Bi-parametric magnetic resonance imaging (bpMRI) has become a pivotal modality in the detection and diagnosis of clinically significant prostate cancer (csPCa). Developing AI-based systems to identify csPCa using bpMRI can transform prostate cancer (PCa) management by improving efficiency and cost-effectiveness. However, current state-of-the-art methods using convolutional neural networks (CNNs) and Transformers are limited in learning in-plane and three-dimensional spatial information from anisotropic bpMRI. Their performances also depend on the availability of large, diverse, and well-annotated bpMRI datasets. To address these challenges, we propose the Zonal-aware Self-supervised Mesh Network (Z-SSMNet), which adaptively integrates multi-dimensional (2D/2.5D/3D) convolutions to learn dense intra-slice information and sparse inter-slice information of the anisotropic bpMRI in a balanced manner. We also propose a self-supervised learning (SSL) technique that effectively captures both intra-slice and inter-slice semantic information using large-scale unlabeled data. Furthermore, we constrain the network to focus on the zonal anatomical regions to improve the detection and diagnosis capability of csPCa. We conducted extensive experiments on the PI-CAI (Prostate Imaging - Cancer AI) dataset comprising 10000+ multi-center and multi-scanner data. Our Z-SSMNet excelled in both lesion-level detection (AP score of 0.633) and patient-level diagnosis (AUROC score of 0.881), securing the top position in the Open Development Phase of the PI-CAI challenge and maintained strong performance, achieving an AP score of 0.690 and an AUROC score of 0.909, and securing the second-place ranking in the Closed Testing Phase. These findings underscore the potential of AI-driven systems for csPCa diagnosis and management.

Sarcopenia is associated with worse survival in non-small cell lung cancer (NSCLC), but less studied in association with toxicity. Here, we investigated the association between imaging-assessed sarcopenia with toxicity in patients with NSCLC. We analyzed a "chemoradiation" cohort (n = 318) of patients with NSCLC treated with chemoradiation, and an external validation "chemo-surgery" cohort (n = 108) who were treated with chemotherapy and surgery from 2002 to 2013 at a different institution. A deep-learning pipeline utilized pretreatment computed tomography scans to estimate SM area at the third lumbar vertebral level. Sarcopenia was defined by dichotomizing SM index, (SM adjusted for height and sex). Primary endpoint was NCI CTCAE v5.0 grade 3 to 5 (G3-5) toxicity within 21-days of first chemotherapy cycle. Multivariable analyses (MVA) of toxicity endpoints with sarcopenia and baseline characteristics were performed by logistic regression, and overall survival (OS) was analyzed using Cox regression. Sarcopenia was identified in 36% and 36% of patients in the chemoradiation and chemo-surgery cohorts, respectively. On MVA, sarcopenia was associated with worse G3-5 toxicity in chemoradiation (HR 2.00, P < .01) and chemo-surgery cohorts (HR 2.95, P = .02). In the chemoradiation cohort, worse OS was associated with G3-5 toxicity (HR 1.42, P = .02) but not sarcopenia on MVA. In chemo-surgery cohort, worse OS was associated with sarcopenia (HR 2.03, P = .02) but not G3-5 toxicity on MVA. Sarcopenia, assessed by an automated deep-learning system, was associated with worse toxicity and survival outcomes in patients with NSCLC. Sarcopenia can be utilized to tailor treatment decisions to optimize adverse events and survival.

