PET/CT based transformer model for multi-outcome prediction in oropharyngeal cancer.

BACKGROUND AND PURPOSE: To optimize our previously proposed TransRP, a model integrating CNN (convolutional neural network) and ViT (Vision Transformer) designed for recurrence-free survival prediction in oropharyngeal cancer and to extend its application to the prediction of multiple clinical outcomes, including locoregional control (LRC), Distant metastasis-free survival (DMFS) and overall survival (OS). MATERIALS AND METHODS: Data was collected from 400 patients (300 for training and 100 for testing) diagnosed with oropharyngeal squamous cell carcinoma (OPSCC) who underwent (chemo)radiotherapy at University Medical Center Groningen. Each patient's data comprised pre-treatment PET/CT scans, clinical parameters, and clinical outcome endpoints, namely LRC, DMFS and OS. The prediction performance of TransRP was compared with CNNs when inputting image data only. Additionally, three distinct methods (m1-3) of incorporating clinical predictors into TransRP training and one method (m4) that uses TransRP prediction as one parameter in a clinical Cox model were compared. RESULTS: TransRP achieved higher test C-index values of 0.61, 0.84 and 0.70 than CNNs for LRC, DMFS and OS, respectively. Furthermore, when incorporating TransRP's prediction into a clinical Cox model (m4), a higher C-index of 0.77 for OS was obtained. Compared with a clinical routine risk stratification model of OS, our model, using clinical variables, radiomics and TransRP prediction as predictors, achieved larger separations of survival curves between low, intermediate and high risk groups. CONCLUSION: TransRP outperformed CNN models for all endpoints. Combining clinical data and TransRP prediction in a Cox model achieved better OS prediction.

A comparative study of federated learning methods for COVID-19 detection.

Deep learning has proven to be highly effective in diagnosing COVID-19; however, its efficacy is contingent upon the availability of extensive data for model training. The data sharing among hospitals, which is crucial for training robust models, is often restricted by privacy regulations. Federated learning (FL) emerges as a solution by enabling model training across multiple hospitals while preserving data privacy. However, the deployment of FL can be resource-intensive, necessitating efficient utilization of computational and network resources. In this study, we evaluate the performance and resource efficiency of five FL algorithms in the context of COVID-19 detection using Convolutional Neural Networks (CNNs) in a decentralized setting. The evaluation involves varying the number of participating entities, the number of federated rounds, and the selection algorithms. Our findings indicate that the Cyclic Weight Transfer algorithm exhibits superior performance, particularly when the number of participating hospitals is limited. These insights hold practical implications for the deployment of FL algorithms in COVID-19 detection and broader medical image analysis.

METhodological RadiomICs Score (METRICS): a quality scoring tool for radiomics research endorsed by EuSoMII.

PURPOSE: To propose a new quality scoring tool, METhodological RadiomICs Score (METRICS), to assess and improve research quality of radiomics studies. METHODS: We conducted an online modified Delphi study with a group of international experts. It was performed in three consecutive stages: Stage#1, item preparation; Stage#2, panel discussion among EuSoMII Auditing Group members to identify the items to be voted; and Stage#3, four rounds of the modified Delphi exercise by panelists to determine the items eligible for the METRICS and their weights. The consensus threshold was 75%. Based on the median ranks derived from expert panel opinion and their rank-sum based conversion to importance scores, the category and item weights were calculated. RESULT: In total, 59 panelists from 19 countries participated in selection and ranking of the items and categories. Final METRICS tool included 30 items within 9 categories. According to their weights, the categories were in descending order of importance: study design, imaging data, image processing and feature extraction, metrics and comparison, testing, feature processing, preparation for modeling, segmentation, and open science. A web application and a repository were developed to streamline the calculation of the METRICS score and to collect feedback from the radiomics community. CONCLUSION: In this work, we developed a scoring tool for assessing the methodological quality of the radiomics research, with a large international panel and a modified Delphi protocol. With its conditional format to cover methodological variations, it provides a well-constructed framework for the key methodological concepts to assess the quality of radiomic research papers. CRITICAL RELEVANCE STATEMENT: A quality assessment tool, METhodological RadiomICs Score (METRICS), is made available by a large group of international domain experts, with transparent methodology, aiming at evaluating and improving research quality in radiomics and machine learning. KEY POINTS: • A methodological scoring tool, METRICS, was developed for assessing the quality of radiomics research, with a large international expert panel and a modified Delphi protocol. • The proposed scoring tool presents expert opinion-based importance weights of categories and items with a transparent methodology for the first time. • METRICS accounts for varying use cases, from handcrafted radiomics to entirely deep learning-based pipelines. • A web application has been developed to help with the calculation of the METRICS score ( https://metricsscore.github.io/metrics/METRICS.html ) and a repository created to collect feedback from the radiomics community ( https://github.com/metricsscore/metrics ).

