Noise-robust breathing-phase estimation on marker-free, ultra low dose X-ray projections for real-time tumor localization via surrogate structures.

PURPOSE: This paper presents a novel strategy for feature-based breathing-phase estimation on ultra low-dose X-ray projections for tumor motion control in radiation therapy. METHODS: Coarse-scaled Curvelet coefficients are identified as motion sensitive but noise-robust features for this purpose. For feature-based breathing-phase estimation, an ensemble strategy with two classifiers is used. This consensus-based estimation substantially increases tracking reliability by rejection of false positives. The algorithm is evaluated on both synthetic and measured phantom data: Monte Carlo simulated ultra low dose projections for a C-arm X-ray and on the basis of 4D-chest-CTs of eight patients on one hand side and real measurements based on a motion phantom. RESULTS: To achieve an accuracy of breathing-phase estimation of more than 95% a fluence between 20 and 400 photons per pixel (open field) is required depending on the patient. Furthermore, the algorithm is evaluated on real ultra low dose projections from an XVI R5.0 system (Elekta AB, Stockholm, Sweden) using an additional lead filter to reduce fluence. The classifiers-consensus-based-gating method estimated the correct position of the test projections in all test cases at a fluence of ∼180 photons per pixel and 92% at a fluence of ∼40 photons per pixel. The deposited dose to patient per image is in the range of nGy. CONCLUSIONS: A novel method is presented for estimation of breathing-phases for real-time tumor localization at ultra low dose both on a simulation and a phantom basis. Its accuracy is comparable to state of the art X-ray based algorithms while the released dose to patients is reduced by two to three orders of magnitude compared to conventional template-based approaches. This allows for continuous motion control during irradiation without the need of external markers.

Reduction of beam hardening artifacts on real C-arm CT data using polychromatic statistical image reconstruction.

PURPOSE: This work aims at the compensation of beam hardening artifacts by the means of an extended three-dimensional polychromatic statistical reconstruction to be applied for flat panel cone-beam CT. METHODS: We implemented this reconstruction technique as being introduced by Elbakri et al. (2002) [1] for a multi-GPU system, assuming the underlying object consists of several well-defined materials. Furthermore, we assume one voxel can only contain an overlap of at most two materials, depending on its density value. Given the X-ray spectrum, the procedure enables to reconstruct the energy-dependent attenuation values of the volume. RESULTS: We evaluated the method by using flat-panel cone-beam CT measurements of structures containing small metal objects and clinical head scan data. In comparison with the water-corrected filtered backprojection, as well as a maximum likelihood reconstruction with a consistency-based beam hardening correction, our method features clearly reduced beam hardening artifacts and a more accurate shape of metal objects. CONCLUSIONS: Our multi-GPU implementation of the polychromatic reconstruction, which does not require any image pre-segmentation, clearly outperforms the standard reconstructions of objects, with respect to beam hardening even in the presence of metal objects inside the volume. However, remaining artifacts, caused mainly by the limited dynamic range of the detector, may have to be addressed in future work.

Prospective trial on telemonitoring of geriatric cancer patients using handheld devices.

PURPOSE AND OBJECTIVE: Randomized trials indicate that electronic or app-based assessment of patient-reported outcomes may improve outcomes in cancer patients. To analyze if an app-based follow-up would be accepted by elderly cancer patients, we conducted a single-center prospective feasibility study (NCT03196050). MATERIALS AND METHODS: Cancer patients (≥60 years) without concurrent uncontrolled severe medical conditions and a Karnofsky performance status (KPS) ≥70 were eligible if they were able to use the smartphone app. The primary endpoint was compliance over 1 year, calculated as patient-specific and study date-specific response rate to questions sent as push notifications; in this interim analysis, we report on 4­month data. Secondary outcomes included a comparison of a subjective health status item (SPHS) with the physician-rated KPS. RESULTS: Out of 225 patients screened, 54 patients agreed to participate and 29 activated the app and participated in the study. The mean age was 66 years (61-78). The individual compliance rate averaged at 58.3% (standard deviation SD = 35%). Daily compliance was 53.3% on average (SD = 10.8%) and declined over time. The average percentage of patients who sent answers at least weekly was 75.0% (SD = 14.8%) and declined from 100% in week 1 to 53.8% in week 17 post-enrollment. Secondary outcomes indicated that questionnaires such as the EORTC-QLQ-C30 are accepted via app and that there is a significant moderate correlation between the SPHS and KPS scores (r = 0.566; p < 0.001). CONCLUSION: Our data indicate that an app-based follow-up incorporating EORTC questionnaires might be possible in highly selected elderly cancer patients with modest compliance rates. Further trials should aim at an increased participation rate.

