Clinical Implementation and Evaluation of Auto-Segmentation Tools for Multi-Site Contouring in Radiotherapy.

Background and purpose: Tools for auto-segmentation in radiotherapy are widely available, but guidelines for clinical implementation are missing. The goal was to develop a workflow for performance evaluation of three commercial auto-segmentation tools to select one candidate for clinical implementation. Materials and Methods: One hundred patients with six treatment sites (brain, head-and-neck, thorax, abdomen, and pelvis) were included. Three sets of AI-based contours for organs-at-risk (OAR) generated by three software tools and manually drawn expert contours were blindly rated for contouring accuracy. The dice similarity coefficient (DSC), the Hausdorff distance, and a dose/volume evaluation based on the recalculation of the original treatment plan were assessed. Statistically significant differences were tested using the Kruskal-Wallis test and the post-hoc Dunn Test with Bonferroni correction. Results: The mean DSC scores compared to expert contours for all OARs combined were 0.80 ± 0.10, 0.75 ± 0.10, and 0.74 ± 0.11 for the three software tools. Physicians' rating identified equivalent or superior performance of some AI-based contours in head (eye, lens, optic nerve, brain, chiasm), thorax (e.g., heart and lungs), and pelvis and abdomen (e.g., kidney, femoral head) compared to manual contours. For some OARs, the AI models provided results requiring only minor corrections. Bowel-bag and stomach were not fit for direct use. During the interdisciplinary discussion, the physicians' rating was considered the most relevant. Conclusion: A comprehensive method for evaluation and clinical implementation of commercially available auto-segmentation software was developed. The in-depth analysis yielded clear instructions for clinical use within the radiotherapy department.

Survey on brachytherapy training among radiation oncology residents in the German-speaking regions of Europe.

PURPOSE: This survey aimed to determine the perception of brachytherapy training among residents in the DACH region, consisting of Austria, Germany and Switzerland. MATERIAL & METHODS: An online questionnaire containing 22 questions related to trainee demographics (n = 5) and to brachytherapy training (n = 17) was sent in two iterations in 11/2019 and 02/2020. The following topics were evaluated: institutional support, barriers to training, extent of training, site-specific training (prostate, gynaecology, breast, gastrointestinal and skin), preferences for further training and outlook on overall development of brachytherapy. The responses were mostly based on a Likert scale of 1 to 5, thereby reflecting strength of opinion. Descriptive statistics were used to describe frequencies. RESULTS: Among the 108 respondents, approximately 69% of residents considered the ability to perform brachytherapy independently to be important or somewhat important. However, only 31% of respondents reported to have a dedicated brachytherapy training during residency. The major limitation to achieve independence in performing brachytherapy was seen in a low case load in Austria, in the lack of training in Switzerland and in both of them in Germany. CONCLUSION: The interest in brachytherapy training among residents in German-speaking countries was generally high, but there is a perceived lack of sufficient case volumes and partially also in formal training opportunities. Fellowships at departments with a high case load as part of a formalised curriculum and dedicated hands-on workshops at national or international conferences might help to overcome these issues.

EviGUIDE - a tool for evidence-based decision making in image-guided adaptive brachytherapy for cervical cancer.

PURPOSE: To develop a novel decision-support system for radiation oncology that incorporates clinical, treatment and outcome data, as well as outcome models from a large clinical trial on magnetic resonance image-guided adaptive brachytherapy (MR-IGABT) for locally advanced cervical cancer (LACC). METHODS: A system, called EviGUIDE, was developed that combines dosimetric information from the treatment planning system, patient and treatment characteristics, and established tumor control probability (TCP), and normal tissue complication probability (NTCP) models, to predict clinical outcome of radiotherapy treatment of LACC. Six Cox Proportional Hazards models based on data from 1341 patients of the EMBRACE-I study have been integrated. One TCP model for local tumor control, and five NTCP models for OAR morbidities. RESULTS: EviGUIDE incorporates TCP-NTCP graphs to help users visualize the clinical impact of different treatment plans and provides feedback on achievable doses based on a large reference population. It enables holistic assessment of the interplay between multiple clinical endpoints and tumour and treatment variables. Retrospective analysis of 45 patients treated with MR-IGABT showed that there exists a sub-cohort of patients (20%) with increased risk factors, that could greatly benefit from the quantitative and visual feedback. CONCLUSION: A novel digital concept was developed that can enhance clinical decision- making and facilitate personalized treatment. It serves as a proof of concept for a new generation of decision support systems in radiation oncology, which incorporate outcome models and high-quality reference data, and aids the dissemination of evidence-based knowledge about optimal treatment and serve as a blueprint for other sites in radiation oncology.

