ASTRA: Atomic Surface Transformations for Radiotherapy Quality Assurance.

Treatment for glioblastoma, an aggressive brain tumour usually relies on radiotherapy. This involves planning how to achieve the desired radiation dose distribution, which is known as treatment planning. Treatment planning is impacted by human errors, inter-expert variability in segmenting (or outlining) the tumor target and organs-at-risk, and differences in segmentation protocols. Erroneous segmentations translate to erroneous dose distributions, and hence sub-optimal clinical outcomes. Reviewing segmentations is time-intensive, significantly reduces the efficiency of radiation oncology teams, and hence restricts timely radiotherapy interventions to limit tumor growth. Moreover, to date, radiation oncologists review and correct segmentations without information on how potential corrections might affect radiation dose distributions, leading to an ineffective and suboptimal segmentation correction workflow. In this paper, we introduce an automated deep-learning based method: atomic surface transformations for radiotherapy quality assurance (ASTRA), that predicts the potential impact of local segmentation variations on radiotherapy dose predictions, thereby serving as an effective dose-aware sensitivity map of segmentation variations. On a dataset of 100 glioblastoma patients, we show how the proposed approach enables assessment and visualization of areas of organs-at-risk being most susceptible to dose changes, providing clinicians with a dose-informed mechanism to review and correct segmentations for radiation therapy planning. These initial results suggest strong potential for employing such methods within a broader automated quality assurance system in the radiotherapy planning workflow. Code to reproduce this is available at https://github.com/amithjkamath/astraClinical Relevance: ASTRA shows promise in indicating what regions of the OARs are more likely to impact the distribution of radiation dose.

Deep-Learning-Based Dose Predictor for Glioblastoma-Assessing the Sensitivity and Robustness for Dose Awareness in Contouring.

External beam radiation therapy requires a sophisticated and laborious planning procedure. To improve the efficiency and quality of this procedure, machine-learning models that predict these dose distributions were introduced. The most recent dose prediction models are based on deep-learning architectures called 3D U-Nets that give good approximations of the dose in 3D almost instantly. Our purpose was to train such a 3D dose prediction model for glioblastoma VMAT treatment and test its robustness and sensitivity for the purpose of quality assurance of automatic contouring. From a cohort of 125 glioblastoma (GBM) patients, VMAT plans were created according to a clinical protocol. The initial model was trained on a cascaded 3D U-Net. A total of 60 cases were used for training, 15 for validation and 20 for testing. The prediction model was tested for sensitivity to dose changes when subject to realistic contour variations. Additionally, the model was tested for robustness by exposing it to a worst-case test set containing out-of-distribution cases. The initially trained prediction model had a dose score of 0.94 Gy and a mean DVH (dose volume histograms) score for all structures of 1.95 Gy. In terms of sensitivity, the model was able to predict the dose changes that occurred due to the contour variations with a mean error of 1.38 Gy. We obtained a 3D VMAT dose prediction model for GBM with limited data, providing good sensitivity to realistic contour variations. We tested and improved the model's robustness by targeted updates to the training set, making it a useful technique for introducing dose awareness in the contouring evaluation and quality assurance process.

PyRaDiSe: A Python package for DICOM-RT-based auto-segmentation pipeline construction and DICOM-RT data conversion.