Cerebrovascular diseases (CVD) can lead to stroke and dementia. Stroke is the second leading cause of death world wide and dementia incidence is increasing by the year. There are several markers of CVD that are visible on brain imaging, including: white matter hyperintensities (WMH), acute and chronic ischaemic stroke lesions (ISL), lacunes, enlarged perivascular spaces (PVS), acute and chronic haemorrhagic lesions, and cerebral microbleeds (CMB). Brain atrophy also occurs in CVD. These markers are important for patient management and intervention, since they indicate elevated risk of future stroke and dementia. We systematically reviewed automated systems designed to support radiologists reporting on these CVD imaging findings. We considered commercially available software and research publications which identify at least two CVD markers. In total, we included 29 commercial products and 13 research publications. Two distinct types of commercial support system were available: those which identify acute stroke lesions (haemorrhagic and ischaemic) from computed tomography (CT) scans, mainly for the purpose of patient triage; and those which measure WMH and atrophy regionally and longitudinally. In research, WMH and ISL were the markers most frequently analysed together, from magnetic resonance imaging (MRI) scans; lacunes and PVS were each targeted only twice and CMB only once. For stroke, commercially available systems largely support the emergency setting, whilst research systems consider also follow-up and routine scans. The systems to quantify WMH and atrophy are focused on neurodegenerative disease support, where these CVD markers are also of significance. There are currently no openly validated systems, commercially, or in research, performing a comprehensive joint analysis of all CVD markers (WMH, ISL, lacunes, PVS, haemorrhagic lesions, CMB, and atrophy).

Accurate brain tumor segmentation from MRI images is critical in the medical industry, directly impacts the efficacy of diagnostic and treatment plans. Accurate segmentation of tumor region can be challenging, especially when noise and abnormalities are present. This research provides a systematic review of automatic brain tumor segmentation techniques, with a specific focus on the design of network architectures. The review categorizes existing methods into unsupervised and supervised learning techniques, as well as machine learning and deep learning approaches within supervised techniques. Deep learning techniques are thoroughly reviewed, with a particular focus on CNN-based, U-Net-based, transfer learning-based, transformer-based, and hybrid transformer-based methods. This survey encompasses a broad spectrum of automatic segmentation methodologies, from traditional machine learning approaches to advanced deep learning frameworks. It provides an in-depth comparison of performance metrics, model efficiency, and robustness across multiple datasets, particularly the BraTS dataset. The study further examines multi-modal MRI imaging and its influence on segmentation accuracy, addressing domain adaptation, class imbalance, and generalization challenges. The analysis highlights the current challenges in Computer-aided Diagnostic (CAD) systems, examining how different models and imaging sequences impact performance. Recent advancements in deep learning, especially the widespread use of U-Net architectures, have significantly enhanced medical image segmentation. This review critically evaluates these developments, focusing the iterative improvements in U-Net models that have driven progress in brain tumor segmentation. Furthermore, it explores various techniques for improving U-Net performance for medical applications, focussing on its potential for improving diagnostic and treatment planning procedures. The efficiency of these automated segmentation approaches is rigorously evaluated using the BraTS dataset, a benchmark dataset, part of the annual Multimodal Brain Tumor Segmentation Challenge (MICCAI). This evaluation provides insights into the current state-of-the-art and identifies key areas for future research and development.

In the HECKTOR 2022 challenge set [1], several state-of-the-art (SOTA, achieving best performance) deep learning models were introduced for predicting recurrence-free period (RFP) in head and neck cancer patients using PET and CT images. This study investigates whether a conventional DenseNet architecture, with optimized numbers of layers and image-fusion strategies, could achieve comparable performance as SOTA models. The HECKTOR 2022 dataset comprises 489 oropharyngeal cancer (OPC) patients from seven distinct centers. It was randomly divided into a training set (n = 369) and an independent test set (n = 120). Furthermore, an additional dataset of 400 OPC patients, who underwent chemo(radiotherapy) at our center, was employed for external testing. Each patients' data included pre-treatment CT- and PET-scans, manually generated GTV (Gross tumour volume) contours for primary tumors and lymph nodes, and RFP information. The present study compared the performance of DenseNet against three SOTA models developed on the HECKTOR 2022 dataset. When inputting CT, PET and GTV using the early fusion (considering them as different channels of input) approach, DenseNet81 (with 81 layers) obtained an internal test C-index of 0.69, a performance metric comparable with SOTA models. Notably, the removal of GTV from the input data yielded the same internal test C-index of 0.69 while improving the external test C-index from 0.59 to 0.63. Furthermore, compared to PET-only models, when utilizing the late fusion (concatenation of extracted features) with CT and PET, DenseNet81 demonstrated superior C-index values of 0.68 and 0.66 in both internal and external test sets, while using early fusion was better in only the internal test set. The basic DenseNet architecture with 81 layers demonstrated a predictive performance on par with SOTA models featuring more intricate architectures in the internal test set, and better performance in the external test. The late fusion of CT and PET imaging data yielded superior performance in the external test.

