Accurate MRI image segmentation is crucial for disease diagnosis, but current Transformer-based methods face two key challenges: limited capability to capture detailed information, leading to blurred boundaries and false localization, and the lack of MRI-specific embedding paradigms for attention modules, which limits their potential and representation capability. To address these challenges, this paper proposes a multi-scheme cross-level attention embedded U-shape Transformer (MSCL-SwinUNet). This model integrates cross-level spatial-wise attention (SW-Attention) to transfer detailed information from encoder to decoder, cross-stage channel-wise attention (CW-Attention) to filter out redundant features and enhance task-related channels, and multi-stage scale-wise attention (ScaleW-Attention) to adaptively process multi-scale features. Extensive experiments on the ACDC, MM-WHS and Synapse datasets demonstrate that the proposed MSCL-SwinUNet surpasses state-of-the-art methods in accuracy and generalizability. Visualization further confirms the superiority of our model in preserving detailed boundaries. This work not only advances Transformer-based segmentation in medical imaging but also provides new insights into designing MRI-specific attention embedding paradigms.Our code is available at https://github.com/waylans/MSCL-SwinUNet .

Accurate segmentation of white matter hyperintensities (WMH) is crucial for clinical decision-making, particularly in the context of multiple sclerosis. However, domain shifts, such as variations in MRI machine types or acquisition parameters, pose significant challenges to model calibration and uncertainty estimation. This comparative study investigates the impact of domain shift on WMH segmentation, proposing maximum-entropy regularization techniques to enhance model calibration and uncertainty estimation. The purpose is to identify errors appearing after model deployment in clinical scenarios using predictive uncertainty as a proxy measure, since it does not require ground-truth labels to be computed. We conducted experiments using a classic U-Net architecture and evaluated maximum entropy regularization schemes to improve model calibration under domain shift on two publicly available datasets: the WMH Segmentation Challenge and the 3D-MR-MS dataset. Performance is assessed with Dice coefficient, Hausdorff distance, expected calibration error, and entropy-based uncertainty estimates. Entropy-based uncertainty estimates can anticipate segmentation errors, both in-distribution and out-of-distribution, with maximum-entropy regularization further strengthening the correlation between uncertainty and segmentation performance, while also improving model calibration under domain shift. Maximum-entropy regularization improves uncertainty estimation for WMH segmentation under domain shift. By strengthening the relationship between predictive uncertainty and segmentation errors, these methods allow models to better flag unreliable predictions without requiring ground-truth annotations. Additionally, maximum-entropy regularization contributes to better model calibration, supporting more reliable and safer deployment of deep learning models in multi-center and heterogeneous clinical environments.

We present the first comparative study of two fundamentally distinct feature extraction techniques: Histogram of Oriented Gradients (HOG) and Topological Data Analysis (TDA), for medical image classification using retinal fundus images. HOG captures local texture and edge patterns through gradient orientation histograms, while TDA, using cubical persistent homology, extracts high-level topological signatures that reflect the global structure of pixel intensities. We evaluate both methods on the large APTOS dataset for two classification tasks: binary detection (normal versus diabetic retinopathy) and five-class diabetic retinopathy severity grading. From each image, we extract 26244 HOG features and 800 TDA features, using them independently to train seven classical machine learning models with 10-fold cross-validation. XGBoost achieved the best performance in both cases: 94.29 percent accuracy (HOG) and 94.18 percent (TDA) on the binary task; 74.41 percent (HOG) and 74.69 percent (TDA) on the multi-class task. Our results show that both methods offer competitive performance but encode different structural aspects of the images. This is the first work to benchmark gradient-based and topological features on retinal imagery. The techniques are interpretable, applicable to other medical imaging domains, and suitable for integration into deep learning pipelines.