BACKGROUND AND OBJECTIVE: Despite fast evolution cycles in deep learning methodologies for medical imaging in radiotherapy, auto-segmentation solutions rarely run in clinics due to the lack of open-source frameworks feasible for processing DICOM RT Structure Sets. Besides this shortage, available open-source DICOM RT Structure Set converters rely exclusively on 2D reconstruction approaches leading to pixelated contours with potentially low acceptance by healthcare professionals. PyRaDiSe, an open-source, deep learning framework independent Python package, addresses these issues by providing a framework for building auto-segmentation solutions feasible to operate directly on DICOM data. In addition, PyRaDiSe provides profound DICOM RT Structure Set conversion and processing capabilities; thus, it applies also to auto-segmentation-related tasks, such as dataset construction for deep learning model training. METHODS: The PyRaDiSe package follows a holistic approach and provides DICOM data handling, deep learning model inference, pre-processing, and post-processing functionalities. The DICOM data handling allows for highly automated and flexible handling of DICOM image series, DICOM RT Structure Sets, and DICOM registrations, including 2D-based and 3D-based conversion from and to DICOM RT Structure Sets. For deep learning model inference, extending given skeleton classes is straightforwardly achieved, allowing for employing any deep learning framework. Furthermore, a profound set of pre-processing and post-processing routines is included that incorporate partial invertibility for restoring spatial properties, such as image origin or orientation. RESULTS: The PyRaDiSe package, characterized by its flexibility and automated routines, allows for fast deployment and prototyping, reducing efforts for auto-segmentation pipeline implementation. Furthermore, while deep learning model inference is independent of the deep learning framework, it can easily be integrated into famous deep learning frameworks such as PyTorch or Tensorflow. The developed package has successfully demonstrated its capabilities in a research project at our institution for organs-at-risk segmentation in brain tumor patients. Furthermore, PyRaDiSe has shown its conversion performance for dataset construction. CONCLUSIONS: The PyRaDiSe package closes the gap between data science and clinical radiotherapy by enabling deep learning segmentation models to be easily transferred into clinical research practice. PyRaDiSe is available on https://github.com/ubern-mia/pyradise and can be installed directly from the Python Package Index using pip install pyradise.

Impact of random outliers in auto-segmented targets on radiotherapy treatment plans for glioblastoma.

AIMS: To save time and have more consistent contours, fully automatic segmentation of targets and organs at risk (OAR) is a valuable asset in radiotherapy. Though current deep learning (DL) based models are on par with manual contouring, they are not perfect and typical errors, as false positives, occur frequently and unpredictably. While it is possible to solve this for OARs, it is far from straightforward for target structures. In order to tackle this problem, in this study, we analyzed the occurrence and the possible dose effects of automated delineation outliers. METHODS: First, a set of controlled experiments on synthetically generated outliers on the CT of a glioblastoma (GBM) patient was performed. We analyzed the dosimetric impact on outliers with different location, shape, absolute size and relative size to the main target, resulting in 61 simulated scenarios. Second, multiple segmentation models where trained on a U-Net network based on 80 training sets consisting of GBM cases with annotated gross tumor volume (GTV) and edema structures. On 20 test cases, 5 different trained models and a majority voting method were used to predict the GTV and edema. The amount of outliers on the predictions were determined, as well as their size and distance from the actual target. RESULTS: We found that plans containing outliers result in an increased dose to healthy brain tissue. The extent of the dose effect is dependent on the relative size, location and the distance to the main targets and involved OARs. Generally, the larger the absolute outlier volume and the distance to the target the higher the potential dose effect. For 120 predicted GTV and edema structures, we found 1887 outliers. After construction of the planning treatment volume (PTV), 137 outliers remained with a mean distance to the target of 38.5 ± 5.0 mm and a mean size of 1010.8 ± 95.6 mm3. We also found that majority voting of DL results is capable to reduce outliers. CONCLUSIONS: This study shows that there is a severe risk of false positive outliers in current DL predictions of target structures. Additionally, these errors will have an evident detrimental impact on the dose and therefore could affect treatment outcome.

Feasibility of postoperative spine stereotactic body radiation therapy in proximity of carbon and titanium hybrid implants using a robotic radiotherapy device.

BACKGROUND AND PURPOSE: To assess the feasibility of postoperative stereotactic body radiation therapy (SBRT) for patients with hybrid implants consisting of carbon fiber reinforced polyetheretherketone and titanium (CFP-T) using CyberKnife. MATERIALS AND METHODS: All essential steps within a radiation therapy (RT) workflow were evaluated. First, the contouring process of target volumes and organs at risk (OAR) was done for patients with CFP-T implants. Second, after RT-planning, the accuracy of the calculated dose distributions was tested in a slab phantom and an anthropomorphic phantom using film dosimetry. As a third step, the accuracy of the mandatory image guided radiation therapy (IGRT) including automatic matching was assessed using the anthropomorphic phantom. For this goal, a standard quality assurance (QA) test was modified to carry out its IGRT part in presence of CFP-T implants. RESULTS: Using CFP-T implants, target volumes could precisely delineated. There was no need for compromising the contours to overcome artifact obstacles. Differences between measured and calculated dose values were below 11% for the slab phantom, and at least 95% of the voxels were within 5% dose difference. The comparisons for the anthropomorphic phantom showed a gamma-passing rate (5%, 1 mm) of at least 97%. Additionally the test results with and without CFP-T implants were comparable. No issues concerning the IGRT were detected. The modified machine QA test resulted in a targeting error of 0.71 mm, which corresponds to the results of the unmodified standard tests. CONCLUSION: Dose calculation and delivery of postoperative spine SBRT is feasible in proximity of CFP-T implants using a CyberKnife system.

