Head and neck cancers (HNC) represent a significant global health burden, with accurate tumor delineation being essential for effective radiotherapy planning. The complexity of the oropharyngeal anatomy, combined with the heterogeneous appearance of tumors on imaging, makes manual segmentation time-intensive and subject to inter-observer variability. Beyond segmentation, predicting long-term clinical outcomes, such as recurrence-free survival (RFS), and determining human papillomavirus (HPV) status from noninvasive imaging, remain challenging yet clinically valuable goals. The HECKTOR 2025 challenge addresses these needs by establishing a comprehensive benchmark for automated HNC analysis using multimodal PET/CT imaging and electronic health records. Building on previous editions (2020-2022), this challenge features an expanded multi-institutional dataset comprising over 1,100 patients from 10 centers worldwide. Participants were tasked with three complementary objectives: (1) segmenting primary gross tumor volumes (GTVp) and metastatic lymph nodes (GTVn), (2) predicting recurrence-free survival, and (3) classifying HPV status. The challenge attracted 35 registered teams, with 15 final submissions evaluated on a held-out test set. Top-performing algorithms achieved a mean Dice similarity coefficient of 0.75 for segmentation, a concordance index of 0.66 for survival prediction, and a balanced accuracy of 0.56 for HPV classification. This paper presents a comprehensive analysis of the submitted methodologies, evaluates their performance across different lesion characteristics, and discusses their implications for clinical translation in automated oncology workflows and decision support systems.

Reconstructive surgery with a flap makes the definition of postoperative radiotherapy volumes challenging. It may also result in errors in automatic segmentation atlases of organ-at-risk and nodal volumes. Automating flap segmentation process could assist clinicians in planning radiotherapy and enable characterization of flap evolution over time and after radiotherapy. Flaps vary significantly in shape, volumes and associated artefacts. We therefore enriched a previously built training dataset with challenging cases to obtain a more robust real-world flap segmentation. Within the framework of the state-of-the-art nnU-Net deep learning architecture, we investigated whether constructing a training dataset with enhanced representation of challenging cases, often associated with poor segmentation performance or outright failures, could improve the overall accuracy and robustness of automated flap segmentation, based on Dice scores compared through paired Wilcoxon signed-rank tests. Clinical trial and real-world data were selected to increase the heterogeneity and enrich the training set with rare challenging cases (such as pedicled flaps, small flaps, unusual location including maxillary flaps, bone resection, presence of dental artefacts or bite block). This enriched training dataset led to improved performance of the nnU-Net model, increasing the mean Dice scores from 0.66 ± 0.29 to 0.74 ± 0.20 (p < 0.001), with median Dice scores rising from 0.76 to 0.80. Robust flap segmentation was achieved without modifying the neural network architecture, loss function, or algorithmic structure, through enrichment of the training set with anatomically and visually challenging cases. This model can be used for detailed analysis of geometrical and textural flap changes over time.

Background and objective: Chest radiography is a medical imaging technique widely used to diagnose thoracic diseases. However, X-ray images may contain artifacts such as irrelevant objects, medical devices, wires and electrodes that can introduce unnecessary noise, making difficult the distinction of relevant anatomical structures, and hindering accurate diagnoses. We aim in this study to address the issue of these artifacts in order to improve lung diseases classification results. Methods: In this paper we present a novel preprocessing approach which begins by detecting images that contain artifacts and then we reduce the artifacts’ noise effect by generating sharper images using a CycleGAN model. The DenseNet-121 model, used for the classification, incorporates channel and spatial attention mechanisms to specifically focus on relevant parts of the image. Additional information contained in the dataset, namely clinical characteristics, were also integrated into the model. Results: We evaluated the performance of the classification model before and after applying our proposed artifact preprocessing approach. These results clearly demonstrate that our preprocessing approach significantly improves the model’s AUC by 5.91% for pneumonia and 6.44% for consolidation classification, outperforming previous studies for the 14 diseases in the ChestX-Ray14 dataset. Conclusion: This research highlights the importance of considering the presence of artifacts when diagnosing lung diseases from radiographic images. By eliminating unwanted noise, our approach enables models to focus on relevant diagnostic features, thereby improving their performance. The results demonstrated that our approach is promising, highlighting its potential for broader applications in lung disease classification.

