Background: Radiomics can provide in-depth characterization of cancers for treatment outcome prediction. Conventional radiomics rely on extraction of image features within a pre-defined image region of interest (ROI) which are typically fed to a classification algorithm for prediction of a clinical endpoint. Deep learning radiomics allows for a simpler workflow where images can be used directly as input to a convolutional neural network (CNN) with or without a pre-defined ROI.Purpose: The purpose of this study was to evaluate (i) conventional radiomics and (ii) deep learning radiomics for predicting overall survival (OS) and disease-free survival (DFS) for patients with head and neck squamous cell carcinoma (HNSCC) using pre-treatment ¹⁸F-fluorodeoxuglucose positron emission tomography (FDG PET) and computed tomography (CT) images.Materials and methods: FDG PET/CT images and clinical data of patients with HNSCC treated with radio(chemo)therapy at Oslo University Hospital (OUS; n = 139) and Maastricht University Medical Center (MAASTRO; n = 99) were collected retrospectively. OUS data was used for model training and initial evaluation. MAASTRO data was used for external testing to assess cross-institutional generalizability. Models trained on clinical and/or conventional radiomics features, with or without feature selection, were compared to CNNs trained on PET/CT images without or with the gross tumor volume (GTV) included. Model performance was measured using accuracy, area under the receiver operating characteristic curve (AUC), Matthew’s correlation coefficient (MCC), and the F1 score calculated for both classes separately.Results: CNNs trained directly on images achieved the highest performance on external data for both endpoints. Adding both clinical and radiomics features to these image-based models increased performance further. Conventional radiomics including clinical data could achieve competitive performance. However, feature selection on clinical and radiomics data lead to overfitting and poor cross-institutional generalizability. CNNs without tumor and node contours achieved close to on-par performance with CNNs including contours.Conclusion: High performance and cross-institutional generalizability can be achieved by combining clinical data, radiomics features and medical images together with deep learning models. However, deep learning models trained on images without contours can achieve competitive performance and could see potential use as an initial screening tool for high-risk patients.

Target volume delineation is a vital but time-consuming and challenging part of radiotherapy, where the goal is to deliver sufficient dose to the target while reducing risks of side effects. For head and neck cancer (HNC) this is complicated by the complex anatomy of the head and neck region and the proximity of target volumes to organs at risk. The purpose of this study was to compare and evaluate conventional PET thresholding methods, six classical machine learning algorithms and a 2D U-Net convolutional neural network (CNN) for automatic gross tumor volume (GTV) segmentation of HNC in PET/CT images. For the latter two approaches the impact of single versus multimodality input on segmentation quality was also assessed. 197 patients were included in the study. The cohort was split into training and test sets (157 and 40 patients, respectively). Five-fold cross-validation was used on the training set for model comparison and selection. Manual GTV delineations represented the ground truth. Tresholding, classical machine learning and CNN segmentation models were ranked separately according to the cross-validation Sørensen–Dice similarity coefficient (Dice). PET thresholding gave a maximum mean Dice of 0.62, whereas classical machine learning resulted in maximum mean Dice scores of 0.24 (CT) and 0.66 (PET; PET/CT). CNN models obtained maximum mean Dice scores of 0.66 (CT), 0.68 (PET) and 0.74 (PET/CT). The difference in cross-validation Dice between multimodality PET/CT and single modality CNN models was significant (p ≤ 0.0001). The top-ranked PET/CT-based CNN model outperformed the best-performing thresholding and classical machine learning models, giving significantly better segmentations in terms of cross-validation and test set Dice, true positive rate, positive predictive value and surface distance-based metrics (p ≤ 0.0001). Thus, deep learning based on multimodality PET/CT input resulted in superior target coverage and less inclusion of surrounding normal tissue.