The predictive value of segmentation metrics on dosimetry in organs at risk of the brain.

BACKGROUND: Fully automatic medical image segmentation has been a long pursuit in radiotherapy (RT). Recent developments involving deep learning show promising results yielding consistent and time efficient contours. In order to train and validate these systems, several geometric based metrics, such as Dice Similarity Coefficient (DSC), Hausdorff, and other related metrics are currently the standard in automated medical image segmentation challenges. However, the relevance of these metrics in RT is questionable. The quality of automated segmentation results needs to reflect clinical relevant treatment outcomes, such as dosimetry and related tumor control and toxicity. In this study, we present results investigating the correlation between popular geometric segmentation metrics and dose parameters for Organs-At-Risk (OAR) in brain tumor patients, and investigate properties that might be predictive for dose changes in brain radiotherapy. METHODS: A retrospective database of glioblastoma multiforme patients was stratified for planning difficulty, from which 12 cases were selected and reference sets of OARs and radiation targets were defined. In order to assess the relation between segmentation quality -as measured by standard segmentation assessment metrics- and quality of RT plans, clinically realistic, yet alternative contours for each OAR of the selected cases were obtained through three methods: (i) Manual contours by two additional human raters. (ii) Realistic manual manipulations of reference contours. (iii) Through deep learning based segmentation results. On the reference structure set a reference plan was generated that was re-optimized for each corresponding alternative contour set. The correlation between segmentation metrics, and dosimetric changes was obtained and analyzed for each OAR, by means of the mean dose and maximum dose to 1% of the volume (Dmax 1%). Furthermore, we conducted specific experiments to investigate the dosimetric effect of alternative OAR contours with respect to the proximity to the target, size, particular shape and relative location to the target. RESULTS: We found a low correlation between the DSC, reflecting the alternative OAR contours, and dosimetric changes. The Pearson correlation coefficient between the mean OAR dose effect and the Dice was -0.11. For Dmax 1%, we found a correlation of -0.13. Similar low correlations were found for 22 other segmentation metrics. The organ based analysis showed that there is a better correlation for the larger OARs (i.e. brainstem and eyes) as for the smaller OARs (i.e. optic nerves and chiasm). Furthermore, we found that proximity to the target does not make contour variations more susceptible to the dose effect. However, the direction of the contour variation with respect to the relative location of the target seems to have a strong correlation with the dose effect. CONCLUSIONS: This study shows a low correlation between segmentation metrics and dosimetric changes for OARs in brain tumor patients. Results suggest that the current metrics for image segmentation in RT, as well as deep learning systems employing such metrics, need to be revisited towards clinically oriented metrics that better reflect how segmentation quality affects dose distribution and related tumor control and toxicity.

Vestibular dose correlates with dizziness after radiosurgery for the treatment of vestibular schwannoma.