Introduction To date, small, imbalanced datasets are considered challenging to efficiently train deep learning (DL) models, especially in the medical domain. Consistently, most Artificial Intelligence (AI) approaches in conjunction with small datasets rely on shallow radiomics where traditional machine learning (ML) is utilized for analysing image-derived features. In this study, we evaluate a recently introduced spatial neural network scheme called Distance-Encoding Biomorphic-Informational Neural Network (DEBI-NN), which trains spatial coordinates of artificial neuron coordinates instead of weights, that are then calculated from neuron distances. This technique dramatically reduces the number of parameters to train, thereby making DEBI-NN eligible for the analysis of small, imbalanced datasets. We refer to this property as spatial plasticity. We hypothesized that DEBI-NNs could systematically outperform baseline NN models in small clinical datasets while requiring less regularizations to be implemented, as spatial plasticity may have self-regularization properties. To test our hypothesis, we aimed to compare DEBI-NNs with baseline NNs while relying on various regularization techniques to investigate how DEBI-NNs perform in the presence of regularizers in small multi-centric medical imaging datasets. Methods Three multi-centric datasets were collected including diffuse large B-cell lymphoma (DLBCL) [18F]FDG positron emission tomography (PET)/computed tomography (CT) with clinical parameters to predict 2-years event-free survival; the head and neck [18F]FDG PET/CT dataset from the 2022 MICCAI challenge (HECKTOR), predicting human papillomavirus status; and [68Ga]Ga-PSMA-11 (PSMA-11) PET/CT as well as PSMA-11 PET/magnetic resonance imaging (MRI) cases to predict histopathology-provided International Society of Urological Pathology (ISUP) grades as low (ISUP ≤2) and high (ISUP >2) risk. Per cohort, 5 different network configurations having 1, 2 and 3 hidden layers and neuron count configurations were defined. Per configuration, DEBI-NNs had 7 regularization techniques and baseline NNs had 6 regularization configurations, totalling 27 = 128 and 26 = 64 regularization variants per network scheme to train and evaluate. Test balanced accuracy (BACC) was measured for each model and correlation of the test BACC in the presence of regularization techniques was evaluated in DEBI-NN and baseline NN models. Results The best-performing DEBI-NN models yielded 84.5 %, 80 % and 80.5 % BACC in DLBCL, HECKTOR and PROSTATE datasets, respectively. In contrast, the highest-performing baseline NN models yielded 71.9 %, 77.3 % and 77.3 % BACC in the same cohorts, respectively. In addition, baseline NNs required the implementation of more regularization techniques to increase test BACC from an average test BACC of 53 % (no regularization) to 60 % (6 regularizations), while DEBI-NNs needed no regularization to achieve 62 % BACC. In return, DEBI-NN BACC monotonously fell down to 56 % BACC as the number of regularizations increased. Conclusions DEBI-NNs exhibit a significantly simpler training complexity compared to baseline NNs, while they also outperform baseline NNs with the presence of minimal or no regularization techniques. Our results strongly imply that DEBI-NNs have a potential to pave the way for the utilization of neural networks in small and imbalanced medical datasets, which the field of medical imaging research routinely operates with.

U-Net is a commonly used model for medical image segmentation. However, when applied to chest X-ray images that show pathologies, it often fails to include these critical pathological areas in the generated masks. To address this limitation, in our study, we tackled the challenge of precise segmentation and mask generation by developing a novel approach, using CycleGAN, that encompasses the areas affected by pathologies within the region of interest, allowing the extraction of relevant radiomic features linked to pathologies. Furthermore, we adopted a feature selection approach to focus the analysis on the most significant features. The results of our proposed pipeline are promising, with an average accuracy of 92.05% and an average AUC of 89.48% for the multi-label classification of effusion and infiltration acquired from the ChestX-ray14 dataset, using the XGBoost model. Furthermore, applying our methodology to the classification of the 14 diseases in the ChestX-ray14 dataset resulted in an average AUC of 83.12%, outperforming previous studies. This research highlights the importance of effective pathological mask generation and features selection for accurate classification of chest diseases. The promising results of our approach underscore its potential for broader applications in the classification of chest diseases.

