-3.6 times the magnitude of the measurement errors, varying widely with model algorithm. The tumor position prediction errors were significantly associated with the patient index but not with the fraction index or tumor site. Partial-least-squares was as accurate as Synchrony and more accurate than ordinary-least-squares. CONCLUSIONS: The accuracy of surrogate-based inferential models of tumor position was affected by all the investigated factors, except for the tumor site and fraction index.

Incidence of changes in respiration-induced tumor motion and its relationship with respiratory surrogates during individual treatment fractions.

PURPOSE: To determine how frequently (1) tumor motion and (2) the spatial relationship between tumor and respiratory surrogate markers change during a treatment fraction in lung and pancreas cancer patients. METHODS AND MATERIALS: A Cyberknife Synchrony system radiographically localized the tumor and simultaneously tracked three respiratory surrogate markers fixed to a form-fitting vest. Data in 55 lung and 29 pancreas fractions were divided into successive 10-min blocks. Mean tumor positions and tumor position distributions were compared across 10-min blocks of data. Treatment margins were calculated from both 10 and 30 min of data. Partial least squares (PLS) regression models of tumor positions as a function of external surrogate marker positions were created from the first 10 min of data in each fraction; the incidence of significant PLS model degradation was used to assess changes in the spatial relationship between tumors and surrogate markers. RESULTS: The absolute change in mean tumor position from first to third 10-min blocks was >5 mm in 13% and 7% of lung and pancreas cases, respectively. Superior-inferior and medial-lateral differences in mean tumor position were significantly associated with the lobe of lung. In 61% and 54% of lung and pancreas fractions, respectively, margins calculated from 30 min of data were larger than margins calculated from 10 min of data. The change in treatment margin magnitude for superior-inferior motion was >1 mm in 42% of lung and 45% of pancreas fractions. Significantly increasing tumor position prediction model error (mean ± standard deviation rates of change of 1.6 ± 2.5 mm per 10 min) over 30 min indicated tumor-surrogate relationship changes in 63% of fractions. CONCLUSIONS: Both tumor motion and the relationship between tumor and respiratory surrogate displacements change in most treatment fractions for patient in-room time of 30 min.

On the accuracy of a moving average algorithm for target tracking during radiation therapy treatment delivery.

Real-time tumor targeting involves the continuous realignment of the radiation beam with the tumor. Real-time tumor targeting offers several advantages such as improved accuracy of tumor treatment and reduced dose to surrounding tissue. Current limitations to this technique include mechanical motion constraints. The purpose of this study was to investigate an alternative treatment scenario using a moving average algorithm. The algorithm, using a suitable averaging period, accounts for variations in the average tumor position, but respiratory induced target position variations about this average are ignored during delivery and can be treated as a random error during planning. In order to test the method a comparison between five different treatment techniques was performed: (1) moving average algorithm, (2) real-time motion tracking, (3) respiration motion gating (at both inhale and exhale), (4) moving average gating (at both inhale and exhale) and (5) static beam delivery. Two data sets were used for the purpose of this analysis: (a) external respiratory-motion traces using different coaching techniques included 331 respiration motion traces from 24 lung-cancer patients acquired using three different breathing types [free breathing (FB), audio coaching (A) and audio-visual biofeedback (AV)]; (b) 3D tumor motion included implanted fiducial motion data for over 160 treatment fractions for 46 thoracic and abdominal cancer patients obtained from the Cyberknife Synchrony. The metrics used for comparison were the group systematic error (M), the standard deviation (SD) of the systematic error (sigma) and the root mean square of the random error (sigma). Margins were calculated using the formula by Stroom et al. [Int. J. Radiat. Oncol., Biol., Phys. 43(4), 905-919 (1999)]: 2sigma + 0.7sigma. The resultant calculations for implanted fiducial motion traces (all values in cm) show that M and sigma are negligible for moving average algorithm, moving average gating, and real-time tracking (i.e., M and sigma = 0 cm) compared to static beam (M = 0.02 cm and sigma = 0.16 cm) or gated beam delivery (M = -0.05 and 0.16 cm at both exhale and inhale, respectively, and sigma = 0.17 and 0.26 cm at both exhale and inhale, respectively). Moving average algorithm sigma = 0.22 cm has a slightly lower random error than static beam delivery sigma = 0.24 cm, though gating, moving average gating, and real-time tracking have much lower random error values for implanted fiducial motion. Similar trends were also observed for the results using the external respiratory motion data. Moving average algorithm delivery significantly reduces M and sigma compared with static beam delivery. The moving average algorithm removes the nonstationary part of the respiration motion which is also achieved by AV, and thus the addition of the moving average algorithm shows little improvement with AV. Overall, a moving average algorithm shows margin reduction compared with gating and static beam delivery, and may have some mechanical advantages over real-time tracking when the beam is aligned with the target and patient compliance advantages over real-time tracking when the target is aligned to the beam.

