Accurate fetal birth weight (FBW) estimation is essential for optimizing delivery decisions and reducing perinatal mortality. However, clinical methods for FBW estimation are inefficient, operator-dependent, and challenging to apply in cases of complex fetal anatomy. Existing deep learning methods are based on 2D standard ultrasound (US) images or videos that lack spatial information, limiting their prediction accuracy. In this study, we propose the first method for directly estimating FBW from 3D fetal US volumes. Our approach integrates a multi-scale feature fusion network (MFFN) and a synthetic sample-based learning framework (SSLF). The MFFN effectively extracts and fuses multi-scale features under sparse supervision by incorporating channel attention, spatial attention, and a ranking-based loss function. SSLF generates synthetic samples by simply combining fetal head and abdomen data from different fetuses, utilizing semi-supervised learning to improve prediction performance. Experimental results demonstrate that our method achieves superior performance, with a mean absolute error of $166.4\pm155.9$ $g$ and a mean absolute percentage error of $5.1\pm4.6$%, outperforming existing methods and approaching the accuracy of a senior doctor. Code is available at: https://github.com/Qioy-i/EFW.

In this paper, we present the VMSE U-Net and VM-Unet CBAM+ model, two cutting-edge deep learning architectures designed to enhance medical image segmentation. Our approach integrates Squeeze-and-Excitation (SE) and Convolutional Block Attention Module (CBAM) techniques into the traditional VM U-Net framework, significantly improving segmentation accuracy, feature localization, and computational efficiency. Both models show superior performance compared to the baseline VM-Unet across multiple datasets. Notably, VMSEUnet achieves the highest accuracy, IoU, precision, and recall while maintaining low loss values. It also exhibits exceptional computational efficiency with faster inference times and lower memory usage on both GPU and CPU. Overall, the study suggests that the enhanced architecture VMSE-Unet is a valuable tool for medical image analysis. These findings highlight its potential for real-world clinical applications, emphasizing the importance of further research to optimize accuracy, robustness, and computational efficiency.

In recent years, accelerated MRI reconstruction based on deep learning has led to significant improvements in image quality with impressive results for high acceleration factors. However, from a clinical perspective image quality is only secondary; much more important is that all clinically relevant information is preserved in the reconstruction from heavily undersampled data. In this paper, we show that existing techniques, even when considering resampling for diffusion-based reconstruction, can fail to reconstruct small and rare pathologies, thus leading to potentially wrong diagnosis decisions (false negatives). To uncover the potentially missing clinical information we propose ``Semantically Diverse Reconstructions'' (\SDR), a method which, given an original reconstruction, generates novel reconstructions with enhanced semantic variability while all of them are fully consistent with the measured data. To evaluate \SDR automatically we train an object detector on the fastMRI+ dataset. We show that \SDR significantly reduces the chance of false-negative diagnoses (higher recall) and improves mean average precision compared to the original reconstructions. The code is available on https://github.com/NikolasMorshuis/SDR

Deep learning-based diagnostic models often suffer performance drops due to distribution shifts between training (source) and test (target) domains. Collecting and labeling sufficient target domain data for model retraining represents an optimal solution, yet is limited by time and scarce resources. Active learning (AL) offers an efficient approach to reduce annotation costs while maintaining performance, but struggles to handle the challenge posed by distribution variations across different datasets. In this study, we propose a novel unsupervised Active learning framework for Domain Adaptation, named ADAptation, which efficiently selects informative samples from multi-domain data pools under limited annotation budget. As a fundamental step, our method first utilizes the distribution homogenization capabilities of diffusion models to bridge cross-dataset gaps by translating target images into source-domain style. We then introduce two key innovations: (a) a hypersphere-constrained contrastive learning network for compact feature clustering, and (b) a dual-scoring mechanism that quantifies and balances sample uncertainty and representativeness. Extensive experiments on four breast ultrasound datasets (three public and one in-house/multi-center) across five common deep classifiers demonstrate that our method surpasses existing strong AL-based competitors, validating its effectiveness and generalization for clinical domain adaptation. The code is available at the anonymized link: https://github.com/miccai25-966/ADAptation.