Ultrafast single breath-hold cone-beam CT lung cancer imaging with faster linac gantry rotation.

PURPOSE: Lung tumors treated with hypo-fractionated deep-inspiration breath-hold stereotactic body radiotherapy benefit from fast imaging and treatment. Single breath-hold cone-beam-CT (CBCT) could reduce motion artifacts and improve treatment precision. Thus, gantry speed was accelerated to 18°/s, limiting acquisition time to 10-20 s. Image quality, dosimetry and registration accuracy were compared with standard-CBCT (3°/s). METHODS AND MATERIALS: For proof-of-concept, image quality was analyzed following customer acceptance tests, CT-dose index measured, and registration accuracy determined with an off-centered ball-bearing-phantom. A lung-tumor patient was simulated with differently shaped tumor-mimicking inlays in a thorax-phantom. Signal-to-noise-ratio, contrast-to-noise-ratio and geometry of the inlays quantified image quality. Dose was measured in representative positions. Registration accuracy was determined with inlays scanned in pre-defined positions. Manual, automatic (clinical software) and objective-automatic (in-house-developed) registration was performed on planning-CT, offsets between results and applied shifts were compared. RESULTS: Image quality of ultrafast-CBCT was adequate for high-contrast areas, despite contrast-reduction of ∼80% due to undersampling. Dose-output was considerably reduced by 60-83% in presented setup; variations are due to gantry-braking characteristics. Registration accuracy was maintained better than 1 mm, mean displacement errors were 0.0 ±â€¯0.2 mm with objective-automatic registration. Ultrafast-CBCT showed no significant registration differences to standard-CBCT. CONCLUSIONS: This study of first tests with faster gantry rotation of 18°/s showed promising results for ultrafast high-contrast lung tumor CBCT imaging within single breath-hold of 10-20 s. Such fast imaging times, in combination with fast treatment delivery, could pave the way for intra-fractional combined imaging and treatment within one breath-hold phase, and thus mitigate residual motion and increase treatment accuracy and patient comfort. Even generally speaking, faster gantry rotation could set a benchmark with immense clinical impact where time matters most: palliative patient care, general reduction in uncertainty, and increase in patient throughput especially important for emerging markets with high patient numbers.

Multiple retrieval case-based reasoning for incomplete datasets.

The performance of case-based reasoning (CBR) depends on an accurate ranking of similar cases in the retrieval phase that affects all subsequent phases and profits from the potential of large databases. Unfortunately, growing databases come along with a rising amount of missing data that reduces the stability of the ranking since incomplete cases cannot be ranked as reliable as complete ones. In context of CBR hardly any work was done so far to rigorously analyze the impact of missing data and solutions to tackle this issue. In particular, a generalized solution which is able to process data under different missingness conditions for different variable types is missing. In this paper we present a multiple retrieval case-based reasoning (MRCBR) framework for incomplete databases that provides a statistically accurate ranking for similar cases. It unifies the advantages of multiple imputation and CBR while it preserves both the data distribution and database structure. Built as generalized CBR system, MRCBR was optimized and tested for medical decision support but can be extended to any CBR requirement as well. It is suitable for numerical and categorical variables and all sorts of missingness conditions. The approach was compared to eight competing methods applicable to handle incomplete databases in context of CBR. The comparison to the true ranking was based on two various error measures. In the evaluation we tested four representative scenarios that considered different conditions for missing data analysis. The outcome for every method in each scenario resulted in 200 miscellaneous setups. MRCBR outperforms all compared CBR methods in presence of missing data and shows reliable and stable results in every scenario. Especially with larger databases and rising number of incomplete variables it enlarges its lead to all other methods. Our study demonstrates that missing data must not be ignored when a correct CBR outcome is required.

Intra-breath-hold residual motion of image-guided DIBH liver-SBRT: An estimation by ultrasound-based monitoring correlated with diaphragm position in CBCT.