Early diagnosis of Neurological Disorders (ND) such as Alzheimer's disease (AD) and Brain Tumors (BT) can be highly challenging since these diseases cause minor changes in the brain's anatomy. Magnetic Resonance Imaging (MRI) is a vital tool for diagnosing and visualizing these ND; however, standard techniques contingent upon human analysis can be inaccurate, require a long-time, and detect early-stage symptoms necessary for effective treatment. Spatial Feature Extraction (FE) has been improved by Convolutional Neural Networks (CNN) and hybrid models, both of which are changes in Deep Learning (DL). However, these analysis methods frequently fail to accept temporal dynamics, which is significant for a complete test. The present investigation introduces the STGCN-ViT, a hybrid model that integrates CNN + Spatial-Temporal Graph Convolutional Networks (STGCN) + Vision Transformer (ViT) components to address these gaps. The model causes the reference to EfficientNet-B0 for FE in space, STGCN for FE in time, and ViT for FE using AM. By applying the Open Access Series of Imaging Studies (OASIS) and Harvard Medical School (HMS) benchmark datasets, the recommended approach proved effective in the investigations, with Group A attaining an accuracy of 93.56%, a precision of 94.41% and an Area under the Receiver Operating Characteristic Curve (AUC-ROC) score of 94.63%. Compared with standard and transformer-based models, the model attains better results for Group B, with an accuracy of 94.52%, precision of 95.03%, and AUC-ROC score of 95.24%. Those results support the model's use in real-time medical applications by providing proof of the probability of accurate but early-stage ND diagnosis.

The generation of medical images presents significant challenges due to their high-resolution and three-dimensional nature. Existing methods often yield suboptimal performance in generating high-quality 3D medical images, and there is currently no universal generative framework for medical imaging. In this paper, we introduce a 3D Medical Latent Diffusion (3D MedDiffusion) model for controllable, high-quality 3D medical image generation. 3D MedDiffusion incorporates a novel, highly efficient Patch-Volume Autoencoder that compresses medical images into latent space through patch-wise encoding and recovers back into image space through volume-wise decoding. Additionally, we design a new noise estimator to capture both local details and global structural information during diffusion denoising process. 3D MedDiffusion can generate fine-detailed, high-resolution images (up to 512x512x512) and effectively adapt to various downstream tasks as it is trained on large-scale datasets covering CT and MRI modalities and different anatomical regions (from head to leg). Experimental results demonstrate that 3D MedDiffusion surpasses state-of-the-art methods in generative quality and exhibits strong generalizability across tasks such as sparse-view CT reconstruction, fast MRI reconstruction, and data augmentation for segmentationand classification. Source code and checkpoints are available at https://github.com/ShanghaiTech-IMPACT/3D-MedDiffusion.

Orthodontically-induced external root resorption (OIERR) is among the most common risks in orthodontic treatment. Traditional OIERR diagnosis is limited by subjective judgement as well as cumbersome manual measurement. The research aims to develop an intelligent detection model for OIERR based on deep convolutional neural networks (CNNs) through cone-beam computed tomography (CBCT) images, thus providing auxiliary diagnosis support for orthodontists. Six pretrained CNN architectures were adopted and 1717 CBCT slices were used for training to construct OIERR detection models. The performance of the models was tested on 429 CBCT slices and the activated regions during decision-making were visualized through heatmaps. The model performance was then compared with that of two orthodontists. The EfficientNet-B1 model, trained through hold-out cross-validation, proved to be the most effective for detecting OIERR. Its accuracy, precision, sensitivity, specificity as well as F1-score were 0.97, 0.98, 0.97, 0.98 and 0.98, respectively. The metrics remarkably outperformed those of orthodontists, whose accuracy, recall and F1-score were 0.86, 0.78, and 0.87 respectively (P < 0.01). The heatmaps suggested that the OIERR detection model primarily relied on root features for decision-making. Automatic detection of OIERR through CNNs as well as CBCT images is both accurate and efficient. The method outperforms orthodontists and is anticipated to serve as a clinical tool for the rapid screening and diagnosis of OIERR.

We propose a compositional graph-based Machine Learning (ML) framework for Alzheimer's disease (AD) detection that constructs complex ML predictors from modular components. In our directed computational graph, datasets are represented as nodes [Formula: see text], and deep learning (DL) models are represented as directed edges [Formula: see text], allowing us to model complex image-processing pipelines [Formula: see text] as end-to-end DL predictors. Each directed path in the graph functions as a DL predictor, supporting both forward propagation for transforming data representations, as well as backpropagation for model finetuning, saliency map computation, and input data optimization. We demonstrate our model on Alzheimer's disease prediction, a complex problem that requires integrating multimodal data containing scans of different modalities and contrasts, genetic data and cognitive tests. We built a graph of 11 nodes (data) and 14 edges (ML models), where each model has been trained on handling a specific task (e.g. skull-stripping MRI scans, AD detection,image2image translation, ...). By using a modular and adaptive approach, our framework effectively integrates diverse data types, handles distribution shifts, and scales to arbitrary complexity, offering a practical tool that remains accurate even when modalities are missing for advancing Alzheimer's disease diagnosis and potentially other complex medical prediction tasks.