Patients with metastatic bone disease are at risk of pathological femoral fractures and may require prophylactic surgical fixation. Current clinical decision support tools often overestimate fracture risk, leading to overtreatment. While novel scores integrating femoral strength assessment via finite element (FE) models show promise, they require 3D imaging, extensive computation, and are difficult to automate. Predicting femoral strength directly from single 2D radiographic projections using convolutional neural networks (CNNs) could address these limitations, but this approach has not yet been explored for femora with metastatic lesions. This study aimed to test whether CNNs can accurately predict strength of femora with metastatic lesions from single 2D radiographic projections. CNNs with various architectures were developed and trained using an FE model generated training dataset. This training dataset was based on 36,000 modified computed tomography (CT) scans, created by randomly inserting artificial lytic lesions into the CT scans of 36 intact anatomical femoral specimens. From each modified CT scan, an anterior-posterior 2D projection was generated and femoral strength in one-legged stance was determined using nonlinear FE models. Following training, the CNN performance was evaluated on an independent experimental test dataset consisting of 31 anatomical femoral specimens (16 intact, 15 with artificial lytic lesions). 2D projections of each specimen were created from corresponding CT scans and femoral strength was assessed in mechanical tests. The CNNs' performance was evaluated using linear regression analysis and compared to 2D densitometric predictors (bone mineral density and content) and CT-based 3D FE models. All CNNs accurately predicted the experimentally measured strength in femora with and without metastatic lesions of the test dataset (R²≥0.80, CCC≥0.81). In femora with metastatic lesions, the performance of the CNNs (best: R²=0.84, CCC=0.86) was considerably superior to 2D densitometric predictors (R²≤0.07) and slightly inferior to 3D FE models (R²=0.90, CCC=0.94). CNNs, trained on a large dataset generated via FE models, predicted experimentally measured strength of femora with artificial metastatic lesions with accuracy comparable to 3D FE models. By eliminating the need for 3D imaging and reducing computational demands, this novel approach demonstrates potential for application in a clinical setting.

Computer-aided diagnosis systems based on deep neural networks heavily rely on datasets with high-quality labels. However, manual annotation for lesion diagnosis relies on image features, often requiring professional experience and complex image analysis process. This inevitably introduces noisy labels, which can misguide the training of classification models. Our goal is to design an effective method to address the challenges posed by label noise in medical images. we propose a novel noise-tolerant medical image classification framework consisting of two phases: fore-training correction and progressive hard-sample enhanced learning. In the first phase, we design a dual-branch sample partition detection scheme that effectively classifies each instance into one of three subsets: clean, hard, or noisy. Simultaneously, we propose a hard-sample label refinement strategy based on class prototypes with confidence-perception weighting and an effective joint correction method for noisy samples, enabling the acquisition of higher-quality training data. In the second phase, we design a progressive hard-sample reinforcement learning method to enhance the model's ability to learn discriminative feature representations. This approach accounts for sample difficulty and mitigates the effects of label noise in medical datasets. Our framework achieves an accuracy of 82.39% on the pneumoconiosis dataset collected by our laboratory. On a five-class skin disease dataset with six different levels of label noise (0, 0.05, 0.1, 0.2, 0.3, and 0.4), the average accuracy over the last ten epochs reaches 88.51%, 86.64%, 85.02%, 83.01%, 81.95%, 77.89%, respectively; For binary polyp classification under noise rates of 0.2, 0.3, and 0.4, the average accuracy over the last ten epochs is 97.90%, 93.77%, 89.33%, respectively. The effectiveness of our proposed framework is demonstrated through its performance on three challenging datasets with both real and synthetic noise. Experimental results further demonstrate the robustness of our method across varying noise rates.