We present a publicly available multimodal dataset for head and neck cancer research, comprising 1123 annotated Positron Emission Tomography/Computed Tomography (PET/CT) studies from patients with histologically confirmed disease, acquired from 10 international medical centers. All studies contain co-registered PET/CT scans with varying acquisition protocols, reflecting real-world clinical diversity from a long-term, multi-institution retrospective collection. Primary gross tumor volumes (GTVp) and involved lymph nodes (GTVn) were manually segmented by experienced radiation oncologists and radiologists following established guidelines. We provide anonymized NifTi files, expert-annotated segmentation masks, comprehensive clinical metadata, and radiotherapy dose distributions for a patient subset. The metadata include TNM staging, HPV status, demographics, long-term follow-up outcomes, survival times, censoring indicators, and treatment information. To demonstrate its utility, we benchmark three key clinical tasks: automated tumor segmentation, recurrence-free survival prediction, and HPV status classification, using state-of-the-art deep learning models like UNet, SegResNet, and multimodal prognostic frameworks.

Reconstructive flap surgery aims to restore the substance and function losses associated with tumor resection. Automatic flap segmentation could allow quantification of flap volume and correlations with functional outcomes after surgery or post-operative RT (poRT). Flaps being ectopic tissues of various components (fat, skin, fascia, muscle, bone) of various volume, shape and texture, the anatomical modifications, inflammation and edema of the postoperative bed make the segmentation task challenging. We built a artificial intelligence-enabled automatic soft-tissue flap segmentation method from CT scans of Head and Neck Cancer (HNC) patients. Ground-truth flap segmentation masks were delineated by two experts on postoperative CT scans of 148 HNC patients undergoing poRT. All CTs and flaps (free or pedicled, soft tissue only or bone) were kept, including those with artefacts, to ensure generalizability. A deep-learning nnUNetv2 framework was built using Hounsfield Units (HU) windowing to mimic radiological assessment. A transformer-based 2D "Segment Anything Model" (MedSAM) was also built and fine-tuned to medical CTs. Models were compared with the Dice Similarity Coefficient (DSC) and Hausdorff Distance 95th percentile (HD95) metrics. Flaps were in the oral cavity (N = 102), oropharynx (N = 26) or larynx/hypopharynx (N = 20). There were free flaps (N = 137), pedicled flaps (N = 11), of soft tissue flap-only (N = 92), reconstructed bone (N = 42), or bone resected without reconstruction (N = 40). The nnUNet-windowing model outperformed the nnUNetv2 and MedSam models. It achieved mean DSCs of 0.69 and HD95 of 25.6 mm using 5-fold cross-validation. Segmentation performed better in the absence of artifacts, and rare situations such as pedicled flaps, laryngeal primaries and resected bone without bone reconstruction (p < 0.01). Automatic flap segmentation demonstrates clinical performances that allow to quantify spontaneous and radiation-induced volume shrinkage of flaps. Free flaps achieved excellent performances; rare situations will be addressed by fine-tuning the network.

Multi-centre variability and HPV-related heterogeneity limit the reliability of PET/CT radiomics for survival prediction in head-and-neck cancer. The research investigates whether different approaches to handle HPV information and feature distribution alignment between centers will enhance performance on external data. The public HECKTOR 2022 cohort with predefined training and test splits served as the basis for our analysis which evaluated CT, PET and early CT+PET fusion with and without harmonization under three HPV strategies (ignored, covariate, stratified). The performance of PET radiomics remained stable when used outside of the centre while harmonization produced neutral or beneficial results for PET but produced mixed effects for CT. PET alone received minimal benefit from the implementation of simple early fusion. The model covariate approach to handle HPV produced unpredictable performance improvements but stratified reporting exposed cohort variations although it was limited by the small number of HPV-negative test cases. The deployment of harmonized PET radiomics provides a practical robust baseline for multi-centre implementation and HPV information proves most useful for transparent subgroup reporting and study design. Future research needs to evaluate calibration methods while verifying these results in external cohorts that have balanced HPV status.