BACKGROUND: Stereotactic radiosurgery (SRS) has been recognized as a first-line treatment option for small to moderate sized vestibular schwannoma (VS). Our aim is to evaluate the impact of SRS doses and other patient and disease characteristics on vestibular function in patients with VS. METHODS: Data on VS patients treated with single-fraction SRS to 12 Gy were retrospectively reviewed. No dose constraints were given to the vestibule during optimization in treatment planning. Patient and tumor characteristics, pre- and post-SRS vestibular examination results and patient-reported dizziness were assessed from patient records. RESULTS: Fifty-three patients were analyzed. Median follow-up was 32 months (range, 6-79). The median minimum, mean and maximum vestibular doses were 2.6 ± 1.6 Gy, 6.7 ± 2.8 Gy, and 11 ± 3.6 Gy, respectively. On univariate analysis, Koos grade (p = 0.04; OR: 3.45; 95% CI 1.01-11.81), tumor volume (median 6.1 cm3; range, 0.8-38; p = 0.01; OR: 4.85; 95% CI 1.43-16.49), presence of pre-SRS dizziness (p = 0.02; OR: 3.98; 95% CI 1.19-13.24) and minimum vestibular dose (p = 0.033; OR: 1.55; 95% CI 1.03-2.32) showed a significant association with patient-reported dizziness. On multivariate analysis, minimum vestibular dose remained significant (p = 0.02; OR: 1.75; 95% CI 1.05-2.89). Patients with improved caloric function had received significantly lower mean (1.5 ± 0.7 Gy, p = 0.01) and maximum doses (4 ± 1.5 Gy, p = 0.01) to the vestibule. CONCLUSIONS: Our results reveal that 5 Gy and above minimum vestibular doses significantly worsened dizziness. Additionally, mean and maximum doses received by the vestibule were significantly lower in patients who had improved caloric function. Further investigations are needed to determine dose-volume parameters and their effects on vestibular toxicity.

Dosimetric analysis of local failures in skull-base chordoma and chondrosarcoma following pencil beam scanning proton therapy.

BACKGROUND: Despite combined modality treatment involving surgery and radiotherapy, a relevant proportion of skull-base chordoma and chondrosarcoma patients develop a local recurrence (LR). This study aims to analyze patterns of recurrence and correlate LR with a detailed dosimetric analysis. METHODS: 222 patients were treated with proton radiotherapy for chordoma (n = 151) and chondrosarcoma (n = 71) at the PSI between 1998 and 2012. All patients underwent surgery, followed by pencil-beam scanning proton therapy to a mean dose of 72.5 ± 2.2GyRBE. A retrospective patterns of recurrence analysis was performed: LR were contoured on follow-up MRI, registered with planning-imaging and the overlap with initial target structures (GTV, PTVhigh-dose, PTVlow-dose) was calculated. DVH parameters of planning structures and recurrences were calculated and correlated with LR using univariate and multivariate cox regression. RESULTS: After a median follow-up of 50 months, 35 (16%) LR were observed. Follow-up MRI imaging was available for 27 (77%) of these recurring patients. Only one (3.7%) recurrence was located completely outside the initial PTV (surgical pathway recurrence). The mean proportions of LR covered by the initial target structures were 48% (range 0-86%) for the GTV, 70% (range 0-100%) for PTVhigh and 83% (range 0-100%) for PTVlow. In the univariate analysis, the following DVH parameters were significantly associated with LR: GTV(V < 66GyRBE, p = 0.01), GTV(volume, p = 0.02), PTVhigh(max, p = 0.02), PTVhigh(V < 66GyRBE, p = 0.03), PTVhigh(V < 59GyRBE, p = 0.02), PTVhigh(volume, p = 0.01) and GTV(D95, p = 0.05). In the multivariate analysis, only histology (chordoma vs. chondrosarcoma, p = 0.01), PTVhigh(volume, p = 0.05) and GTV(V < 66GyRBE, p = 0.02) were independent prognostic factors for LR. CONCLUSION: This study identified DVH parameters, which are associated with the risk of local recurrence after proton therapy using pencil-beam scanning for patients with skull-base chordoma and chondrosarcoma.

Fully automated brain resection cavity delineation for radiation target volume definition in glioblastoma patients using deep learning.