BACKGROUND AND PURPOSE: Craniocaudal motion during image-guided abdominal SBRT can be reduced by computer-controlled deep-inspiratory-breath-hold (DIBH). However, a residual motion can occur in the DIBH-phases which can only be detected with intrafractional real-time-monitoring. We assessed the intra-breath-hold residual motion of DIBH and compared residual motion of target structures during DIBH detected by ultrasound (US). US data were compared with residual motion of the diaphragm-dome (DD) detected in the DIBH-CBCT-projections. PATIENTS AND METHODS: US-based monitoring was performed with an experimental US-system simultaneously to DIBH-CBCT acquisition. A total of 706 DIBHs during SBRT-treatments of metastatic lesions (liver, spleen, adrenal) of various primaries were registered in 13 patients. Residual motion of the target structure was documented with US during each DIBH. Motion of the DD was determined by comparison to a reference phantom-scan taking the individual geometrical setting at a given projection angle into account. Residual motion data detected by US were correlated to those of the DD (DIBH-CBCT-projection). RESULTS: US-based monitoring could be performed in all cases and was well tolerated by all patients. Additional time for daily US-based setup required 8 ±â€¯4 min. 385 DIBHs of 706 could be analyzed. In 59% of all DIBHs, residual motion was below 2 mm. In 36%, residual motion of 2-5 mm and in 4% of 5-8 mm was observed. Only 1% of all DIBHs and 0.16% of all readings revealed a residual motion of >8 mm during DIBH. For DIBHs with a residual motion over 2 mm, 137 of 156 CBCT-to-US curves had a parallel residual motion and showed a statistical correlation. DISCUSSION AND CONCLUSION: Soft-tissue monitoring with ultrasound is a fast real-time method without additional radiation exposure. Computer-controlled DIBH has a residual motion of <5 mm in >95% which is in line with the published intra-breath-hold-precision. Larger intrafractional deviations can be avoided if the beam is stopped at an US-defined threshold.

An offline technique to evaluate residual motion of the diaphragm during deep inspiratory breath-hold from cone-beam CT datasets.

PURPOSE: In radiation therapy, the computer-assisted deep inspiration breath-hold (DIBH) technique is one approach to deal with respiratory motion of tumors in the lung, liver, or upper abdomen. However, inter- and intra-breath-hold deviations from an optimal static tumor position might occur. A novel method is presented to noninvasively measure the diaphragm position and thus estimate its residual deviation (as surrogate for the tumor position) based on cone-beam computed tomography (CBCT) projection data using active breathing control during acquisition. METHODS: The diaphragm dome (DD) position relative to the isocenter of a linear accelerator is known from the static (DIBH) planning CT. A ball-bearing phantom (BB) is placed at this position, a CBCT dataset is acquired, and in each projection the position of the projected BB is determined automatically based on thresholding. The position of the DD is determined manually in CBCT projections of a patient. The distance between DD and BB (ideal static setting) in craniocaudal direction is calculated for a given angle based on the distance in the projection plane and the relative position of the BB referring to the source and the detector. An angle-dependent correction factor is introduced which takes this geometrical setting into account. The accuracy of the method is assessed. RESULTS: The method allows a CBCT projection-based estimation of the deviation between the DD and its optimal position as defined in the planning CT, i.e., the residual motion of the DD can be assessed. The error of this estimation is 2.2 mm in craniocaudal direction. CONCLUSIONS: The developed method allows an offline estimation of the inspiration depth (inter- and intra-breath-hold) over time. It will be useful as a reference for comparison to other methods of residual motion estimation, e.g., surface scanning.

Automated ultrafast kilovoltage-megavoltage cone-beam CT for image guided radiotherapy of lung cancer: System description and real-time results.