Purpose: Identification and delineation of the gross tumour and malignant nodal volume (GTV) in medical images are vital in radiotherapy. We assessed the applicability of convolutional neural networks (CNNs) for fully automatic delineation of the GTV from FDG-PET/CT images of patients with head and neck cancer (HNC). CNN models were compared to manual GTV delineations made by experienced specialists. New structure-based performance metrics were introduced to enable in-depth assessment of auto-delineation of multiple malignant structures in individual patients. Methods: U-Net CNN models were trained and evaluated on images and manual GTV delineations from 197 HNC patients. The dataset was split into training, validation and test cohorts (n= 142, n = 15 and n = 40, respectively). The Dice score, surface distance metrics and the new structure-based metrics were used for model evaluation. Additionally, auto-delineations were manually assessed by an oncologist for 15 randomly selected patients in the test cohort. Results: The mean Dice scores of the auto-delineations were 55%, 69% and 71% for the CT-based, PET-based and PET/CT-based CNN models, respectively. The PET signal was essential for delineating all structures. Models based on PET/CT images identified 86% of the true GTV structures, whereas models built solely on CT images identified only 55% of the true structures. The oncologist reported very high-quality auto-delineations for 14 out of the 15 randomly selected patients. Conclusions: CNNs provided high-quality auto-delineations for HNC using multimodality PET/CT. The introduced structure-wise evaluation metrics provided valuable information on CNN model strengths and weaknesses for multi-structure auto-delineation.

Background: Solid tumors are known to be spatially heterogeneous. Detection of treatment-resistant tumor regions can improve clinical outcome, by enabling implementation of strategies targeting such regions. In this study, K-means clustering was used to group voxels in dynamic contrast enhanced magnetic resonance images (DCE-MRI) of cervical cancers. The aim was to identify clusters reflecting treatment resistance that could be used for targeted radiotherapy with a dose-painting approach. Material and methods: Eighty-one patients with locally advanced cervical cancer underwent DCE-MRI prior to chemoradiotherapy. The resulting image time series were fitted to two pharmacokinetic models, the Tofts model (yielding parameters K^trans and ν_e) and the Brix model (A_Brix, k_ep and k_el). K-means clustering was used to group similar voxels based on either the pharmacokinetic parameter maps or the relative signal increase (RSI) time series. The associations between voxel clusters and treatment outcome (measured as locoregional control) were evaluated using the volume fraction or the spatial distribution of each cluster. Results: One voxel cluster based on the RSI time series was significantly related to locoregional control (adjusted p-value 0.048). This cluster consisted of low-enhancing voxels. We found that tumors with poor prognosis had this RSI-based cluster gathered into few patches, making this cluster a potential candidate for targeted radiotherapy. None of the voxels clusters based on Tofts or Brix parameter maps were significantly related to treatment outcome. Conclusion: We identified one group of tumor voxels significantly associated with locoregional relapse that could potentially be used for dose painting. This tumor voxel cluster was identified using the raw MRI time series rather than the pharmacokinetic maps.

Background: Tumour delineation is a challenging, time-consuming and complex part of radiotherapy planning. In this study, an automatic method for delineating locally advanced cervical cancers was developed using a machine learning approach. Materials and methods: A method for tumour segmentation based on image voxel classification using Fisher?s Linear Discriminant Analysis (LDA) was developed. This was applied to magnetic resonance (MR) images of 78 patients with locally advanced cervical cancer. The segmentation was based on multiparametric MRI consisting of T2- weighted (T2w), T1-weighted (T1w) and dynamic contrast-enhanced (DCE) sequences, and included intensity and spatial information from the images. The model was trained and assessed using delineations made by two radiologists. Results: Segmentation based on T2w or T1w images resulted in mean sensitivity and specificity of 94% and 52%, respectively. Including DCE-MR images improved the segmentation model?s performance significantly, giving mean sensitivity and specificity of 85?93%. Comparisons with radiologists? tumour delineations gave Dice similarity coefficients of up to 0.44. Conclusion: Voxel classification using a machine learning approach is a flexible and fully automatic method for tumour delineation. Combining all relevant MR image series resulted in high sensitivity and specificity. Moreover, the presented method can be extended to include additional imaging modalities.

Tumor delineation is time- and labor-intensive and prone to inter- and intraobserver variations. Magnetic resonance imaging (MRI) provides good soft tissue contrast, and functional MRI captures tis...