BACKGROUND: Automated brain tumor segmentation methods are computational algorithms that yield tumor delineation from, in this case, multimodal magnetic resonance imaging (MRI). We present an automated segmentation method and its results for resection cavity (RC) in glioblastoma multiforme (GBM) patients using deep learning (DL) technologies. METHODS: Post-operative, T1w with and without contrast, T2w and fluid attenuated inversion recovery MRI studies of 30 GBM patients were included. Three radiation oncologists manually delineated the RC to obtain a reference segmentation. We developed a DL cavity segmentation method, which utilizes all four MRI sequences and the reference segmentation to learn to perform RC delineations. We evaluated the segmentation method in terms of Dice coefficient (DC) and estimated volume measurements. RESULTS: Median DC of the three radiation oncologist were 0.85 (interquartile range [IQR]: 0.08), 0.84 (IQR: 0.07), and 0.86 (IQR: 0.07). The results of the automatic segmentation compared to the three different raters were 0.83 (IQR: 0.14), 0.81 (IQR: 0.12), and 0.81 (IQR: 0.13) which was significantly lower compared to the DC among raters (chi-square = 11.63, p = 0.04). We did not detect a statistically significant difference of the measured RC volumes for the different raters and the automated method (Kruskal-Wallis test: chi-square = 1.46, p = 0.69). The main sources of error were due to signal inhomogeneity and similar intensity patterns between cavity and brain tissues. CONCLUSIONS: The proposed DL approach yields promising results for automated RC segmentation in this proof of concept study. Compared to human experts, the DC are still subpar.

Early results and volumetric analysis after spot-scanning proton therapy with concomitant hyperthermia in large inoperable sacral chordomas.

OBJECTIVE: Large inoperable sacral chordomas show unsatisfactory local control rates even when treated with high dose proton therapy (PT). The aim of this study is assessing feasibility and reporting early results of patients treated with PT and concomitant hyperthermia (HT). METHODS:: Patients had histologically proven unresectable sacral chordomas and received 70 Gy (relative biological effectiveness) in 2.5 Gy fractions with concomitant weekly HT. Toxicity was assessed according to CTCAE_v4. A volumetric tumor response analysis was performed. RESULTS:: Five patients were treated with the combined approach. Median baseline tumor volume was 735 cc (range, 369-1142). All patients completed PT and received a median of 5 HT sessions (range, 2-6). Median follow-up was 18 months (range, 9-26). The volumetric analysis showed an objective response of all tumors (median shrinkage 46%; range, 9-72). All patients experienced acute Grade 2-3 local pain. One patient presented with a late Grade 3 iliac fracture. CONCLUSION: Combining PT and HT in large inoperable sacral chordomas is feasible and causes acceptable toxicity. Volumetric analysis shows promising early results, warranting confirmation in the framework of a prospective trial. ADVANCES IN KNOWLEDGE:: This is an encouraging first report of the feasibility and early results of concomitant HT and PT in treating inoperable sacral chordoma.

Dosimetric comparison of protons vs photons in re-irradiation of intracranial meningioma.

OBJECTIVES: Re-irradiation of recurrent intracranial meningiomas represents a major challenge due to dose limits of critical structures and the necessity of sufficient dose coverage of the recurrent tumor for local control. The aim of this study was to investigate dosimetric differences between pencil beam scanning protons (PBS) and volumetric modulated arc therapy (VMAT) photons for intracranial re-irradiation of meningiomas. METHODS: Nine patients who received an initial dose >50 Gy for intracranial meningioma and who were re-irradiated for recurrence were selected for plan comparison. A volumetric modulated arc therapy photon and a pencil beam scanning proton plan were generated (prescription dose: 15 × 3 Gy) based on the targets used in the re-irradiation treatment. RESULTS: In all cases, where the cumulative dose exceeded 100 or 90 Gy, these high dose volumes were larger for the proton plans. The integral doses were significantly higher in all photon plans (reduction with protons: 48.6%, p < 0.01). In two cases (22.2%), organ at risk (OAR) sparing was superior with the proton plan. In one case (11.1%), the photon plan showed a dosimetric advantage. In the remaining six cases (66.7%), we found no clinically relevant differences in dose to the OARs. CONCLUSIONS: The dosimetric results of the accumulated dose for a re-irradiation with protons and with photons were very similar. The photon plans had a steeper dose falloff directly outside the target and were superior in minimizing the high dose volumes. The proton plans achieved a lower integral dose. Clinically relevant OAR sparing was extremely case specific. The optimal treatment modality should be assessed individually. ADVANCES IN KNOWLEDGE: Dose sparing in re-irradiation of intracranial meningiomas with protons or photons is highly case specific and the optimal treatment modality needs to be assessed on an individual basis.