Tools for large-scale data analytics of an international multi-center study in radiation oncology for cervical cancer.

PURPOSE: To develop and implement a software that enables centers, treating patients with state-of-the-art radiation oncology, to compare their patient, treatment, and outcome data to a reference cohort, and to assess the quality of their treatment approach. MATERIALS AND METHODS: A comprehensive data dashboard was designed, which al- lowed holistic assessment of institutional treatment approaches. The software was tested in the ongoing EMBRACE-II study for locally advanced cervical cancer. The tool created individualized dashboards and automatic analysis scripts, verified pro- tocol compliance and checked data for inconsistencies. Identified quality assurance (QA) events were analysed. A survey among users was conducted to assess usability. RESULTS: The survey indicated favourable feedback to the prototype and highlighted its value for internal monitoring. Overall, 2302 QA events were identified (0.4% of all collected data). 54% were due to missing or incomplete data, and 46% originated from other causes. At least one QA event was found in 519/1001 (52%) of patients. QA events related to primary study endpoints were found in 16% of patients. Sta- tistical methods demonstrated good performance in detecting anomalies, with precisions ranging from 71% to 100%. Most frequent QA event categories were Treatment Technique (27%), Patient Characteristics (22%), Dose Reporting (17%), Outcome 156 (15%), Outliers (12%), and RT Structures (8%). CONCLUSION: A software tool was developed and tested within a clinical trial in radia- tion oncology. It enabled the quantitative and qualitative comparison of institutional patient and treatment parameters with a large multi-center reference cohort. We demonstrated the value of using statistical methods to automatically detect implau- sible data points and highlighted common pitfalls and uncertainties in radiotherapy for cervical cancer.

Brachytherapy training survey among radiation oncology residents in Europe.

We aim to investigate the current state of brachytherapy (BT) training among the radiation oncology trainees in Europe. MATERIAL AND METHODS: A 22-question online survey based on the one by the American Association of Radiation Oncology Residents (2017) with added queries pertinent to training in Europe was sent to 1450 residents in two iterations. These included site-specific training, volume of experience, barriers to training, institutional support, and preferences for further education. Responses to individual statements were given on a 1 to 5 Likert-type scale. The answers were reported by junior (≤3 years of training) and senior years of training (year of training 4/5/6 and junior staff). Descriptive statistics were used to describe frequencies. RESULTS: Residents from 21 European countries participated, 445 (31%) responded. 205 (47%) were senior residents. 60% residents consider that performing BT independently at the end of residency is very or somewhat important. Confidence in joining a brachytherapy practice at the end of residency was high or somewhat high in 34% of senior residents. They reported as barriers to achieving independence in BT to be lack of appropriate didactic/procedural training from supervisors (47%) and decreased case load (31%). 68% reported their program lacks a formal BT curriculum and standardized training assessment. CONCLUSIONS: Residents in Europe, feel independent BT practice is very or somewhat important, but do not feel confident they will achieve this goal. To address this gap, efforts are needed to develop and implement a formal and comprehensive BT curriculum with easy access to trained instructors.

Neural network-assisted automated image registration for MRI-guided adaptive brachytherapy in cervical cancer.

PURPOSE: In image-guided adaptive brachytherapy (IGABT) a quantitative evaluation of the dosimetric changes between fractions due to anatomical variations, can be implemented via rigid registration of images from subsequent fractions based on the applicator as a reference structure. With available treatment planning systems (TPS), this is a manual and time-consuming process. The aim of this retrospective study was to automate this process. A neural network (NN) was trained to predict the applicator structure from MR images. The resulting segmentation was used to automatically register MR-volumes. MATERIAL AND METHODS: DICOM images and plans of 56 patients treated for cervical cancer with high dose-rate (HDR) brachytherapy were used in the study. A 2D and a 3D NN were trained to segment applicator structures on clinical T2-weighted MRI datasets. Different rigid registration algorithms were investigated and compared. To evaluate a fully automatic registration workflow, the NN-predicted applicator segmentations (AS) were used for rigid image registration with the best performing algorithm. The DICE coefficient and mean distance error between dwell positions (MDE) were used to evaluate segmentation and registration performance. RESULTS: The mean DICE coefficient for the predicted AS was 0.70 ±â€¯0.07 and 0.58 ±â€¯0.04 for the 3D NN and 2D NN, respectively. Registration algorithms achieved MDE errors from 8.1 ±â€¯3.7 mm (worst) to 0.7 ±â€¯0.5 mm (best), using ground-truth AS. Using the predicted AS from the 3D NN together with the best registration algorithm, an MDE of 2.7 ±â€¯1.4 mm was achieved. CONCLUSION: Using a combination of deep learning models and state of the art image registration techniques has been demonstrated to be a promising solution for automatic image registration in IGABT. In combination with auto-contouring of organs at risk, the auto-registration workflow from this study could become part of an online-dosimetric interfraction evaluation workflow in the future.