Brain Tumor Segmentation (BraTS) plays a critical role in clinical diagnosis, treatment planning, and monitoring the progression of brain tumors. However, due to the variability in tumor appearance, size, and intensity across different MRI modalities, automated segmentation remains a challenging task. In this study, we propose a novel Transformer-based framework, multiPI-TransBTS, which integrates multi-physical information to enhance segmentation accuracy. The model leverages spatial information, semantic information, and multi-modal imaging data, addressing the inherent heterogeneity in brain tumor characteristics. The multiPI-TransBTS framework consists of an encoder, an Adaptive Feature Fusion (AFF) module, and a multi-source, multi-scale feature decoder. The encoder incorporates a multi-branch architecture to separately extract modality-specific features from different MRI sequences. The AFF module fuses information from multiple sources using channel-wise and element-wise attention, ensuring effective feature recalibration. The decoder combines both common and task-specific features through a Task-Specific Feature Introduction (TSFI) strategy, producing accurate segmentation outputs for Whole Tumor (WT), Tumor Core (TC), and Enhancing Tumor (ET) regions. Comprehensive evaluations on the BraTS2019 and BraTS2020 datasets demonstrate the superiority of multiPI-TransBTS over the state-of-the-art methods. The model consistently achieves better Dice coefficients, Hausdorff distances, and Sensitivity scores, highlighting its effectiveness in addressing the BraTS challenges. Our results also indicate the need for further exploration of the balance between precision and recall in the ET segmentation task. The proposed framework represents a significant advancement in BraTS, with potential implications for improving clinical outcomes for brain tumor patients.

The shear wave elastography (SWE) provides quantitative markers for tissue characterization by measuring the shear wave speed (SWS), which reflects tissue stiffness. SWE uses an acoustic radiation force pulse sequence to generate shear waves that propagate laterally through tissue with transient displacements. These waves travel perpendicular to the applied force, and their displacements are tracked using high-frame-rate ultrasound. Estimating the SWS map involves two main steps: speckle tracking and SWS estimation. Speckle tracking calculates particle velocity by measuring RF/IQ data displacement between adjacent firings, while SWS estimation methods typically compare particle velocity profiles of samples that are laterally a few millimeters apart. Deep learning (DL) methods have gained attention for SWS estimation, often relying on supervised training using simulated data. However, these methods may struggle with real-world data, which can differ significantly from the simulated training data, potentially leading to artifacts in the estimated SWS map. To address this challenge, we propose a physics-inspired learning approach that utilizes real data without known SWS values. Our method employs an adaptive unsupervised loss function, allowing the network to train with the real noisy data to minimize the artifacts and improve the robustness. We validate our approach using experimental phantom data and in vivo liver data from two human subjects, demonstrating enhanced accuracy and reliability in SWS estimation compared with conventional and supervised methods. This hybrid approach leverages the strengths of both data-driven and physics-inspired learning, offering a promising solution for more accurate and robust SWS mapping in clinical applications.

The increasing digitalization of multi-modal data in medicine and novel artificial intelligence (AI) algorithms opens up a large number of opportunities for predictive models. In particular, deep learning models show great performance in the medical field. A major limitation of such powerful but complex models originates from their 'black-box' nature. Recently, a variety of explainable AI (XAI) methods have been introduced to address this lack of transparency and trust in medical AI. However, the majority of such methods have solely been evaluated on single data modalities. Meanwhile, with the increasing number of XAI methods, integrative XAI frameworks and benchmarks are essential to compare their performance on different tasks. For that reason, we developed BenchXAI, a novel XAI benchmarking package supporting comprehensive evaluation of fifteen XAI methods, investigating their robustness, suitability, and limitations in biomedical data. We employed BenchXAI to validate these methods in three common biomedical tasks, namely clinical data, medical image and signal data, and biomolecular data. Our newly designed sample-wise normalization approach for post-hoc XAI methods enables the statistical evaluation and visualization of performance and robustness. We found that the XAI methods Integrated Gradients, DeepLift, DeepLiftShap, and GradientShap performed well over all three tasks, while methods like Deconvolution, Guided Backpropagation, and LRP-α1-β0 struggled for some tasks. With acts such as the EU AI Act the application of XAI in the biomedical domain becomes more and more essential. Our evaluation study represents a first step towards verifying the suitability of different XAI methods for various medical domains.