Proton and photon radiotherapy in stage III NSCLC: Effects on hematological toxicity and adjuvant immune therapy.

BACKGROUND AND PURPOSE: Concurrent chemo-radiotherapy (CCRT) followed by adjuvant durvalumab is standard-of-care for fit patients with unresectable stage III NSCLC. Intensity modulated proton therapy (IMPT) results in different doses to organs than intensity modulated photon therapy (IMRT). We investigated whether IMPT compared to IMRT reduce hematological toxicity and whether it affects durvalumab treatment. MATERIALS AND METHODS: Prospectively collected series of consecutive patients with stage III NSCLC receiving CCRT between 06.16 and 12.22 (staged with FDG-PET-CT and brain imaging) were retrospectively analyzed. The primary endpoint was the incidence of lymphopenia grade ≥ 3 in IMPT vs IMRT treated patients. RESULTS: 271 patients were enrolled (IMPT: n = 71, IMRT: n = 200) in four centers. All patients received platinum-based chemotherapy. Median age: 66 years, 58 % were male, 36 % had squamous NSCLC. The incidence of lymphopenia grade ≥ 3 during CCRT was 67 % and 47 % in the IMRT and IMPT group, respectively (OR 2.2, 95 % CI: 1.0-4.9, P = 0.03). The incidence of anemia grade ≥ 3 during CCRT was 26 % and 9 % in the IMRT and IMPT group respectively (OR = 4.9, 95 % CI: 1.9-12.6, P = 0.001). IMPT was associated with a lower rate of Performance Status (PS) ≥ 2 at day 21 and 42 after CCRT (13 % vs. 26 %, P = 0.04, and 24 % vs. 39 %, P = 0.02). Patients treated with IMPT had a higher probability of receiving adjuvant durvalumab (74 % vs. 52 %, OR 0.35, 95 % CI: 0.16-0.79, P = 0.01). CONCLUSION: IMPT was associated with a lower incidence of severe lymphopenia and anemia, better PS after CCRT and a higher probability of receiving adjuvant durvalumab.

Deep learning-based outcome prediction using PET/CT and automatically predicted probability maps of primary tumor in patients with oropharyngeal cancer.

BACKGROUND AND OBJECTIVE: Recently, deep learning (DL) algorithms showed to be promising in predicting outcomes such as distant metastasis-free survival (DMFS) and overall survival (OS) using pre-treatment imaging in head and neck cancer. Gross Tumor Volume of the primary tumor (GTVp) segmentation is used as an additional channel in the input to DL algorithms to improve model performance. However, the binary segmentation mask of the GTVp directs the focus of the network to the defined tumor region only and uniformly. DL models trained for tumor segmentation have also been used to generate predicted tumor probability maps (TPM) where each pixel value corresponds to the degree of certainty of that pixel to be classified as tumor. The aim of this study was to explore the effect of using TPM as an extra input channel of CT- and PET-based DL prediction models for oropharyngeal cancer (OPC) patients in terms of local control (LC), regional control (RC), DMFS and OS. METHODS: We included 399 OPC patients from our institute that were treated with definitive (chemo)radiation. For each patient, CT and PET scans and GTVp contours, used for radiotherapy treatment planning, were collected. We first trained a previously developed 2.5D DL framework for tumor probability prediction by 5-fold cross validation using 131 patients. Then, a 3D ResNet18 was trained for outcome prediction using the 3D TPM as one of the possible inputs. The endpoints were LC, RC, DMFS, and OS. We performed 3-fold cross validation on 168 patients for each endpoint using different combinations of image modalities as input. The final prediction in the test set (100) was obtained by averaging the predictions of the 3-fold models. The C-index was used to evaluate the discriminative performance of the models. RESULTS: The models trained replacing the GTVp contours with the TPM achieved the highest C-indexes for LC (0.74) and RC (0.60) prediction. For OS, using the TPM or the GTVp as additional image modality resulted in comparable C-indexes (0.72 and 0.74). CONCLUSIONS: Adding predicted TPMs instead of GTVp contours as an additional input channel for DL-based outcome prediction models improved model performance for LC and RC.

An umbrella review of candidate predictors of response, remission, recovery, and relapse across mental disorders.

