Early diagnosis of skin cancer can significantly improve patient survival. Currently, skin lesion classification faces challenges such as lesion–background semantic entanglement, high intra-class variability, artifactual interference, and more, while existing classification models lack modeling of physicians’ diagnostic paradigms. To this end, we propose DermViT, a medically driven deep learning architecture that addresses the above issues through a medically-inspired modular design. DermViT consists of three main modules: (1) Dermoscopic Context Pyramid (DCP), which mimics the multi-scale observation process of pathological diagnosis to adapt to the high intraclass variability of lesions such as melanoma, then extract stable and consistent data at different scales; (2) Dermoscopic Hierarchical Attention (DHA), which can reduce computational complexity while realizing intelligent focusing on lesion areas through a coarse screening–fine inspection mechanism; (3). Dermoscopic Feature Gate (DFG), which simulates the observation–verification operation of doctors through a convolutional gating mechanism and effectively suppresses semantic leakage of artifact regions. Our experimental results show that DermViT significantly outperforms existing methods in terms of classification accuracy (86.12%, a 7.8% improvement over ViT-Base) and number of parameters (40% less than ViT-Base) on the ISIC2018 and ISIC2019 datasets. Our visualization results further validate DermViT’s ability to locate lesions under interference conditions. By introducing a modular design that mimics a physician’s observation mode, DermViT achieves more logical feature extraction and decision-making processes for medical diagnosis, providing an efficient and reliable solution for dermoscopic image analysis.

Medical image registration is essential in clinical practices such as surgical navigation and image-guided diagnosis. The Transformer architecture of TransMorph demonstrates better accuracy in non-rigid registration tasks. However, its weaker spatial locality priors necessitate large-scale training datasets and a heavy number of parameters, which conflict with the limited annotated data and real-time demands of clinical workflows. Moreover, traditional downsampling and upsampling always degrade high-frequency anatomical features such as tissue boundaries or small lesions. We proposed WaveMorph, a wavelet-guided multi-scale ConvNeXt method for unsupervised medical image registration. A novel multi-scale wavelet feature fusion downsampling module is proposed by integrating the ConvNeXt architecture with Haar wavelet lossless decomposition to extract and fuse features from eight frequency sub-images using multi-scale convolution kernels. Additionally, a lightweight dynamic upsampling module is introduced in the decoder to reconstruct fine-grained anatomical structures. WaveMorph integrates the inductive bias of CNNs with the advantages of Transformers, effectively mitigating topological distortions caused by spatial information loss while supporting real-time inference. In both atlas-to-patient (IXI) and inter-patient (OASIS) registration tasks, WaveMorph demonstrates state-of-the-art performance, achieving Dice scores of 0.779 ± 0.015 and 0.824 ± 0.021, respectively, and real-time inference (0.072 s/image), validating the effectiveness of our model in medical image registration.

Accurate prediction of Electrocardiogram (ECG) signals is crucial for early diagnosis and continuous cardiac monitoring. To address the challenges posed by the non-stationary nature of ECG signals, this study introduces a novel deep learning prediction method that synergistically combines Variational Mode Decomposition (VMD), the Cao method, and a Long Short-Term Memory (LSTM) neural network. The proposed method first applies VMD to adaptively decompose the raw ECG signal into several Intrinsic Mode Functions (IMFs), which effectively reduces nonlinearity and non-stationarity. Subsequently, the Cao method is utilized to compute the minimum embedding dimension for each IMF, thereby optimally configuring the input structure for the LSTM network. Each IMF component is then predicted independently by an LSTM model trained with the Adam optimizer. The final reconstructed ECG signal is derived from the aggregate of these individual IMF predictions. Evaluated on the benchmark MIT-BIH Arrhythmia Database, the proposed method achieves a root mean square error (RMSE) of 0.001326 and a mean absolute error (MAE) of 0.001044, demonstrating high predictive precision. The results from a comparative study indicate that the proposed method surpasses several established prediction methods, confirming its effectiveness and potential for practical application in enhancing ECG signal analysis.