Deformable medical image registration is traditionally formulated as an optimization problem. While classical methods solve this problem iteratively, recent learning-based approaches use recurrent neural networks (RNNs) to mimic this process by unrolling the prediction of deformation fields in a fixed number of steps. However, classical methods typically converge after sufficient iterations, but learning-based unrolling methods lack a theoretical convergence guarantee and show instability empirically. In addition, unrolling methods have a practical bottleneck at training time: GPU memory usage grows linearly with the unrolling steps due to backpropagation through time (BPTT). To address both theoretical and practical challenges, we propose DEQReg, a novel registration framework based on Deep Equilibrium Models (DEQ), which formulates registration as an equilibrium-seeking problem, establishing a natural connection between classical optimization and learning-based unrolling methods. DEQReg maintains constant memory usage, enabling theoretically unlimited iteration steps. Through extensive evaluation on the public brain MRI and lung CT datasets, we show that DEQReg can achieve competitive registration performance, while substantially reducing memory consumption compared to state-of-the-art unrolling methods. We also reveal an intriguing phenomenon: the performance of existing unrolling methods first increases slightly then degrades irreversibly when the inference steps go beyond the training configuration. In contrast, DEQReg achieves stable convergence with its inbuilt equilibrium-seeking mechanism, bridging the gap between classical optimization-based and modern learning-based registration methods.

Objectives Parametric tissue mapping enables quantitative cardiac tissue characterization but is limited by inter-observer variability during manual delineation. Traditional approaches relying on average relaxation values and single cutoffs may oversimplify myocardial complexity. This study evaluates whether deep learning (DL) can achieve segmentation accuracy comparable to inter-observer variability, explores the utility of statistical features beyond mean T1/T2 values, and assesses whether machine learning (ML) combining multiple features enhances disease detection. Materials & Methods T1 and T2 maps were manually segmented. The test subset was independently annotated by two observers, and inter-observer variability was assessed. A DL model was trained to segment left ventricle blood pool and myocardium. Average (A), lower quartile (LQ), median (M), and upper quartile (UQ) were computed for the myocardial pixels and employed in classification by applying cutoffs or in ML. Dice similarity coefficient (DICE) and mean absolute percentage error evaluated segmentation performance. Bland-Altman plots assessed inter-user and model-observer agreement. Receiver operating characteristic analysis determined optimal cutoffs. Pearson correlation compared features from model and manual segmentations. F1-score, precision, and recall evaluated classification performance. Wilcoxon test assessed differences between classification methods, with p < 0.05 considered statistically significant. Results 144 subjects were split into training (100), validation (15) and evaluation (29) subsets. Segmentation model achieved a DICE of 85.4%, surpassing inter-observer agreement. Random forest applied to all features increased F1-score (92.7%, p < 0.001). Conclusion DL facilitates segmentation of T1/ T2 maps. Combining multiple features with ML improves disease detection.

Oral cancer is a hazardous disease and a major cause of morbidity and mortality worldwide. The purpose of this study was to develop the deep convolutional neural networks (CNN)-based multiclass classification and object detection models for distinguishing and detection of oral carcinoma and sarcoma in contrast-enhanced CT images. This study included 3,259 slices of CT images of oral cancer cases from the cancer hospital and two regional hospitals from 2016 to 2020. Multiclass classification models were constructed using DenseNet-169, ResNet-50, EfficientNet-B0, ConvNeXt-Base, and ViT-Base-Patch16-224 to accurately differentiate between oral carcinoma and sarcoma. Additionally, multiclass object detection models, including Faster R-CNN, YOLOv8, and YOLOv11, were designed to autonomously identify and localize lesions by placing bounding boxes on CT images. Performance evaluation on a test dataset showed that the best classification model achieved an accuracy of 0.97, while the best detection models yielded a mean average precision (mAP) of 0.87. In conclusion, the CNN-based multiclass models have a great promise for accurately determining and distinguishing oral carcinoma and sarcoma in CT imaging, potentially enhancing early detection and informing treatment strategies.