We aimed to identify diagnosis-specific/transdiagnostic/transoutcome multivariable candidate predictors (MCPs) of key outcomes in mental disorders. We conducted an umbrella review (protocol  link ), searching MEDLINE/Embase (19/07/2022), including systematic reviews of studies reporting on MCPs of response, remission, recovery, or relapse, in DSM/ICD-defined mental disorders. From published predictors, we filtered MCPs, validating MCP criteria. AMSTAR2/PROBAST measured quality/risk of bias of systematic reviews/individual studies. We included 117 systematic reviews, 403 studies, 299,888 individuals with mental disorders, testing 796 prediction models. Only 4.3%/1.2% of the systematic reviews/individual studies were at low risk of bias. The most frequently targeted outcome was remission (36.9%), the least frequent was recovery (2.5%). Studies mainly focused on depressive (39.4%), substance-use (17.9%), and schizophrenia-spectrum (11.9%) disorders. We identified numerous MCPs within disorders for response, remission and relapse, but none for recovery. Transdiagnostic MCPs of remission included lower disease-specific symptoms (disorders = 5), female sex/higher education (disorders = 3), and quality of life/functioning (disorders = 2). Transdiagnostic MCPs of relapse included higher disease-specific symptoms (disorders = 5), higher depressive symptoms (disorders = 3), and younger age/higher anxiety symptoms/global illness severity/ number of previous episodes/negative life events (disorders = 2). Finally, positive trans-outcome MCPs for depression included less negative life events/depressive symptoms (response, remission, less relapse), female sex (response, remission) and better functioning (response, less relapse); for schizophrenia, less positive symptoms/higher depressive symptoms (remission, less relapse); for substance use disorder, marital status/higher education (remission, less relapse). Male sex, younger age, more clinical symptoms and comorbid mental/physical symptoms/disorders were poor prognostic factors, while positive factors included social contacts and employment, absent negative life events, higher education, early access/intervention, lower disease-specific and comorbid mental and physical symptoms/conditions, across mental disorders. Current data limitations include high risk of bias of studies and extraction of single predictors from multivariable models. Identified MCPs can inform future development, validation or refinement of prediction models of key outcomes in mental disorders.

Comparison of computed tomography image features extracted by radiomics, self-supervised learning and end-to-end deep learning for outcome prediction of oropharyngeal cancer.

Background and purpose: To compare the prediction performance of image features of computed tomography (CT) images extracted by radiomics, self-supervised learning and end-to-end deep learning for local control (LC), regional control (RC), locoregional control (LRC), distant metastasis-free survival (DMFS), tumor-specific survival (TSS), overall survival (OS) and disease-free survival (DFS) of oropharyngeal squamous cell carcinoma (OPSCC) patients after (chemo)radiotherapy. Methods and materials: The OPC-Radiomics dataset was used for model development and independent internal testing and the UMCG-OPC set for external testing. Image features were extracted from the Gross Tumor Volume contours of the primary tumor (GTVt) regions in CT scans when using radiomics or a self-supervised learning-based method (autoencoder). Clinical and combined (radiomics, autoencoder or end-to-end) models were built using multivariable Cox proportional-hazard analysis with clinical features only and both clinical and image features for LC, RC, LRC, DMFS, TSS, OS and DFS prediction, respectively. Results: In the internal test set, combined autoencoder models performed better than clinical models and combined radiomics models for LC, RC, LRC, DMFS, TSS and DFS prediction (largest improvements in C-index: 0.91 vs. 0.76 in RC and 0.74 vs. 0.60 in DMFS). In the external test set, combined radiomics models performed better than clinical and combined autoencoder models for all endpoints (largest improvements in LC, 0.82 vs. 0.71). Furthermore, combined models performed better in risk stratification than clinical models and showed good calibration for most endpoints. Conclusions: Image features extracted using self-supervised learning showed best internal prediction performance while radiomics features have better external generalizability.

CT-based deep multi-label learning prediction model for outcome in patients with oropharyngeal squamous cell carcinoma.

BACKGROUND: Personalized treatment is increasingly required for oropharyngeal squamous cell carcinoma (OPSCC) patients due to emerging new cancer subtypes and treatment options. Outcome prediction model can help identify low or high-risk patients who may be suitable to receive de-escalation or intensified treatment approaches. PURPOSE: To develop a deep learning (DL)-based model for predicting multiple and associated efficacy endpoints in OPSCC patients based on computed tomography (CT). METHODS: Two patient cohorts were used in this study: a development cohort consisting of 524 OPSCC patients (70% for training and 30% for independent testing) and an external test cohort of 396 patients. Pre-treatment CT-scans with the gross primary tumor volume contours (GTVt) and clinical parameters were available to predict endpoints, including 2-year local control (LC), regional control (RC), locoregional control (LRC), distant metastasis-free survival (DMFS), disease-specific survival (DSS), overall survival (OS), and disease-free survival (DFS). We proposed DL outcome prediction models with the multi-label learning (MLL) strategy that integrates the associations of different endpoints based on clinical factors and CT-scans. RESULTS: The multi-label learning models outperformed the models that were developed based on a single endpoint for all endpoints especially with high AUCs ≥ 0.80 for 2-year RC, DMFS, DSS, OS, and DFS in the internal independent test set and for all endpoints except 2-year LRC in the external test set. Furthermore, with the models developed, patients could be stratified into high and low-risk groups that were significantly different for all endpoints in the internal test set and for all endpoints except DMFS in the external test set. CONCLUSION: MLL models demonstrated better discriminative ability for all 2-year efficacy endpoints than single outcome models in the internal test and for all endpoints except LRC in the external set.

