Prevalent knee osteoarthritis (OA) imposes substantial burden on health systems with no cure available. Its ultimate treatment is total knee replacement (TKR). Complications from surgery and recovery are difficult to predict in advance, and numerous factors may affect them. Radiographic knee alignment is one of the key factors that impacts TKR outcomes, affecting outcomes such as postoperative pain or function. Recently, artificial intelligence (AI) has been introduced to the automatic analysis of knee radiographs, for example, to automate knee alignment measurements. Existing review articles tend to focus on knee OA diagnosis and segmentation of bones or cartilages in MRI rather than exploring knee alignment biomarkers for TKR outcomes and their assessment. In this review, we first examine the current scoring protocols for evaluating TKR outcomes and potential knee alignment biomarkers associated with these outcomes. We then discuss existing AI-based approaches for generating knee alignment biomarkers from knee radiographs, and explore future directions for knee alignment assessment and TKR outcome prediction.

Explainable Artificial Intelligence (XAI) encompasses a broad spectrum of methods that aim to enhance the transparency of deep learning models, with Class Activation Mapping (CAM) methods widely used for visual interpretability. However, systematic evaluations of these methods in veterinary radiography remain scarce. This study presents a comparative analysis of eleven CAM methods, including GradCAM, XGradCAM, ScoreCAM, and EigenCAM, on a dataset of 7362 canine and feline X-ray images. A ResNet18 model was chosen based on the specificity of the dataset and preliminary results where it outperformed other models. Quantitative and qualitative evaluations were performed to determine how well each CAM method produced interpretable heatmaps relevant to clinical decision-making. Among the techniques evaluated, EigenGradCAM achieved the highest mean score and standard deviation (SD) of 2.571 (SD = 1.256), closely followed by EigenCAM at 2.519 (SD = 1.228) and GradCAM++ at 2.512 (SD = 1.277), with methods such as FullGrad and XGradCAM achieving worst scores of 2.000 (SD = 1.300) and 1.858 (SD = 1.198) respectively. Despite variations in saliency visualization, no single method universally improved veterinarians' diagnostic confidence. While certain CAM methods provide better visual cues for some pathologies, they generally offered limited explainability and didn't substantially improve veterinarians' diagnostic confidence.

In this work, we present GazeLT, a human visual attention integration-disintegration approach for long-tailed disease classification. A radiologist's eye gaze has distinct patterns that capture both fine-grained and coarser level disease related information. While interpreting an image, a radiologist's attention varies throughout the duration; it is critical to incorporate this into a deep learning framework to improve automated image interpretation. Another important aspect of visual attention is that apart from looking at major/obvious disease patterns, experts also look at minor/incidental findings (few of these constituting long-tailed classes) during the course of image interpretation. GazeLT harnesses the temporal aspect of the visual search process, via an integration and disintegration mechanism, to improve long-tailed disease classification. We show the efficacy of GazeLT on two publicly available datasets for long-tailed disease classification, namely the NIH-CXR-LT (n=89237) and the MIMIC-CXR-LT (n=111898) datasets. GazeLT outperforms the best long-tailed loss by 4.1% and the visual attention-based baseline by 21.7% in average accuracy metrics for these datasets. Our code is available at https://github.com/lordmoinak1/gazelt.

This study aimed to systematically review the current performance of large language models (LLMs) in dento-maxillofacial radiology (DMFR). Five electronic databases were used to identify studies that developed, fine-tuned, or evaluated LLMs for DMFR-related tasks. Data extracted included study purpose, LLM type, images/text source, applied language, dataset characteristics, input and output, performance outcomes, evaluation methods, and reference standards. Customized assessment criteria adapted from the TRIPOD-LLM reporting guideline were used to evaluate the risk-of-bias in the included studies specifically regarding the clarity of dataset origin, the robustness of performance evaluation methods, and the validity of the reference standards. The initial search yielded 1621 titles, and nineteen studies were included. These studies investigated the use of LLMs for tasks including the production and answering of DMFR-related qualification exams and educational questions (n = 8), diagnosis and treatment recommendations (n = 7), and radiology report generation and patient communication (n = 4). LLMs demonstrated varied performance in diagnosing dental conditions, with accuracy ranging from 37-92.5% and expert ratings for differential diagnosis and treatment planning between 3.6-4.7 on a 5-point scale. For DMFR-related qualification exams and board-style questions, LLMs achieved correctness rates between 33.3-86.1%. Automated radiology report generation showed moderate performance with accuracy ranging from 70.4-81.3%. LLMs demonstrate promising potential in DMFR, particularly for diagnostic, educational, and report generation tasks. However, their current accuracy, completeness, and consistency remain variable. Further development, validation, and standardization are needed before LLMs can be reliably integrated as supportive tools in clinical workflows and educational settings.