PURPOSE: To establish a fully automated kV-MV CBCT imaging method on a clinical linear accelerator that allows image acquisition of thoracic targets for patient positioning within one breath-hold (∼15s) under realistic clinical conditions. METHODS AND MATERIALS: Our previously developed FPGA-based hardware unit which allows synchronized kV-MV CBCT projection acquisition is connected to a clinical linear accelerator system via a multi-pin switch; i.e. either kV-MV imaging or conventional clinical mode can be selected. An application program was developed to control the relevant linac parameters automatically and to manage the MV detector readout as well as the gantry angle capture for each MV projection. The kV projections are acquired with the conventional CBCT system. GPU-accelerated filtered backprojection is performed separately for both data sets. After appropriate grayscale normalization both modalities are combined and the final kV-MV volume is re-imported in the CBCT system to enable image matching. To demonstrate adequate geometrical accuracy of the novel imaging system the Penta-Guide phantom QA procedure is performed. Furthermore, a human plastinate and different tumor shapes in a thorax phantom are scanned. Diameters of the known tumor shapes are measured in the kV-MV reconstruction. RESULTS: An automated kV-MV CBCT workflow was successfully established in a clinical environment. The overall procedure, from starting the data acquisition until the reconstructed volume is available for registration, requires ∼90s including 17s acquisition time for 100° rotation. It is very simple and allows target positioning in the same way as for conventional CBCT. Registration accuracy of the QA phantom is within ±1mm. The average deviation from the known tumor dimensions measured in the thorax phantom was 0.7mm which corresponds to an improvement of 36% compared to our previous kV-MV imaging system. CONCLUSIONS: Due to automation the kV-MV CBCT workflow is speeded up by a factor of >10 compared to the manual approach. Thus, the system allows a simple, fast and reliable imaging procedure and fulfills all requirements to be successfully introduced into the clinical workflow now, enabling single-breath-hold volume imaging.

Phantom-based evaluation of dose exposure of ultrafast combined kV-MV-CBCT towards clinical implementation for IGRT of lung cancer.

PURPOSE: Combined ultrafast 90°+90° kV-MV-CBCT within single breath-hold of 15s has high clinical potential for accelerating imaging for lung cancer patients treated with deep inspiration breath-hold (DIBH). For clinical feasibility of kV-MV-CBCT, dose exposure has to be small compared to prescribed dose. In this study, kV-MV dose output is evaluated and compared to clinically-established kV-CBCT. METHODS: Accurate dose calibration was performed for kV and MV energy; beam quality was determined. For direct comparison of MV and kV dose output, relative biological effectiveness (RBE) was considered. CT dose index (CTDI) was determined and measurements in various representative locations of an inhomogeneous thorax phantom were performed to simulate the patient situation. RESULTS: A measured dose of 20.5mGE (Gray-equivalent) in the target region was comparable to kV-CBCT (31.2mGy for widely-used, and 9.1mGy for latest available preset), whereas kV-MV spared healthy tissue and reduced dose to 6.6mGE (30%) due to asymmetric dose distribution. The measured weighted CTDI of 12mGE for kV-MV lay in between both clinical presets. CONCLUSIONS: Dosimetric properties were in agreement with established imaging techniques, whereas exposure to healthy tissue was reduced. By reducing the imaging time to a single breath-hold of 15s, ultrafast combined kV-MV CBCT shortens patient time at the treatment couch and thus improves patient comfort. It is therefore usable for imaging of hypofractionated lung DIBH patients.

Towards clinical implementation of ultrafast combined kV-MV CBCT for IGRT of lung cancer : Evaluation of registration accuracy based on phantom study.

PURPOSE: Combined kV-MV cone-beam CT (CBCT) is a promising approach to accelerate imaging for patients with lung tumors treated with deep inspiration breath-hold. During a single breath-hold (15 s), a 3D kV-MV CBCT can be acquired, thus minimizing motion artifacts and increasing patient comfort. Prior to clinical implementation, positioning accuracy was evaluated and compared to clinically established imaging techniques. METHODS AND MATERIALS: An inhomogeneous thorax phantom with four tumor-mimicking inlays was imaged in 10 predefined positions and registered to a planning CT. Novel kV-MV CBCT imaging (90° arc) was compared to clinically established kV-chest CBCT (360°) as well as nonclinical kV-CBCT and low-dose MV-CBCT (each 180°). Manual registration, automatic registration provided by the manufacturer and an additional in-house developed manufacturer-independent framework based on the MATLAB registration toolkit were applied. RESULTS: Systematic setup error was reduced to 0.05 mm by high-precision phantom positioning with optical tracking. Stochastic mean displacement errors were 0.5 ± 0.3 mm in right-left, 0.4 ± 0.4 mm in anteroposterior and 0.0 ± 0.4 mm in craniocaudal directions for kV-MV CBCT with manual registration (maximum errors of no more than 1.4 mm). Clinical kV-chest CBCT resulted in mean errors of 0.2 mm (other modalities: 0.4-0.8 mm). Similar results were achieved with both automatic registration methods. CONCLUSION: The comparison study of repositioning accuracy between novel kV-MV CBCT and clinically established volume imaging demonstrated that registration accuracy is maintained below 1 mm. Since imaging time is reduced to one breath-hold, kV-MV CBCT is ideal for image guidance, e.g., in lung stereotactic ablative radiotherapy.