Increased risk of stress urinary incontinence surgery after hysterectomy for benign indication-a population-based cohort study.

BACKGROUND: Hysterectomy is a common procedure used to treat different gynecologic conditions. The association between hysterectomy for benign indication and stress urinary incontinence has previously been established. Stress urinary incontinence can be treated surgically, and options have improved after introduction of the midurethral sling procedure in 1998. OBJECTIVE: This study aimed to estimate the risk of stress urinary incontinence surgery after hysterectomy for benign indication. STUDY DESIGN: The study was carried out as a matched register-based cohort study including Danish women born from 1947 to 2000. Women who underwent hysterectomy for benign indication were matched to nonhysterectomized women in a 1:5 ratio on the basis of age and calendar year of hysterectomy. The risk of stress urinary incontinence surgery after hysterectomy was estimated. We adjusted for income, educational level, and parity. The risk of stress urinary incontinence surgery was further estimated in a subcohort excluding all vaginal hysterectomies. The joint effect of hysterectomy and parity was estimated in the main cohort, and the joint effect of hysterectomy and vaginal birth or cesarean delivery on stress urinary incontinence surgery was explored in a subgroup of women who only had 1 mode of delivery. All analyses were made using the Cox proportional hazards model. RESULTS: We included 83,370 women who underwent hysterectomy and 413,969 reference women. The overall risk of stress urinary incontinence surgery was more than doubled for women who underwent hysterectomy (adjusted hazard ratio, 2.6; 95% confidence interval, 2.4-2.8). The adjusted hazard ratio decreased slightly to 2.4 (95% confidence interval, 2.3-2.6) when excluding all vaginal hysterectomies. We found a trend of increasing risk of stress urinary incontinence surgery with increased parity among both women who underwent hysterectomy and the reference group. In the subgroup of women who only had 1 mode of delivery, we found the risk of stress urinary incontinence surgery to be particularly increased for women with a history of ≥1 vaginal births. The hazard ratio was 15.1 (95% confidence interval, 10.3-22.1) for women with a history of 1 vaginal birth who underwent hysterectomy, whereas the hazard ratio for women in the reference group with 1 vaginal birth was 5.1 (95% confidence interval, 3.8-8.1). Overall, women who underwent hysterectomy had a 3 times higher risk of stress urinary incontinence surgery than the reference group, irrespective of the number of vaginal births. CONCLUSION: This study indicates, in accordance with previous studies, that hysterectomy increases the risk of subsequent stress urinary incontinence surgery. Women should be informed and gynecologists include this knowledge in decision-making. Further precautions should be taken when treating parous women, particularly those with a history of ≥1 vaginal births.

Deep learning aided oropharyngeal cancer segmentation with adaptive thresholding for predicted tumor probability in FDG PET and CT images.

Objective. Tumor segmentation is a fundamental step for radiotherapy treatment planning. To define an accurate segmentation of the primary tumor (GTVp) of oropharyngeal cancer patients (OPC) each image volume is explored slice-by-slice from different orientations on different image modalities. However, the manual fixed boundary of segmentation neglects the spatial uncertainty known to occur in tumor delineation. This study proposes a novel deep learning-based method that generates probability maps which capture the model uncertainty in the segmentation task.Approach. We included 138 OPC patients treated with (chemo)radiation in our institute. Sequences of 3 consecutive 2D slices of concatenated FDG-PET/CT images and GTVp contours were used as input. Our framework exploits inter and intra-slice context using attention mechanisms and bi-directional long short term memory (Bi-LSTM). Each slice resulted in three predictions that were averaged. A 3-fold cross validation was performed on sequences extracted from the axial, sagittal, and coronal plane. 3D volumes were reconstructed and single- and multi-view ensembling were performed to obtain final results. The output is a tumor probability map determined by averaging multiple predictions.Main Results. Model performance was assessed on 25 patients at different probability thresholds. Predictions were the closest to the GTVp at a threshold of 0.9 (mean surface DSC of 0.81, median HD95of 3.906 mm).Significance. The promising results of the proposed method show that is it possible to offer the probability maps to radiation oncologists to guide them in a in a slice-by-slice adaptive GTVp segmentation.

Survival prediction for stage I-IIIA non-small cell lung cancer using deep learning.