To construct and validate an interpretable machine learning (ML) radiomics model derived from multiparametric magnetic resonance imaging (MRI) images to differentiate between luminal and non-luminal breast cancer (BC) subtypes. This study enrolled 1098 BC participants from four medical centers, categorized into a training cohort (n = 580) and validation cohorts 1-3 (n = 252, 89, and 177, respectively). Multiparametric MRI-based radiomics features, including T1-weighted imaging (T1WI), T2-weighted imaging (T2WI), diffusion-weighted imaging (DWI), apparent diffusion coefficient (ADC), and dynamic contrast-enhanced (DCE) imaging, were extracted. Five ML algorithms were applied to develop various radiomics models, from which the best performing model was identified. A ML-based combined model including optimal radiomics features and clinical predictors was constructed, with performance assessed through receiver operating characteristic (ROC) analysis. The Shapley additive explanation (SHAP) method was utilized to assess model interpretability. Tumor size and MR-reported lymph node status were chosen as significant clinical variables. Thirteen radiomics features were identified from multiparametric MRI images. The extreme gradient boosting (XGBoost) radiomics model performed the best, achieving area under the curves (AUCs) of 0.941, 0.903, 0.862, and 0.894 across training and validation cohorts 1-3, respectively. The XGBoost combined model showed favorable discriminative power, with AUCs of 0.956, 0.912, 0.894, and 0.906 in training and validation cohorts 1-3, respectively. The SHAP visualization facilitated global interpretation, identifying "ADC_wavelet-HLH_glszm_ZoneEntropy" and "DCE_wavelet-HLL_gldm_DependenceVariance" as the most significant features for the model's predictions. The XGBoost combined model derived from multiparametric MRI may proficiently differentiate between luminal and non-luminal BC and aid in treatment decision-making. An interpretable machine learning radiomics model can preoperatively predict luminal and non-luminal subtypes in breast cancer, thereby aiding therapeutic decision-making.

The stomach is one of the main digestive organs in the GIT, essential for digestion and nutrient absorption. However, various gastrointestinal diseases, including gastritis, ulcers, and cancer, affect health and quality of life severely. The precise diagnosis of gastrointestinal (GI) tract diseases is a significant challenge in the field of healthcare, as misclassification leads to late prescriptions and negative consequences for patients. Even with the advancement in machine learning and explainable AI for medical image analysis, existing methods tend to have high false negative rates which compromise critical disease cases. This paper presents a hybrid deep learning based explainable artificial intelligence (XAI) approach to improve the accuracy of gastrointestinal disorder diagnosis, including stomach diseases, from images acquired endoscopically. Swin Transformer with DCNN (EfficientNet-B3, ResNet-50) is integrated to improve both the accuracy of diagnostics and the interpretability of the model to extract robust features. Stacked machine learning classifiers with meta-loss and XAI techniques (Grad-CAM) are combined to minimize false negatives, which helps in early and accurate medical diagnoses in GI tract disease evaluation. The proposed model successfully achieved an accuracy of 93.79% with a lower misclassification rate, which is effective for gastrointestinal tract disease classification. Class-wise performance metrics, such as precision, recall, and F1-score, show considerable improvements with false-negative rates being reduced. AI-driven GI tract disease diagnosis becomes more accessible for medical professionals through Grad-CAM because it provides visual explanations about model predictions. This study makes the prospect of using a synergistic DL with XAI open for improvement towards early diagnosis with fewer human errors and also guiding doctors handling gastrointestinal diseases.

This study aimed to identify the hyoid bone (HB) using the nnU-Net based artificial intelligence (AI) model in cone beam computed tomography (CBCT) images and assess the model's success in automatic segmentation. CBCT images of 190 patients were randomly selected. The raw data was converted to DICOM format and transferred to the 3D Slicer Imaging Software (Version 4.10.2; MIT, Cambridge, MA, USA). HB was labeled manually using the 3D Slicer. The dataset was divided into training, validation, and test sets in a ratio of 8:1:1. The nnU-Net v2 architecture was utilized to process the training and test datasets, generating the algorithm weight factors. To assess the model's accuracy and performance, a confusion matrix was employed. F1-score, Dice coefficient (DC), 95% Hausdorff distance (95% HD), and Intersection over Union (IoU) metrics were calculated to evaluate the results. The model's performance metrics were as follows: DC = 0.9434, IoU = 0.8941, F1-score = 0.9446, and 95% HD = 1.9998. The receiver operating characteristic (ROC) curve was generated, yielding an AUC value of 0.98. The results indicated that the nnU-Net v2 model achieved high precision and accuracy in HB segmentation on CBCT images. Automatic segmentation of HB can enhance clinicians' decision-making speed and accuracy in diagnosing and treating various clinical conditions. Not applicable.