To address issues such as loss of detailed information, blurred target boundaries, and unclear structural hierarchy in medical image fusion, this paper proposes an adaptive feature medical image fusion network based on a full-scale diffusion model. First, a region-level feature map is generated using a kernel-based saliency map to enhance local features and boundary details. Then, a full-scale diffusion feature extraction network is employed for global feature extraction, alongside a multi-scale denoising U-shaped network designed to fully capture cross-layer information. A multi-scale feature integration module is introduced to reinforce texture details and structural information extracted by the encoder. Finally, an adaptive fusion scheme is applied to progressively fuse region-level features, global features, and source images layer by layer, enhancing the preservation of detail information. To validate the effectiveness of the proposed method, this paper validates the proposed model on the publicly available Harvard dataset and an abdominal dataset. By comparing with nine other representative image fusion methods, the proposed approach achieved improvements across seven evaluation metrics. The results demonstrate that the proposed method effectively extracts both global and local features of medical images, enhances texture details and target boundary clarity, and generates fusion image with high contrast and rich information, providing more reliable support for subsequent clinical diagnosis.
Humans ; Image Processing, Computer-Assisted/methods* ; Algorithms ; Neural Networks, Computer ; Diagnostic Imaging/methods* ; Image Interpretation, Computer-Assisted/methods*

Accurate detection of cephalometric landmarks is crucial for orthodontic diagnosis and treatment planning. Current landmark detection methods are mainly divided into heatmap-based and regression-based approaches. However, these methods often rely on parallel computation of multiple models to improve accuracy, significantly increasing the complexity of training and deployment. This paper presented a novel regression method that can simultaneously detect all cephalometric landmarks in high-resolution X-ray images. By leveraging the encoder module of Transformer, a dual-encoder model was designed to achieve coarse-to-fine localization of cephalometric landmarks. The entire model consisted of three main components: a feature extraction module, a reference encoder module, and a fine-tuning encoder module, responsible for feature extraction and fusion of X-ray images, coarse localization of cephalometric landmarks, and fine localization of landmarks, respectively. The model was fully end-to-end differentiable and could learn the intercorrelation relationships between cephalometric landmarks. Experimental results showed that the successful detection rate (SDR) of our algorithm was superior to other existing methods. It attained the highest 2 mm SDR of 89.51% on test set 1 of the ISBI2015 dataset and 90.68% on the test set of the ISBI2023 dataset. Meanwhile, it reduces memory consumption and enhances the model's popularity and applicability, providing more reliable technical support for orthodontic diagnosis and treatment plan formulation.
Cephalometry/methods* ; Humans ; Algorithms ; Anatomic Landmarks/diagnostic imaging* ; Image Processing, Computer-Assisted/methods* ; X-Rays

Subtype classification of age-related macular degeneration (AMD) based on optical coherence tomography (OCT) images serves as an effective auxiliary tool for clinicians in diagnosing disease progression and formulating treatment plans. To improve the classification accuracy of AMD subtypes, this study proposes a keypoint-based, attention-enhanced residual network (KPA-ResNet). The proposed architecture adopts a 50-layer residual network (ResNet-50) as the backbone, preceded by a keypoint localization module based on heatmap regression to outline critical lesion regions. A two-dimensional relative self-attention mechanism is incorporated into convolutional layers to enhance the representation of key lesion areas. Furthermore, the network depth is appropriately increased and an improved residual module, ConvNeXt, is introduced to enable comprehensive extraction of high-dimensional features and enrich the detail of lesion boundary contours, ultimately achieving higher classification accuracy of AMD subtypes. Experimental results demonstrate that KPA-ResNet achieves significant improvements in overall classification accuracy compared with conventional convolutional neural networks. Specifically, for the wet AMD subtypes, the classification accuracies for inactive choroidal neovascularization (CNV) and active CNV reach 92.8% and 95.2%, respectively, representing substantial improvement over ResNet-50. These findings validate the superior performance of KPA-ResNet in AMD subtype classification tasks. This work provides a high-accuracy, generalizable network architecture for OCT-based AMD subtype classification and offers new insights into integrating attention mechanisms with convolutional neural networks in ophthalmic image analysis.
Tomography, Optical Coherence/methods* ; Humans ; Macular Degeneration/diagnostic imaging* ; Neural Networks, Computer