BACKGROUND AND PURPOSE: The aim of this study was to develop and evaluate a prediction model for 2-year overall survival (OS) in stage I-IIIA non-small cell lung cancer (NSCLC) patients who received definitive radiotherapy by considering clinical variables and image features from pre-treatment CT-scans. MATERIALS AND METHODS: NSCLC patients who received stereotactic radiotherapy were prospectively collected at the UMCG and split into a training and a hold out test set including 189 and 81 patients, respectively. External validation was performed on 228 NSCLC patients who were treated with radiation or concurrent chemoradiation at the Maastro clinic (Lung1 dataset). A hybrid model that integrated both image and clinical features was implemented using deep learning. Image features were learned from cubic patches containing lung tumours extracted from pre-treatment CT scans. Relevant clinical variables were selected by univariable and multivariable analyses. RESULTS: Multivariable analysis showed that age and clinical stage were significant prognostic clinical factors for 2-year OS. Using these two clinical variables in combination with image features from pre-treatment CT scans, the hybrid model achieved a median AUC of 0.76 [95 % CI: 0.65-0.86] and 0.64 [95 % CI: 0.58-0.70] on the complete UMCG and Maastro test sets, respectively. The Kaplan-Meier survival curves showed significant separation between low and high mortality risk groups on these two test sets (log-rank test: p-value < 0.001, p-value = 0.012, respectively) CONCLUSION: We demonstrated that a hybrid model could achieve reasonable performance by utilizing both clinical and image features for 2-year OS prediction. Such a model has the potential to identify patients with high mortality risk and guide clinical decision making.

Head and neck cancer predictive risk estimator to determine control and therapeutic outcomes of radiotherapy (HNC-PREDICTOR): development, international multi-institutional validation, and web implementation of clinic-ready model-based risk stratification for head and neck cancer.

BACKGROUND: Personalised radiotherapy can improve treatment outcomes of patients with head and neck cancer (HNC), where currently a 'one-dose-fits-all' approach is the standard. The aim was to establish individualised outcome prediction based on multi-institutional international 'big-data' to facilitate risk-based stratification of patients with HNC. METHODS: The data of 4611 HNC radiotherapy patients from three academic cancer centres were split into four cohorts: a training (n = 2241), independent test (n = 786), and external validation cohorts 1 (n = 1087) and 2 (n = 497). Tumour- and patient-related clinical variables were considered in a machine learning pipeline to predict overall survival (primary end-point) and local and regional tumour control (secondary end-points); serially, imaging features were considered for optional model improvement. Finally, patients were stratified into high-, intermediate-, and low-risk groups. RESULTS: Performance score, AJCC8thstage, pack-years, and Age were identified as predictors for overall survival, demonstrating good performance in both the training cohort (c-index = 0.72 [95% CI, 0.66-0.77]) and in all three validation cohorts (c-indices: 0.76 [0.69-0.83], 0.73 [0.68-0.77], and 0.75 [0.68-0.80]). Excellent stratification of patients with HNC into high, intermediate, and low mortality risk was achieved; with 5-year overall survival rates of 17-46% for the high-risk group compared to 92-98% for the low-risk group. The addition of morphological image feature further improved the performance (c-index = 0.73 [0.64-0.81]). These models are integrated in a clinic-ready interactive web interface: https://uic-evl.github.io/hnc-predictor/ CONCLUSIONS: Robust model-based prediction was able to stratify patients with HNC in distinct high, intermediate, and low mortality risk groups. This can effectively be capitalised for personalised radiotherapy, e.g., for tumour radiation dose escalation/de-escalation.

Daily dose evaluation based on corrected CBCTs for breast cancer patients: accuracy of dose and complication risk assessment.

OBJECTIVES: The goal of this study is to validate different CBCT correction methods to select the superior method that can be used for dose evaluation in breast cancer patients with large anatomical changes treated with photon irradiation. MATERIALS AND METHOD: Seventy-six breast cancer patients treated with a partial VMAT photon technique (70% conformal, 30% VMAT) were included in this study. All patients showed at least a 5 mm variation (swelling or shrinkage) of the breast on the CBCT compared to the planning-CT (pCT) and had a repeat-CT (rCT) for dose evaluation acquired within 3 days of this CBCT. The original CBCT was corrected using four methods: (1) HU-override correction (CBCTHU), (2) analytical correction and conversion (CBCTCC), (3) deep learning (DL) correction (CTDL) and (4) virtual correction (CTV). Image quality evaluation consisted of calculating the mean absolute error (MAE) and mean error (ME) within the whole breast clinical target volume (CTV) and the field of view of the CBCT minus 2 cm (CBCT-ROI) with respect to the rCT. The dose was calculated on all image sets using the clinical treatment plan for dose and gamma passing rate analysis. RESULTS: The MAE of the CBCT-ROI was below 66 HU for all corrected CBCTs, except for the CBCTHU with a MAE of 142 HU. No significant dose differences were observed in the CTV regions in the CBCTCC, CTDL and CTv. Only the CBCTHU deviated significantly (p < 0.01) resulting in 1.7% (± 1.1%) average dose deviation. Gamma passing rates were > 95% for 2%/2 mm for all corrected CBCTs. CONCLUSION: The analytical correction and conversion, deep learning correction and virtual correction methods can be applied for an accurate CBCT correction that can be used for dose evaluation during the course of photon radiotherapy of breast cancer patients.