ObjectivePoor outcomes in acute respiratory distress syndrome (ARDS) can be alleviated with tools that support early diagnosis. Current machine learning methods for detecting ARDS do not take full advantage of the multimodality of ARDS pathophysiology. We developed a multimodal deep learning model that uses imaging data, continuously collected ventilation data, and tabular data derived from a patients electronic health record (EHR) to make ARDS predictions. Materials and MethodsA chest radiograph (x-ray), at least two hours of ventilator waveform (VWD) data within the first 24 hours of intubation, and EHR-derived tabular data were used from 220 patients admitted to the ICU to train a deep learning model. The model uses pretrained encoders for the x-rays and ventilation data and trains a feature extractor on tabular data. Encoded features for a patient are combined to make a single ARDS prediction. Ablation studies for each modality assessed their effect on the models predictive capability. ResultsThe trimodal model achieved an area under the receiver operator curve (AUROC) of 0.86 with a 95% confidence interval of 0.01. This was a statistically significant improvement (p<0.05) over single modality models and bimodal models trained on VWD+tabular and VWD+x-ray data. Discussion and ConclusionOur results demonstrate the potential utility of using deep learning to address complex conditions with heterogeneous data. More work is needed to determine the additive effect of modalities on ARDS detection. Our framework can serve as a blueprint for building performant multimodal deep learning models for conditions with small, heterogeneous datasets.

The aim of this study was to evaluate GPT-4o's multimodal reasoning ability to review panoramic radiograph (PR) and verify its radiologic findings, while exploring the role of prompt engineering in enhancing its performance. The study included 230 PRs from West China Hospital of Stomatology in 2024, which were interpreted to generate the PR findings. A total of 300 instances of interpretation errors, were manually inserted into the PR findings. The ablation study was conducted to assess whether GPT-4o can perform reasoning on PR under a zero-shot prompt. Prompt engineering was employed to enhance the reasoning capabilities of GPT-4o in identifying interpretation errors with PRs. The prompt strategies included chain-of-thought, self-consistency, in-context learning, multimodal in-context learning, and their systematic integration into a meta-prompt. Recall, accuracy, and F1 score were employed to evaluate the outputs. Subsequently, the localization capability of GPT-4o and its influence on reasoning capability were evaluated. In the ablation study, GPT-4o's recall increased significantly from 2.67 to 43.33% upon acquiring PRs (P < 0.001). GPT-4o with the meta prompt demonstrated improvements in recall (43.33% vs. 52.67%, P = 0.022), accuracy (39.95% vs. 68.75%, P < 0.001), and F1 score (0.42 vs. 0.60, P < 0.001) compared to the zero-shot prompt and other prompt strategies. The localization accuracy of GPT-4o was 45.67% (137 out of 300, 95% CI: 40.00 to 51.34). A significant correlation was observed between its localization accuracy and reasoning capability under the meta prompt (φ coefficient = 0.33, p < 0.001). The model's recall increased by 5.49% (P = 0.031) by providing accurate localization cues within the meta prompt. GPT-4o demonstrated a certain degree of multimodal capability for PR, with performance enhancement through prompt engineering. Nevertheless, its performance remains inadequate for clinical requirements. Future efforts will be necessary to identify additional factors influencing the model's reasoning capability or to develop more advanced models. Evaluating GPT-4o's capability to interpret and reason through PRs and exploring potential methods to enhance its performance before clinical application in assisting radiological assessments.

Radiology report generation (RRG) tasks leverage computer-aided technology to automatically produce descriptive text reports for medical images, aiming to ease radiologists' workload, reduce misdiagnosis rates, and lessen the pressure on medical resources. However, previous works have yet to focus on enhancing feature extraction of low-quality images, incorporating cross-modal interaction information, and mitigating latency in report generation. We propose an Image-Enhanced Cross-Modal Fusion Network (IFNet) for automatic RRG to tackle these challenges. IFNet includes three key components. First, the image enhancement module enhances the detailed representation of typical and atypical structures in X-ray images, thereby boosting detection success rates. Second, the cross-modal fusion networks efficiently and comprehensively capture the interactions of cross-modal features. Finally, a more efficient transformer report generation module is designed to optimize report generation efficiency while being suitable for low-resource devices. Experimental results on public datasets IU X-ray and MIMIC-CXR demonstrate that IFNet significantly outperforms the current state-of-the-art methods.