Colorectal polyps are important early markers of colorectal cancer, and their early detection is crucial for cancer prevention. Although existing polyp segmentation models have achieved certain results, they still face challenges such as diverse polyp morphology, blurred boundaries, and insufficient feature extraction. To address these issues, this study proposes a parallel coordinate fusion network (PCFNet), aiming to improve the accuracy and robustness of polyp segmentation. PCFNet integrates parallel convolutional modules and a coordinate attention mechanism, enabling the preservation of global feature information while precisely capturing detailed features, thereby effectively segmenting polyps with complex boundaries. Experimental results on Kvasir-SEG and CVC-ClinicDB demonstrate the outstanding performance of PCFNet across multiple metrics. Specifically, on the Kvasir-SEG dataset, PCFNet achieved an F1-score of 0.897 4 and a mean intersection over union (mIoU) of 0.835 8; on the CVC-ClinicDB dataset, it attained an F1-score of 0.939 8 and an mIoU of 0.892 3. Compared with other methods, PCFNet shows significant improvements across all performance metrics, particularly in multi-scale feature fusion and spatial information capture, demonstrating its innovativeness. The proposed method provides a more reliable AI-assisted diagnostic tool for early colorectal cancer screening.
Humans ; Colonic Polyps/diagnostic imaging* ; Colorectal Neoplasms/diagnostic imaging* ; Neural Networks, Computer ; Image Processing, Computer-Assisted/methods* ; Algorithms ; Early Detection of Cancer

Most current medical image segmentation models are primarily built upon the U-shaped network (U-Net) architecture, which has certain limitations in capturing both global contextual information and fine-grained details. To address this issue, this paper proposes a novel U-shaped network model, termed the Multi-View U-Net (MUNet), which integrates self-attention and multi-view attention mechanisms. Specifically, a newly designed multi-view attention module is introduced to aggregate semantic features from different perspectives, thereby enhancing the representation of fine details in images. Additionally, the MUNet model leverages a self-attention encoding block to extract global image features, and by fusing global and local features, it improves segmentation performance. Experimental results demonstrate that the proposed model achieves superior segmentation performance in coronary artery image segmentation tasks, significantly outperforming existing models. By incorporating self-attention and multi-view attention mechanisms, this study provides a novel and efficient modeling approach for medical image segmentation, contributing to the advancement of intelligent medical image analysis.
Humans ; Image Processing, Computer-Assisted/methods* ; Neural Networks, Computer ; Algorithms ; Attention ; Coronary Vessels/diagnostic imaging* ; Diagnostic Imaging/methods*

Endometrial cancer (EC) is one of the most common gynecological malignancies, with an increasing incidence rate worldwide. Accurate segmentation of lesion areas in computed tomography (CT) images is a critical step in assisting clinical diagnosis. In this study, we propose a novel deep learning-based segmentation model, termed spatial choice and weight union network (SCWU-Net), which incorporates two newly designed modules: the spatial selection module (SSM) and the combination weight module (CWM). The SSM enhances the model's ability to capture contextual information through deep convolutional blocks, while the CWM, based on joint attention mechanisms, is employed within the skip connections to further boost segmentation performance. By integrating the strengths of both modules into a U-shaped multi-scale architecture, the model achieves precise segmentation of EC lesion regions. Experimental results on a public dataset demonstrate that SCWU-Net achieves a Dice similarity coefficient (DSC) of 82.98%, an intersection over union (IoU) of 78.63%, a precision of 92.36%, and a recall of 84.10%. Its overall performance is significantly outperforming other state-of-the-art models. This study enhances the accuracy of lesion segmentation in EC CT images and holds potential clinical value for the auxiliary diagnosis of endometrial cancer.
Humans ; Endometrial Neoplasms/diagnostic imaging* ; Female ; Tomography, X-Ray Computed/methods* ; Deep Learning ; Algorithms ; Image Processing, Computer-Assisted/methods* ; Neural Networks, Computer