Deep Inspiration Breath Hold-Based Radiation Therapy: A Clinical Review.

Several recent developments in linear accelerator-based radiation therapy (RT) such as fast multileaf collimators, accelerated intensity modulation paradigms like volumeric modulated arc therapy and flattening filter-free (FFF) high-dose-rate therapy have dramatically shortened the duration of treatment fractions. Deliverable photon dose distributions have approached physical complexity limits as a consequence of precise dose calculation algorithms and online 3-dimensional image guided patient positioning (image guided RT). Simultaneously, beam quality and treatment speed have continuously been improved in particle beam therapy, especially for scanned particle beams. Applying complex treatment plans with steep dose gradients requires strategies to mitigate and compensate for motion effects in general, particularly breathing motion. Intrafractional breathing-related motion results in uncertainties in dose delivery and thus in target coverage. As a consequence, generous margins have been used, which, in turn, increases exposure to organs at risk. Particle therapy, particularly with scanned beams, poses additional problems such as interplay effects and range uncertainties. Among advanced strategies to compensate breathing motion such as beam gating and tracking, deep inspiration breath hold (DIBH) gating is particularly advantageous in several respects, not only for hypofractionated, high single-dose stereotactic body RT of lung, liver, and upper abdominal lesions but also for normofractionated treatment of thoracic tumors such as lung cancer, mediastinal lymphomas, and breast cancer. This review provides an in-depth discussion of the rationale and technical implementation of DIBH gating for hypofractionated and normofractionated RT of intrathoracic and upper abdominal tumors in photon and proton RT.

Multiple breath-hold CBCT for online image guided radiotherapy of lung tumors: simulation with a dynamic phantom and first patient data.

BACKGROUND AND PURPOSE: Computer controlled breath-hold effectively reduces organ motion for image-guided precision radiotherapy of lung tumors. However, the acquisition time of 3D cone-beam-CT (CBCT) exceeds maximum breath-hold times. We have developed an approach enabling online verification using CBCT image acquisition with ABC®-based breath-hold. METHODS: Patient CBCT images were acquired with ABC®-based repeat breath-hold. The clinical situation was also simulated with a Motion Phantom. Reconstruction of patient and phantom images with selection of free-breathing and breath-hold projections only was performed. RESULTS: CBCT-imaging in repeat breath-hold resulted in a precisely spherical appearance of a tumor-mimicking structure in the phantom. A faint "ghost" structure (free-breathing phases) can be clearly discriminated. Mean percentage of patient breath-hold time was 66%. Reconstruction based on free-breathing-only shows blurring of both tumor and diaphragm, reconstruction based on breath-hold projections only resulted in sharp contours of the same structures. From the phantom experiments, a maximal repositioning error of 1mm in each direction can be estimated. DISCUSSION AND CONCLUSION: CBCT during repetitive breath hold provides reliable soft-tissue-based positioning. Fast 3D-imaging during one breath-hold is currently under development and has the potential to accelerate clinical linac-based volume imaging.

A phenomenological kV beam model for cone-beam imaging.