Diaphragm-Based Position Verification to Improve Daily Target Dose Coverage in Proton and Photon Radiation Therapy Treatment of Distal Esophageal Cancer.

PURPOSE: In modern conformal radiation therapy of distal esophageal cancer, target coverage can be affected by variations in the diaphragm position. We investigated if daily position verification (PV) extended by a diaphragm position correction would optimize target dose coverage for esophageal cancer treatment. METHODS AND MATERIALS: For 15 esophageal cancer patients, intensity modulated proton therapy (IMPT) and volumetric modulated arc therapy (VMAT) plans were computed. Displacements of the target volume were correlated with diaphragm displacements using repeated 4-dimensional computed tomography images to determine the correction needed to account for diaphragm variations. Afterwards, target coverage was evaluated for 3 PV approaches based on: (1) bony anatomy (PV_B), (2) bony anatomy corrected for the diaphragm position (PV_BD) and (3) target volume (PV_T). RESULTS: The cranial-caudal mean target displacement was congruent with almost half of the diaphragm displacement (y = 0.459x), which was used for the diaphragm correction in PV_BD. Target dose coverage using PV_B was adequate for most patients with diaphragm displacements up till 10 mm (≥94% of the dose in 98% of the volume [D98%]). For larger displacements, the target coverage was better maintained by PV_T and PV_BD. Overall, PV_BD accounted best for target displacements, especially in combination with tissue density variations (D98%: IMPT 94% ± 5%, VMAT 96% ± 5%). Diaphragm displacements of more than 10 mm were observed in 22% of the cases. CONCLUSIONS: PV_B was sufficient to achieve adequate target dose coverage in case of small deviations in diaphragm position. However, large deviations of the diaphragm were best mitigated by PV_BD. To detect the cases where target dose coverage could be compromised due to diaphragm position variations, we recommend monitoring of the diaphragm position before treatment through online imaging.

A novel semi auto-segmentation method for accurate dose and NTCP evaluation in adaptive head and neck radiotherapy.

BACKGROUND AND PURPOSE: Accurate segmentation of organs-at-risk (OARs) is crucial but tedious and time-consuming in adaptive radiotherapy (ART). The purpose of this work was to automate head and neck OAR-segmentation on repeat CT (rCT) by an optimal combination of human and auto-segmentation for accurate prediction of Normal Tissue Complication Probability (NTCP). MATERIALS AND METHODS: Human segmentation (HS) of 3 observers, deformable image registration (DIR) based contour propagation and deep learning contouring (DLC) were carried out to segment 15 OARs on 15 rCTs. The original treatment plan was re-calculated on rCT to obtain mean dose (Dmean) and consequent NTCP-predictions. The average Dmean and NTCP-predictions of the three observers were referred to as the gold standard to calculate the absolute difference of Dmean and NTCP-predictions (|ΔDmean| and |ΔNTCP|). RESULTS: The average |ΔDmean| of parotid glands in HS was 1.40 Gy, lower than that obtained with DIR and DLC (3.64 Gy, p < 0.001 and 3.72 Gy, p < 0.001, respectively). DLC showed the highest |ΔDmean| in middle Pharyngeal Constrictor Muscle (PCM) (5.13 Gy, p = 0.01). DIR showed second highest |ΔDmean| in the cricopharyngeal inlet (2.85 Gy, p = 0.01). The semi auto-segmentation (SAS) adopted HS, DIR and DLC for segmentation of parotid glands, PCM and all other OARs, respectively. The 90th percentile |ΔNTCP|was 2.19%, 2.24%, 1.10% and 1.50% for DIR, DLC, HS and SAS respectively. CONCLUSIONS: Human segmentation of the parotid glands remains necessary for accurate interpretation of mean dose and NTCP during ART. Proposed semi auto-segmentation allows NTCP-predictions within 1.5% accuracy for 90% of the cases.

Validation of separate multi-atlases for auto segmentation of cardiac substructures in CT-scans acquired in deep inspiration breath hold and free breathing.