To address the challenges in blood cell recognition caused by diverse morphology, dense distribution, and the abundance of small target information, this paper proposes a blood cell detection algorithm - the "You Only Look Once" model based on hybrid mixing attention and deep over-parameters (HADO-YOLO). First, a hybrid attention mechanism is introduced into the backbone network to enhance the model's sensitivity to detailed features. Second, the standard convolution layers with downsampling in the neck network are replaced with deep over-parameterized convolutions to expand the receptive field and improve feature representation. Finally, the detection head is decoupled to enhance the model's robustness for detecting abnormal cells. Experimental results on the Blood Cell Counting Dataset (BCCD) demonstrate that the HADO-YOLO algorithm achieves a mean average precision of 90.2% and a precision of 93.8%, outperforming the baseline YOLO model. Compared with existing blood cell detection methods, the proposed algorithm achieves state-of-the-art detection performance. In conclusion, HADO-YOLO offers a more efficient and accurate solution for identifying various types of blood cells, providing valuable technical support for future clinical diagnostic applications.
Algorithms ; Humans ; Blood Cells/cytology* ; Blood Cell Count/methods* ; Neural Networks, Computer ; Deep Learning ; Detection Algorithms

Tumor treating fields (TTF) therapy is an innovative tumor treatment modality. Currently, the TTF devices predominantly employ insulated ceramic electrodes as the electric field transmission medium, resulting in low energy transfer efficiency of the electric field and poor portability of the devices. This study proposed an innovative TTF transmission mode and independently designed a conducted-electrode TTF cell culture dish utilizing inert titanium materials. The electric field conduction characteristics were verified through finite element simulations and experimental tests. Finally, based on the self-manufactured conducted-electrode TTF cell culture dish, experiments on the proliferation inhibition of U87 tumor cells by TTF were conducted. The results demonstrated that under an applied TTF voltage of 10 V and frequency of 200 kHz, the electric field intensities within the medium for conducted and insulated electrodes are approximately 2.5 V/cm and 0.7 V/cm, respectively. Compared to conventional insulated TTF systems, the conducted-electrode TTF configuration exhibited a lower electrode voltage drop and a higher electric field intensity in the culture medium, indicating superior electric field transmission efficiency. Following 36 hours of treatment with conducted-electrode TTF on U87 cells, the proliferation inhibition rate reached approximately 50%, demonstrating effective suppression of tumor cell growth. This approach presents a potential direction for optimizing TTF treatment modality and device design.
Humans ; Electrodes ; Neoplasms/pathology* ; Cell Line, Tumor ; Cell Proliferation/radiation effects* ; Electric Stimulation Therapy/methods* ; Electromagnetic Fields

Current epilepsy prediction methods are not effective in characterizing the multi-domain features of complex long-term electroencephalogram (EEG) data, leading to suboptimal prediction performance. Therefore, this paper proposes a novel multi-scale sparse adaptive convolutional network based on multi-head attention mechanism (MS-SACN-MM) model to effectively characterize the multi-domain features. The model first preprocesses the EEG data, constructs multiple convolutional layers to effectively avoid information overload, and uses a multi-layer perceptron and multi-head attention mechanism to focus the network on critical pre-seizure features. Then, it adopts a focal loss training strategy to alleviate class imbalance and enhance the model's robustness. Experimental results show that on the publicly created dataset (CHB-MIT) by MIT and Boston Children's Hospital, the MS-SACN-MM model achieves a maximum accuracy of 0.999 for seizure prediction 10 ~ 15 minutes in advance. This demonstrates good predictive performance and holds significant importance for early intervention and intelligent clinical management of epilepsy patients.
Humans ; Electroencephalography/methods* ; Epilepsy/physiopathology* ; Neural Networks, Computer ; Seizures/physiopathology* ; Signal Processing, Computer-Assisted ; Algorithms

To address the current issues of data imbalance and scarcity in photoplethysmography (PPG) data for type 2 diabetes mellitus (T2DM) prediction, this study proposes an improved conditional Wasserstein generative adversarial network with gradient penalty (CWGAN-GP). The algorithm integrated gated recurrent unit (GRU) networks and self-attention mechanisms to construct a generator, aiming to produce high-quality PPG signals. Various data augmentation methods, including the improved CWGAN-GP, were employed to expand the PPG dataset, and multiple classifiers were applied for T2DM prediction analysis. Experimental results showed that the model trained on data generated by the improved CWGAN-GP achieved the optimal prediction performance. The highest accuracy reached 0.895 0, and compared with other data enhancement methods, this approach exhibited significant advantages in terms of precision and F1-score. The generated data notably enhances the accuracy and generalization ability of T2DM prediction models, providing a more reliable technical basis for non-invasive early T2DM screening based on PPG signals.
Photoplethysmography/methods* ; Diabetes Mellitus, Type 2/diagnosis* ; Humans ; Algorithms ; Neural Networks, Computer ; Signal Processing, Computer-Assisted ; Prediction Algorithms