Image-guided brachytherapy in cervical cancer including fractionation.

Image-guided brachytherapy in cervical cancer has been developed to be a feasible and very efficient component of the treatment of locally advanced cervical cancer in addition to concurrent chemoradiation treatment. This technique allows effective dose coverage of the target while sparing the organs at risk through adjustment of the implants (intracavitary and interstitial needles) and multi-pararametric three-dimensional treatment planning. Emerging evidence from prospective studies shows a high rate of local control throughout all stages, superior to two-dimensional brachytherapy, with limited toxicity for each organ site. This is associated with a high rate of pelvic control and overall survival. Based on clinical evidence, there is a dose-effect relationship for both disease and morbidity endpoints from which clear dose constraints for the target and organs at risk were derived. This review gives an overview of the major milestones that occurred in the development of image-guided adaptive brachytherapy in the last two decades, including outcome data and a summary of the hard and soft dose constraints recommended for targets and organs at risk.

Toward 3D-TRUS image-guided interstitial brachytherapy for cervical cancer.

PURPOSE: To qualitatively and quantitatively analyze needle visibility in combined intracavitary and interstitial cervical cancer brachytherapy on 3D transrectal ultrasound (TRUS) in comparison to gold standard MRI. METHODS AND MATERIALS: Image acquisition was done with a customized TRUS stepper unit and software (Medcom, Germany; Elekta, Sweden; ACMIT, Austria) followed by an MRI on the same day with the applicator in place. Qualitative assessment was done with following scoring system: 0 = no visibility 1 (= poor), 2 (= fair), 3 (= excellent) discrimination, quantitative assessment was done by measuring the distance between each needle and the tandem two centimeters (cm) above the ring and comparing to the respective measurement on MRI. RESULTS: Twenty-nine implants and a total of 188 needles (132 straight, 35 oblique, 21 free-hand) were available. Overall, 79% were visible (87% straight, 51% oblique, 76% free-hand). Mean visibility score was 1.4 ± 0.5 for all visible needles. Distance of the visible needles to tandem was mean ± standard deviation (SD) 21.3 millimeters (mm) ± 6.5 mm on MRI and 21.0 mm ± 6.4 mm on TRUS, respectively. Difference between MRI and TRUS was max 14 mm, mean ± SD -0.3 mm ± 2.6 mm. 11% differed more than 3 mm. CONCLUSIONS: Straight needles were better detectable than oblique needles (87% vs. 51%). Detectability was impaired by insufficient rotation of the TRUS probe, poor image quality or anatomic variation. As needles show a rather indistinct signal on TRUS, online detection with a standardized imaging protocol in combination with tracking should be investigated, aiming at the development of real time image guidance and online treatment planning.

Hybrid TRUS/CT with optical tracking for target delineation in image-guided adaptive brachytherapy for cervical cancer.