BACKGROUND AND PURPOSE: Developing NTCP-models for cardiac complications after breast cancer (BC) radiotherapy requires cardiac dose-volume parameters for many patients. These can be obtained by using multi-atlas based automatic segmentation (MABAS) of cardiac structures in planning CT scans. We investigated the relevance of separate multi-atlases for deep inspiration breath hold (DIBH) and free breathing (FB) CT scans. MATERIALS AND METHODS: BC patients scanned in DIBH (n = 10) and in FB (n = 20) were selected to create separate multi-atlases consisting of expert panel delineations of the whole heart, atria and ventricles. The accuracy of atlas-generated contours was validated with expert delineations in independent datasets (n = 10 for DIBH and FB) and reported as Dice coefficients, contour distances and dose-volume differences in relation to interobserver variability of manual contours. Dependency of MABAS contouring accuracy on breathing technique was assessed by validation of a FB atlas in DIBH patients and vice versa (cross-validation). RESULTS: For all structures the FB and DIBH atlases resulted in Dice coefficients with their respective reference contours ≥ 0.8 and average contour distances ≤ 2 mm smaller than slice thickness of (CTs). No significant differences were found for dose-volume parameters in volumes receiving relevant dose levels (WH, LV and RV). Accuracy of the DIBH atlas was at least similar to, and for the ventricles better than, the interobserver variation in manual delineation. Cross-validation between breathing techniques showed a reduced MABAS performance. CONCLUSION: Multi-atlas accuracy was at least similar to interobserver delineation variation. Separate atlases for scans made in DIBH and FB could benefit atlas performance because accuracy depends on breathing technique.

Assessment of a diaphragm override strategy for robustly optimized proton therapy planning for esophageal cancer patients.

PURPOSE: To ensure target coverage in the treatment of esophageal cancer, a density override to the region of diaphragm motion can be applied in the optimization process. Here, we evaluate the benefit of this approach during robust optimization for intensity modulated proton therapy (IMPT) planning. MATERIALS AND METHODS: For 10 esophageal cancer patients, two robustly optimized IMPT plans were created either using (WDO) or not using (NDO) a diaphragm density override of 1.05 g/cm3 during plan optimization. The override was applied to the excursion of the diaphragm between exhale and inhale. Initial robustness evaluation was performed for plan acceptance (setup errors of 8 mm, range errors of ±3%), and subsequently, on all weekly repeated 4DCTs (setup errors of 2 mm, range errors of ±3%). Target coverage and hotspots were analyzed on the resulting voxel-wise minimum (Vwmin ) and voxel-wise maximum (Vwmax ) dose distributions. RESULTS: The nominal dose distributions were similar for both WDO and NDO plans. However, visual inspection of the Vwmax of the WDO plans showed hotspots behind the right diaphragm override region. For one patient, target coverage and hotspots improved by applying the diaphragm override. We found no differences in target coverage in the weekly evaluations between the two approaches. CONCLUSION: The diaphragm override approach did not result in a clinical benefit in terms of planning and interfractional robustness. Therefore, we do not see added value in employing this approach as a default option during robust optimization for IMPT planning in esophageal cancer.

Estimating the real-world performance of the PROMISE minimal-risk tool.

BACKGROUND: Stable chest pain is a common indication for cardiac catheterization. We assessed the prognostic value of the Prospective Multicenter Imaging Study for Evaluation (PROMISE) Minimal-Risk Tool in identifying patients who are at very low risk of obstructive coronary artery disease (CAD) or downstream cardiovascular adverse outcomes. METHODS: We applied the PROMISE Minimal-Risk Tool to consecutive patients without known CAD who underwent elective cardiac catheterization for stable angina from January 1, 2000 to December 31, 2014 in the Duke Databank for Cardiovascular Disease (DDCD). Patients with scores >0.46 (top decile of lowest-risk from the PROMISE cohort) were classified as low-risk. Logistic regression modeling compared likelihood of freedom from obstructive coronary artery disease on index angiography, 2-year survival, and 2-year survival free of myocardial infarction (MI) and MI/revascularization between low- and non low-risk patients. Alternative cut points to define low- risk patients were also explored. RESULTS: Among 6251 patients undergoing cardiac catheterization for stable chest pain, 1082 (17.3%) were low-risk per the PROMISE minimal-risk tool. Among low risk patients, obstructive coronary artery disease was observed in 14.9% and left main disease (≥ 50% Stenosis) was rare (0.9%). Compared with other patients, low risk patients had a higher likelihood of freedom from obstructive coronary disease on index catheterization (85.1% vs. 44.2%, OR 4.84, 95% CI 4.06-5.77). Low risk patients had significantly higher survival (98.2% vs. 94.4%, OR 3.18, 95% CI 1.99-5.08), MI-free survival (97.2% vs. 91.9%, OR 3.03, 95% CI 2.07-4.45), and MI/revascularization-free survival (86.2 vs. 59.9%, OR 4.19, 95% CI 3.48-5.05) at 2 years than non-low risk patients. Operating characteristics for predicting the outcomes of interest varied modestly depending on the low-risk cut-point used but the positive predictive value for 2 year freedom from death was >98% regardless. CONCLUSION: The PROMISE minimal-risk tool identifies 17% of stable chest pain patients referred to cardiac catheterization as low risk. These patients have a low prevalence of obstructive CAD and better survival than non-low risk patients. While this suggests that these patients are unlikely to benefit from catheterization, further research is needed to confirm a favorable downstream prognosis with medical management alone.