Purpose/Objective(s) The PAIR prostate study aims to demonstrate the dosimetric feasibility of tumoral and hypoxia-guided dose escalated radiation therapy in prostate cancer on a novel ring gantry system TPS, based on conventionally fractionated and hypofractionated regimens. Materials/Methods The GTV was delineated on MRI and the hypoxic region (biological tumor volume [BTV]) within the PTV on F-MISO PET uptake was delineated on pre-therapeutic F-MISO PET scans of 7 patients showing uptake in the tumor before undergoing external-beam RT for intermediate-risk prostate cancer. IMRT planning was performed using three different prescriptions regimen 1) Standard fractionation (SF) 77Gy/35 fractions (fx) to PTV, with a boost to 95Gy and 118Gy in 35 fx to the GTV and to the hypoxic region defined by F-MISO PET, respectively, 2) moderate hypofractionation (MH) 60 Gy/20 fx with a boost to 67Gy and 91Gy in 20 fx to the GTV and to the hypoxic region, respectively, and 3) high hypofractionation (HH) 40 Gy/5 fx with a boost to 50Gy in 5 fx to the GTV and as high as possible to the hypoxic region (SBRT). Planning was performed on a treatment planning system. Results The average size of the GTV and BTV was 11.4 ± 11.3 cc, and 1.30 ± 1.20 cc, respectively. For the SF regimen, the average D98% and D2% to the PTV, GTV and BTV were 71.3 ± 11.7 Gy and 105.0 ± 7.6 Gy; 90.3 ± 6.9 Gy and 113.2 ± 6.9 Gy; and 107.4 ± 5.6 Gy, and 116.0 ± 5.6 Gy, respectively. For the MH and HH techniques, the average D98% to the PTV was 54.3 ± 2.5 Gy, and 31.7 ± 5.3 Gy, and the average D2% was 83.6 ± 7.0 Gy, and 55.8 ± 2.9 Gy. In addition, the average D98%, and D2% to the GTV was 61.2 ± 6.2 Gy, and 89.8 ± 5.9 Gy using the MH treatment technique, while the average D98%, and D2% to the BTV was 85.2 ± 4.1 Gy, and 92.3 ± 3.2 Gy, respectively. Also, for the HH treatment planning method, the average D98%, and D2% to the GTV was 45.7 ± 4.9 Gy, and 59.2 ± 3.2 Gy, while the average D98%, and D2% to the BTV was 53.6 ± 4.2 Gy and 60.6 ± 2.5 Gy, respectively. The average D50% of the bladder, rectum, and femoral heads was 12.6 ± 5.4 Gy, 24.8 ± 9.9 Gy, and 9.6 ± 2.5 Gy, while the average V15Gy was 43.0 ± 11.5%, 59.4 ± 12.5 %, and 12.5 ± 14.1%, respectively. Conclusion This study showed the dosimetric feasibility of dose-escalating the tumor and the hypoxic region in patients with prostate cancer using standard, moderate and high hypofractionated regimen on a treatment planning system while keeping the doses within acceptable limits for the organs at risk.

The purpose of this survey is to provide a comprehensive review of the most recent publications on lung disease classification from chest X-ray images using deep learning algorithms. Methods: This research aims to present several common chest radiography datasets and to introduce briefly the general image preprocessing procedures that are applied to chest X-ray images. Then, the classification of specific and multiple lung diseases is described, focusing on the method and dataset used in the selected studies, the evaluation measures and the results. In addition, the problems and future direction of lung diseases classification are discussed to provide an important research base for researchers in the future. As the most common examination tool, Chest X-ray (CXR) is crucial in the medical field for disease diagnosis. Thus, the classification of chest diseases based on chest X-ray has gained significant attention from researchers. In recent years, deep learning methods have been used and have emerged as powerful techniques in medical imaging fields. One hundred ten articles published from 2016 to 2023 were reviewed and summarized, confirming that this particular research area is very important and has great potential for future research.