Pulmonary disease can severely impair respiratory function and be life-threatening. Accurately recognizing pulmonary diseases in chest X-ray images is challenging due to overlapping body structures and the complex anatomy of the chest. We propose an adaptive multiscale feature fusion model for recognizing Chest X-ray images of pneumonia, tuberculosis, and COVID-19, which are common pulmonary diseases. We introduce an Adaptive Multiscale Fusion Network (AMFNet) for pulmonary disease classification in chest X-ray images. AMFNet consists of a lightweight Multiscale Fusion Network (MFNet) and ResNet50 as the secondary feature extraction network. MFNet employs Fusion Blocks with self-calibrated convolution (SCConv) and Attention Feature Fusion (AFF) to capture multiscale semantic features, and integrates a custom activation function, MFReLU, which is employed to reduce the model's memory access time. A fusion module adaptively combines features from both networks. Experimental results show that AMFNet achieves 97.48% accuracy and an F1 score of 0.9781 on public datasets, outperforming models like ResNet50, DenseNet121, ConvNeXt-Tiny, and Vision Transformer while using fewer parameters.

Generative image models have achieved remarkable progress in both natural and medical imaging. In the medical context, these techniques offer a potential solution to data scarcity-especially for low-prevalence anomalies that impair the performance of AI-driven diagnostic and segmentation tools. However, questions remain regarding the fidelity and clinical utility of synthetic images, since poor generation quality can undermine model generalizability and trust. In this study, we evaluate the effectiveness of state-of-the-art generative models-Generative Adversarial Networks (GANs) and Diffusion Models (DMs)-for synthesizing chest X-rays conditioned on four abnormalities: Atelectasis (AT), Lung Opacity (LO), Pleural Effusion (PE), and Enlarged Cardiac Silhouette (ECS). Using a benchmark composed of real images from the MIMIC-CXR dataset and synthetic images from both GANs and DMs, we conducted a reader study with three radiologists of varied experience. Participants were asked to distinguish real from synthetic images and assess the consistency between visual features and the target abnormality. Our results show that while DMs generate more visually realistic images overall, GANs can report better accuracy for specific conditions, such as absence of ECS. We further identify visual cues radiologists use to detect synthetic images, offering insights into the perceptual gaps in current models. These findings underscore the complementary strengths of GANs and DMs and point to the need for further refinement to ensure generative models can reliably augment training datasets for AI diagnostic systems.

Although the use of artificial intelligence (AI) in healthcare is increasing, stakeholder engagement remains poor, particularly relating to understanding parent/carer acceptance of AI tools in paediatric imaging. We explore these perceptions and compare them to the opinions of children and young people (CYAP). A UK national online survey was conducted, inviting parents, carers and guardians of children to participate. The survey was "live" from June 2022 to 2023. The survey included questions asking about respondents' views of AI in general, as well as in specific circumstances (e.g. fractures) with respect to children's healthcare. One hundred forty-six parents/carers (mean age = 45; range = 21-80) from all four nations of the UK responded. Most respondents (93/146, 64%) believed that AI would be more accurate at interpreting paediatric musculoskeletal radiographs than healthcare professionals, but had a strong preference for human supervision (66%). Whilst male respondents were more likely to believe that AI would be more accurate (55/72, 76%), they were twice as likely as female parents/carers to believe that AI use could result in their child's data falling into the wrong hands. Most respondents would like to be asked permission before AI is used for the interpretation of their child's scans (104/146, 71%). Notably, 79% of parents/carers prioritised accuracy over speed compared to 66% of CYAP. Parents/carers feel positively about AI for paediatric imaging but strongly discourage autonomous use. Acknowledging the diverse opinions of the patient population is vital in aiding the successful integration of AI for paediatric imaging. Parents/carers demonstrate a preference for AI use with human supervision that prioritises accuracy, transparency and institutional accountability. AI is welcomed as a supportive tool, but not as a substitute for human expertise. Parents/carers are accepting of AI use, with human supervision. Over half believe AI would replace doctors/nurses looking at bone X-rays within 5 years. Parents/carers are more likely than CYAP to trust AI's accuracy. Parents/carers are also more sceptical about AI data misuse.