Artificial intelligence (AI) has introduced numerous opportunities for human assistance and task automation in medicine. However, it suffers from poor generalization in the presence of shifts in the data distribution. In the context of AI-based computed tomography (CT) analysis, significant data distribution shifts can be caused by changes in scanner manufacturer, reconstruction technique or dose. AI harmonization techniques can address this problem by reducing distribution shifts caused by various acquisition settings. This paper presents an open-source benchmark dataset containing CT scans of an anthropomorphic phantom acquired with various scanners and settings, which purpose is to foster the development of AI harmonization techniques. Using a phantom allows fixing variations attributed to inter- and intra-patient variations. The dataset includes 1378 image series acquired with 13 scanners from 4 manufacturers across 8 institutions using a harmonized protocol as well as several acquisition doses. Additionally, we present a methodology, baseline results and open-source code to assess image- and feature-level stability and liver tissue classification, promoting the development of AI harmonization strategies.

Medical image encryption is important for maintaining the confidentiality of sensitive medical data and protecting patient privacy. Contemporary healthcare systems store significant patient data in text and graphic form. This research proposes a New 5D hyperchaotic system combined with a customised U-Net architecture. Chaotic maps have become an increasingly popular method for encryption because of their remarkable characteristics, including statistical randomness and sensitivity to initial conditions. The significant region is segmented from the medical images using the U-Net network, and its statistics are utilised as initial conditions to generate the new random sequence. Initially, zig-zag scrambling confuses the pixel position of a medical image and applies further permutation with a new 5D hyperchaotic sequence. Two stages of diffusion are used, such as dynamic DNA flip and dynamic DNA XOR, to enhance the encryption algorithm's security against various attacks. The randomness of the New 5D hyperchaotic system is verified using the NIST SP800-22 statistical test, calculating the Lyapunov exponent and plotting the attractor diagram of the chaotic sequence. The algorithm validates with statistical measures such as PSNR, MSE, NPCR, UACI, entropy, and Chi-square values. Evaluation is performed for test images yields average horizontal, vertical, and diagonal correlation coefficients of -0.0018, -0.0002, and 0.0007, respectively, Shannon entropy of 7.9971, Kolmogorov Entropy value of 2.9469, NPCR of 99.61%, UACI of 33.49%, Chi-square "PASS" at both the 5% (293.2478) and 1% (310.4574) significance levels, key space is 2⁵⁰⁰ and an average encryption time of approximately 2.93 s per 256 × 256 image on a standard desktop CPU. The performance comparisons use various encryption methods and demonstrate that the proposed method ensures secure reliability against various challenges.

Alzheimer's disease (AD) is a progressive neurodegenerative disorder that significantly impacts cognitive function, posing a major global health challenge. Despite its rising prevalence, particularly in low and middle-income countries, early diagnosis remains inadequate, with projections estimating over 55 million affected individuals by 2022, expected to triple by 2050. Accurate early detection is critical for effective intervention. This study presents Neuroimaging-based Early Detection of Alzheimer's Disease using Deep Learning (NEDA-DL), a novel computer-aided diagnostic (CAD) framework leveraging a hybrid ResNet-50 and AlexNet architecture optimized with CUDA-based parallel processing. The proposed deep learning model processes MRI and PET neuroimaging data, utilizing depthwise separable convolutions to enhance computational efficiency. Performance evaluation using key metrics including accuracy, sensitivity, specificity, and F1-score demonstrates state-of-the-art classification performance, with the Softmax classifier achieving 99.87% accuracy. Comparative analyses further validate the superiority of NEDA-DL over existing methods. By integrating structural and functional neuroimaging insights, this approach enhances diagnostic precision and supports clinical decision-making in Alzheimer's disease detection.