OBJECTIVE: To prospectively compare the interobserver variability of combined transrectal ultrasound (TRUS)/computed tomography (CT)- vs. CT only- vs. magnetic resonance imaging (MRI) only-based contouring of the high-risk clinical target volume (CTVHR) in image-guided adaptive brachytherapy (IGABT) for locally advanced cervical cancer (LACC). METHODS: Five patients with LACC (FIGO stages IIb-IVa) treated with radiochemotherapy and IGABT were included. CT, TRUS, and T2-weighted MRI images were performed after brachytherapy applicator insertion. 3D-TRUS image acquisition was performed with a customized ultrasound stepper device and software. Automatic applicator reconstruction using optical tracking was performed in the TRUS dataset and TRUS and CT images were fused with rigid image registration with the applicator as reference structure. The CTVHR (based on the GEC-ESTRO recommendations) was contoured by five investigators on the three modalities (CTVHR_CT, CTVHR_TRUS-CT, and CTVHR_MRI). A consensus reference CTVHR_MRI (MRIref) was defined for each patient. Descriptive statistics and overlap measures were calculated using RTslicer (SlicerRT Community and Percutaneous Surgery Laboratory, Queen's University, Canada), comparing contours of every observer with one another and with the MRIref. RESULTS: The interobserver coefficient of variation was 0.18 ± 0.05 for CT, 0.10 ± 0.04 for TRUS-CT, and 0.07 ± 0.03 for MRI. Interobserver concordance in relation to the MRIref expressed by the generalized conformity index was 0.75 ± 0.04 for MRI, 0.51 ± 0.10 for TRUS-CT, and 0.48 ± 0.06 for CT. The mean CTVHR_CT volume of all observers was 71% larger than the MRIref volume, whereas the mean CTVHR_TRUS-CT volume was 15% larger. CONCLUSION: Hybrid TRUS-CT as an imaging modality for contouring the CTVHR in IGABT for LACC is feasible and reproducible among multiple observers. TRUS-CT substantially reduces overestimation of the CTVHR volume of CT alone while maintaining similar interobserver variability.

Association between pathology and texture features of multi parametric MRI of the prostate.

The role of multi-parametric (mp)MRI in the diagnosis and treatment of prostate cancer has increased considerably. An alternative to visual inspection of mpMRI is the evaluation using histogram-based (first order statistics) parameters and textural features (second order statistics). The aims of the present work were to investigate the relationship between benign and malignant sub-volumes of the prostate and textures obtained from mpMR images. The performance of tumor prediction was investigated based on the combination of histogram-based and textural parameters. Subsequently, the relative importance of mpMR images was assessed and the benefit of additional imaging analyzed. Finally, sub-structures based on the PI-RADS classification were investigated as potential regions to automatically detect maligned lesions. Twenty-five patients who received mpMRI prior to radical prostatectomy were included in the study. The imaging protocol included T2, DWI, and DCE. Delineation of tumor regions was performed based on pathological information. First and second order statistics were derived from each structure and for all image modalities. The resulting data were processed with multivariate analysis, using PCA (principal component analysis) and OPLS-DA (orthogonal partial least squares discriminant analysis) for separation of malignant and healthy tissue. PCA showed a clear difference between tumor and healthy regions in the peripheral zone for all investigated images. The predictive ability of the OPLS-DA models increased for all image modalities when first and second order statistics were combined. The predictive value reached a plateau after adding ADC and T2, and did not increase further with the addition of other image information. The present study indicates a distinct difference in the signatures between malign and benign prostate tissue. This is an absolute prerequisite for automatic tumor segmentation, but only the first step in that direction. For the specific identified signature, DCE did not add complementary information to T2 and ADC maps.

Feasibility of dominant intraprostatic lesion boosting using advanced photon-, proton- or brachytherapy.

BACKGROUND AND PURPOSE: Advancements in imaging and dose delivery enable boosting of the dominant intraprostatic lesions (DIL), while maintaining organs-at-risk (OAR) tolerances. This study aimed to assess the feasibility of DIL boosting for volumetric modulated arc therapy (VMAT), intensity modulated proton therapy (IMPT) and high dose rate brachytherapy (HDR-BT). MATERIAL AND METHODS: DILs were defined on multiparametric magnetic resonance imaging and fused with planning CT for twelve patients. VMAT, IMPT and HDR-BT plans were created for each patient with an EQD2(α/ß) DIL aimed at 111.6 Gy, PTV(initial) D(pres) was 80.9 Gy (EBRT) with CTV D90%=81.9 Gy (HDR-BT). Hard dose constraints were applied for OARs. RESULTS: Higher boost doses were achieved with IMPT compared to VMAT, keeping major OAR doses at similar levels. Patient averaged EQD2(α/ß) D50% to DIL were 110.7, 114.2 and 150.1 Gy(IsoE) for VMAT, IMPT and HDR-BT, respectively. Respective rectal wall D(mean) were 30.5±5.0, 16.7±3.6, 9.5±2.5 Gy(IsoE) and bladder wall D(mean) were 21.0±5.5, 15.6±4.3 and 6.3±2.2 Gy(IsoE). CONCLUSIONS: DIL boosting was found to be feasible with all investigated techniques. Although OAR doses were higher than for standard treatment approach, the risk levels were reasonably low. HDR-BT was superior to VMAT and IMPT, both in terms of OAR sparing and DIL boosting.