The Importance of Radiation Dose to the Atherosclerotic Plaque in the Left Anterior Descending Coronary Artery for Radiation-Induced Cardiac Toxicity of Breast Cancer Patients?

PURPOSE: Radiation-induced acute coronary events (ACEs) may occur as a treatment-related late adverse effect of breast cancer (BC) radiation. However, the underlying mechanisms behind this radiation-induced cardiac disease remain to be determined. The objective of this study was to test the hypothesis that radiation dose to calcified atherosclerotic plaques in the left anterior descending coronary artery (LAD) is a better predictor for ACEs than radiation dose to the whole heart or left ventricle in patients with BC treated with radiation therapy. METHODS AND MATERIALS: The study cohort consisted of 910 patients with BC treated with postoperative radiation therapy after breast-conserving surgery. In total, 163 patients had an atherosclerotic plaque in the LAD. The endpoint was the occurrence of an ACE after treatment. For each individual patient, the mean heart dose, volume of the left ventricle receiving ≥5 Gy (LV-V5), mean LAD dose, and mean dose to calcified atherosclerotic plaques in the LAD, if present, were acquired based on planning computed tomography scans. Cox regression analysis was used to analyze the effects on the cumulative incidence of ACEs. RESULTS: The median follow-up time was 9.2 years (range, 0.1-14.3 years). In total, 38 patients (4.2%) developed an ACE during follow-up. For patients with an atherosclerotic plaque (n = 163), the mean dose to the atherosclerotic plaque was the strongest predictor for ACEs, even after correction for cardiovascular risk factors (hazard ratio [HR], 1.269; 95% CI, 1.090-1.477; P = .002). The LV-V5 was associated with ACEs in patients without atherosclerotic plaques in the LAD (n = 680) (HR, 1.021; 95% CI, 1.003-1.039; P = .023). CONCLUSIONS: The results of this study suggest that radiation dose to pre-existing calcified atherosclerotic plaques in the LAD is strongly associated with the development of ACEs in patients with BC.

Towards the clinical implementation of intensity-modulated proton therapy for thoracic indications with moderate motion: Robust optimised plan evaluation by means of patient and machine specific information.

PURPOSE: Compared to volumetric modulated arc therapy (VMAT), clinical benefits are anticipated when treating thoracic tumours with intensity-modulated proton therapy (IMPT). However, the current concern of plan robustness as a result of motion hampers its wide clinical implementation. To define an optimal protocol to treat lung and oesophageal cancers, we present a comprehensive evaluation of IMPT planning strategies, based on patient 4DCTs and machine log files. MATERIALS AND METHODS: For ten lung and ten oesophageal cancer patients, a planning 4DCT and weekly repeated 4DCTs were collected. For these twenty patients, the CTV volume and motion were assessed based on the 4DCTs. In addition to clinical VMAT plans, layered rescanned 3D and 4D robust optimised IMPT plans (IMPT_3D and IMPT_4D respectively) were generated, and approved clinically, for all patients. The IMPT plans were then delivered in dry runs at our proton facility to obtain log files, and subsequently evaluated through our 4D robustness evaluation method (4DREM). With this method, for each evaluated plan, fourteen 4D accumulated scenario doses were obtained, representing 14 possible fractionated treatment courses. RESULTS: From VMAT to IMPT_3D, nominal Dmean(lungs-GTV) decreased 2.75 ± 0.56 GyRBE and 3.76 ± 0.92 GyRBE over all lung and oesophageal cancer patients, respectively. A more pronounced reduction was verified for Dmean(heart): 5.38 ± 7.36 GyRBE (lung cases) and 9.51 ± 2.25 GyRBE (oesophagus cases). Target coverage robustness of IMPT_3D was sufficient for 18/20 patients. Averaged dose in critical structures over all 4DREM scenarios changed only slightly for both IMPT_3D and IMPT_4D. Relative to IMPT_3D, no gain in IMPT_4D was observed. CONCLUSION: The dosimetric superiority of IMPT over VMAT has been established. For most thoracic tumours, our IMPT_3D planning protocol showed to be robust and clinically suitable. Nevertheless, accurate patient positioning and adapting to anatomical variations over the course of treatment remain compulsory.