Development and preliminary evaluation of a prototype audiovisual biofeedback device incorporating a patient-specific guiding waveform.

The aim of this research was to investigate the effectiveness of a novel audio-visual biofeedback respiratory training tool to reduce respiratory irregularity. The audiovisual biofeedback system acquires sample respiratory waveforms of a particular patient and computes a patient-specific waveform to guide the patient's subsequent breathing. Two visual feedback models with different displays and cognitive loads were investigated: a bar model and a wave model. The audio instructions were ascending/descending musical tones played at inhale and exhale respectively to assist in maintaining the breathing period. Free-breathing, bar model and wave model training was performed on ten volunteers for 5 min for three repeat sessions. A total of 90 respiratory waveforms were acquired. It was found that the bar model was superior to free breathing with overall rms displacement variations of 0.10 and 0.16 cm, respectively, and rms period variations of 0.77 and 0.33 s, respectively. The wave model was superior to the bar model and free breathing for all volunteers, with an overall rms displacement of 0.08 cm and rms periods of 0.2 s. The reduction in the displacement and period variations for the bar model compared with free breathing was statistically significant (p = 0.005 and 0.002, respectively); the wave model was significantly better than the bar model (p = 0.006 and 0.005, respectively). Audiovisual biofeedback with a patient-specific guiding waveform significantly reduces variations in breathing. The wave model approach reduces cycle-to-cycle variations in displacement by greater than 50% and variations in period by over 70% compared with free breathing. The planned application of this device is anatomic and functional imaging procedures and radiation therapy delivery.

Audio-visual biofeedback for respiratory-gated radiotherapy: impact of audio instruction and audio-visual biofeedback on respiratory-gated radiotherapy.

PURPOSE: Respiratory gating is a commercially available technology for reducing the deleterious effects of motion during imaging and treatment. The efficacy of gating is dependent on the reproducibility within and between respiratory cycles during imaging and treatment. The aim of this study was to determine whether audio-visual biofeedback can improve respiratory reproducibility by decreasing residual motion and therefore increasing the accuracy of gated radiotherapy. METHODS AND MATERIALS: A total of 331 respiratory traces were collected from 24 lung cancer patients. The protocol consisted of five breathing training sessions spaced about a week apart. Within each session the patients initially breathed without any instruction (free breathing), with audio instructions and with audio-visual biofeedback. Residual motion was quantified by the standard deviation of the respiratory signal within the gating window. RESULTS: Audio-visual biofeedback significantly reduced residual motion compared with free breathing and audio instruction. Displacement-based gating has lower residual motion than phase-based gating. Little reduction in residual motion was found for duty cycles less than 30%; for duty cycles above 50% there was a sharp increase in residual motion. CONCLUSIONS: The efficiency and reproducibility of gating can be improved by: incorporating audio-visual biofeedback, using a 30-50% duty cycle, gating during exhalation, and using displacement-based gating.

The clinical implementation of respiratory-gated intensity-modulated radiotherapy.

The clinical use of respiratory-gated radiotherapy and the application of intensity-modulated radiotherapy (IMRT) are 2 relatively new innovations to the treatment of lung cancer. Respiratory gating can reduce the deleterious effects of intrafraction motion, and IMRT can concurrently increase tumor dose homogeneity and reduce dose to critical structures including the lungs, spinal cord, esophagus, and heart. The aim of this work is to describe the clinical implementation of respiratory-gated IMRT for the treatment of non-small cell lung cancer. Documented clinical procedures were developed to include a tumor motion study, gated CT imaging, IMRT treatment planning, and gated IMRT delivery. Treatment planning procedures for respiratory-gated IMRT including beam arrangements and dose-volume constraints were developed. Quality assurance procedures were designed to quantify both the dosimetric and positional accuracy of respiratory-gated IMRT, including film dosimetry dose measurements and Monte Carlo dose calculations for verification and validation of individual patient treatments. Respiratory-gated IMRT is accepted by both treatment staff and patients. The dosimetric and positional quality assurance test results indicate that respiratory-gated IMRT can be delivered accurately. If carefully implemented, respiratory-gated IMRT is a practical alternative to conventional thoracic radiotherapy. For mobile tumors, respiratory-gated radiotherapy is used as the standard of care at our institution. Due to the increased workload, the choice of IMRT is taken on a case-by-case basis, with approximately half of the non-small cell lung cancer patients receiving respiratory-gated IMRT. We are currently evaluating whether superior tumor coverage and limited normal tissue dosing will lead to improvements in local control and survival in non-small cell lung cancer.