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.

Background: In laparoscopic surgery with CO₂ pneumoperitoneum, serious complications often occur for elderly patients and those who undergo long operations. These complications mainly include respiratory and circulatory system changes. In patients with tumors, release of free tumor cells into the abdominal cavity is believed to be possible. Gasless laparoscopic techniques can avoid these complications of CO₂ pneumoperitoneum. Currently, the main shortcoming of gasless laparoscopic techniques is inadequate operative space. Because of this shortcoming, gasless techniques have not been widely applied in clinical practice. Materials and Methods: We herein describe a new technique of gasless laparoscopic cholecystectomy in pigs using a self-designed umbrella-like abdominal wall-lifting device. This device lifts up the anterior abdominal wall by opening the umbrella leaf in the abdominal cavity. Results: Five pigs underwent laparoscopic cholecystectomy using this technique. The operation times were 85, 40, 28, 21, and 24 minutes. The corresponding bleeding volumes were 11, 20, 5, 2, and 8 mL. Conclusions: These preliminary outcomes suggest that the umbrella-like abdominal wall-lifting technique is safe and feasible in gasless laparoscopic surgery and can provide sufficient exposure of the operative field. Further study in the form of randomized controlled trials is needed to investigate the advantages of this new technique.

Aim Postoperative gastrointestinal dysfunction (POGD) is a common complication following colorectal surgery. This study aimed to investigate the incidence and risk factors of POGD after minimally invasive surgery and to assess the relationship between robotic surgery, POGD, and their outcomes. Method Patients who had undergone minimally invasive colorectal surgery at our institution between July 2018 and November 2023 were retrospectively enrolled. POGD was diagnosed based on the presence of two or more intestinal symptoms within 72 h or more after surgery. Risk factors were identified through regression analyses, and the impact of POGD on outcomes was assessed using linear regression.The association between those factors was assessed using subgroup analysis and hierarchical regression. Results A total of 226 patients were included in the analysis, including 33 with POGD. POGD occurred in 14.6% of patients, with a lower incidence in robotic surgery (7.3%) than in laparoscopic surgery (19.8%). Multivariate analysis indicated that robotic surgery had a protective effect, while blood loss exceeding 50 ml was an independent risk factor for POGD. POGD was also correlated with longer length of stays and higher costs. The association between POGD, length of stay, and cost varied depending on the surgical platform. Robotic surgery exacerbated the effect of POGD on short-term outcomes, which aligned with the observed significant interaction effect. Conclusion POGD remains a prevalent postoperative disease. Preventive strategies, including meticulous hemostasis techniques and robotic surgery, should be prioritized by healthcare professionals to reduce POGD risk, improve short-term outcomes, and preserve healthcare resources.

Textbook outcome (TO), a combined quality indicator, encompasses key postoperative indicators such as the absence of complications, R0 resection, and no prolonged length of day. It has been suggested to be of additional value over single outcome parameters in short-term outcomes of surgical treatment. The main objective of this research was to assess the relationship between TO and neoadjuvant therapy (NT), thereby providing insights into NT’s role in surgical quality. Patients who underwent minimally invasive rectal surgery were enrolled between January 2019 and June 2024. TO was defined as achieving R0 resection, at least 12 lymph nodes harvested, no adverse outcomes (Clavien–Dindo score ≥ 3, readmission, or mortality within 30 days), and length of stay within the ≤ 75th percentile for the treatment year. The relationship between TO and NT was analyzed using regression analyses. Subgroup analysis and hierarchical regression were conducted to investigate potential influencing factors and interactions. 405 patients were enrolled, with 204 achieving TO. NT was associated with a reduction in TO (OR: 0.37, 95% CI: 0.21 ~ 0.65, p < 0.001), while robotic surgery (OR: 2.88, 95% CI: 1.62 ~ 5.11), total laparoscopic surgery (OR: 2.79, 95% CI: 1.71 ~ 4.56), enhanced recovery after surgery (OR: 1.62, 95% CI: 1.02 ~ 2.59), and stoma (OR: 1.87, 95% CI: 1.18 ~ 2.96) were associated with an increased rate of TO. The impact of NT on TO varied depending on surgery duration; prolonged surgical time exacerbated the negative effect of NT on TO. This observation was consistent with a significant interaction effect. NT is associated with a lower TO rate, especially in patients with prolonged surgical time. Robotic surgery, total laparoscopic surgery, enhanced recovery after surgery, and stoma can improve achieve TO.

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.