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.
Objective To investigate and compare the value of clinical, intratumoral, and peritumoral radiomic features in the preoperative differentiation of minimally invasive adenocarcinoma (MIA) from invasive adenocarcinoma (IAC) of the lung, and to develop a high-performance integrated predictive model to guide surgical decision-making. MethodsWe retrospectively enrolled patients with postoperative pathologically confirmed MIA or IAC at the Department of Thoracic Surgery, Northern Jiangsu People's Hospital Affiliated to Yangzhou University, from 2020 to 2022. Clinical data and preoperative CT images were collected. Intratumoral and peritumoral (a 3 mm extension from the tumor margin) regions of interest were delineated, and high-throughput radiomic features were extracted using PyRadiomics. After feature selection, various machine learning algorithms were employed to construct predictive models based on clinical features, intratumoral features, peritumoral features, and combined features. Model performance was evaluated using the area under the receiver operating characteristic curve (AUC), decision curve analysis, and calibration curves. Results A total of 665 patients were included. There were 208 males and 457 females with a mean age of (57.67±11.21) years. The radiomics model combining intratumoral and peritumoral features (AUC=0.924) outperformed the single-region models. After further integrating the optimal radiomics model with the clinical model, the AUC of the resulting fusion model increased to 0.935 in the validation set. Furthermore, Delong's test, the net reclassification improvement, and the integrated discrimination improvement all confirmed that the predictive efficacy of the fusion model was significantly superior to that of any individual model. Conclusion The machine learning model integrating clinical and radiomic features can effectively differentiate MIA from IAC preoperatively, providing reliable decision support for the selection of personalized surgical strategies.
ObjectiveTo address the inter-observer annotation variability in echocardiographic image segmentation caused by image boundary ambiguity and differences in experts’ clinical backgrounds. Methods A framework was proposed that leveraged a conditional diffusion model, guided by cardiac ultrasound images, to learn the probability distribution of multi-expert annotations. A probabilistic consensus strategy was further introduced to dynamically fuse these annotations. This approach was designed to simultaneously aggregate clinical consensus among experts and preserve the diversity of the annotation distribution, thereby jointly modeling both the stable morphological features of the target anatomy and the reasonable variations inherent in expert annotations. Given the scarcity of authentic multi-expert annotations in public datasets like CAMUS, a synthetic multi-expert annotation system was developed using morphological operations. This system was designed to simulate three typical clinical annotation style (conservative, optimal, and aggressive), to provide a reliable data foundation for validating the proposed method. ResultsThe proposed model significantly outperformed state-of-the-art methods in both left ventricular endocardium (LVEndo) and left ventricular epicardium (LVEpi) segmentation tasks. For LVEndo, the generalized energy distance (\begin{document}$ GED{_{5}}{_{0}} $\end{document}) at the end-diastolic (ED) phase reached 0.073 1, representing a 43.9% reduction compared to the D-Persona model. For LVEpi, the \begin{document}$ GED{_{5}}{_{0}} $\end{document} decreased by 42.0% and 39.0% at the ED and end-systolic phases, respectively, relative to D-Persona. Furthermore, the structural fidelity improved by 2.7-3.5 percentage points for LVEndo and 1.4-2.2 percentage points for LVEpi compared to single-expert models, indicating a superior ability to capture diverse expert preferences. Conclusion By jointly modeling population-level consensus and annotation diversity, this work enables a unified characterization of anatomical structures and their associated annotation uncertainty in cardiac ultrasound images, offering a novel approach toward robust and interpretable segmentation in clinical settings.