A phenomenological kV beam model was developed to address attenuation and scatter in radiographic images for the purpose of cone-beam imaging. Characterization of a kV beam in terms of the minimal number of parameters and calculation of attenuation and scatter in radiographs of scanned objects are the main applications of this model. Model parameters are derived from radiographs of homogeneous solid water phantoms for various depths and field sizes. The response of the cone-beam detector to kV beams is factorized into different contributions such as output factor, tissue-air ratio and off-axis ratio, with each contribution having an analytical representation. The formulas which are used to characterize the beam model in uniform phantoms are then extended to arbitrary objects using the concept of the water-equivalent pathlength. A weighted sum of three Gaussians in each direction models the dose deposition kernel. Detector response arising from the first Gaussian term can be interpreted as the primary signal while the second and third Gaussians constitute short- and long-range scatter. The model is then applied to predict the primary and scatter signals for arbitrary objects. A technique of scatter removal from the measured radiographs is investigated. The model accurately predicts detector response of varying-thickness phantoms such as multi-step and cylindrical phantoms. The scatter contributes over 90% to the total signal for 20 cm thick phantoms. The calculated scatter-to-primary ratio as a function of spatial coordinates agrees with Monte Carlo studies reported in the literature. Water-equivalent thickness related to primary and scatter contributions calculated from an analysis of radiographs results in an improved calibration technique suitable for CB-CT reconstruction. The kV beam model and the associated theoretical formulations can be utilized to characterize any kV beam line; however, for the specific study the OBI system (Varian) was used to obtain experimental radiographs.

Fast kilovoltage/megavoltage (kVMV) breathhold cone-beam CT for image-guided radiotherapy of lung cancer.

Long image acquisition times of 60-120 s for cone-beam CT (CBCT) limit the number of patients with lung cancer who can undergo volume image guidance under breathhold. We developed a low-dose dual-energy kilovoltage-megavoltage-cone-beam CT (kVMV-CBCT) based on a clinical treatment unit reducing imaging time to < or =15 s. Simultaneous kVMV-imaging was achieved by dedicated synchronization hardware controlling the output of the linear accelerator (linac) based on detector panel readout signals, preventing imaging artifacts from interference of the linac's MV-irradiation and panel readouts. Optimization was performed to minimize the imaging dose. Single MV-projections, reconstructed MV-CBCT images and images of simultaneous 90 degrees kV- and 90 degrees MV-CBCT (180 degrees kVMV-CBCT) were acquired with different parameters. Image quality and imaging dose were evaluated and compared to kV-imaging. Hardware-based kVMV synchronization resulted in artifact-free projections. A combined 180 degrees kVMV-CBCT scan with a total MV-dose of 5 monitor units was acquired in 15 s and with sufficient image quality. The resolution was 5-6 line pairs cm(-1) (Catphan phantom). The combined kVMV-scan dose was equivalent to a kV-radiation scan dose of approximately 33 mGy. kVMV-CBCT based on a standard linac is promising and can provide ultra-fast online volume image guidance with low imaging dose and sufficient image quality for fast and accurate patient positioning for patients with lung cancer under breathhold.

Breath-hold target localization with simultaneous kilovoltage/megavoltage cone-beam computed tomography and fast reconstruction.

PURPOSE: Hypofractionated high-dose radiotherapy for small lung tumors has typically been based on stereotaxy. Cone-beam computed tomography and breath-hold techniques have provided a noninvasive basis for precise cranial and extracranial patient positioning. The cone-beam computed tomography acquisition time of 60 s, however, is beyond the breath-hold capacity of patients, resulting in respiratory motion artifacts. By combining megavoltage (MV) and kilovoltage (kV) photon sources (mounted perpendicularly on the linear accelerator) and accelerating the gantry rotation to the allowed limit, the data acquisition time could be reduced to 15 s. METHODS AND MATERIALS: An Elekta Synergy 6-MV linear accelerator, with iViewGT as the MV- and XVI as the kV-imaging device, was used with a Catphan phantom and an anthropomorphic thorax phantom. Both image sources performed continuous image acquisition, passing an angle interval of 90° within 15 s. For reconstruction, filtered back projection on a graphics processor unit was used. It reconstructed 100 projections acquired to a 512 × 512 × 512 volume within 6 s. RESULTS: The resolution in the Catphan phantom (CTP528 high-resolution module) was 3 lines/cm. The spatial accuracy was within 2-3 mm. The diameters of different tumor shapes in the thorax phantom were determined within an accuracy of 1.6 mm. The signal-to-noise ratio was 68% less than that with a 180°-kV scan. The dose generated to acquire the MV frames accumulated to 82.5 mGy, and the kV contribution was <6 mGy. CONCLUSION: The present results have shown that fast breath-hold, on-line volume imaging with a linear accelerator using simultaneous kV-MV cone-beam computed tomography is promising and can potentially be used for image-guided radiotherapy for lung cancer patients in the near future.