Data augmentation is an effective technique for regularizing deep networks, which helps to enhance the generalizability and robustness of the model. However, in the field of medical imaging, traditional data augmentation techniques such as cropping, rotation, and degradation may inadvertently alter the critical characteristics of pathological lesions. Conventional semantic augmentation methods, such as altering the color and contrast of the object background, may also affect the structural features of medical images in uncontrolled semantic directions. Such operational conditions compromise the model's diagnostic reliability in medical contexts. To address this issue, we propose a surprisingly efficient implicit augmentation-invariant learning strategy (AILS) via variational Bayesian inference on differentially constrained feature manifolds. Parameterizing probability measures over tangent space through deep networks enables precise estimation of semantic direction distributions. Subsequently, geodesic-aware semantic features are sampled from the reparameterized variational posterior, achieving semantic-consistent augmentation. Simultaneously, to mine augmentation distribution invariance, we design the AiHLoss, which constrains the augmentation distribution to facilitate the network to learn augmentation invariance. Extensive experiments demonstrate that AILS exhibits high performance on public medical image datasets, outperforming existing augmentation methods.

As individuals age, simultaneous declines in cognitive function, physical ability, and visual capacity (e.g., dynamic visual acuity) pose challenges to maintaining functional independence in later life. Despite advances in immersive virtual environments for age-related functional support, dynamic visual acuity remains underexplored, particularly in relation to its interaction with cognitive and motor processes. Meanwhile, existing interventions often rely solely on performance-based difficulty adjustment, neglecting the dual need to maintain engagement and promote skill progression. To address these gaps, we developed the Pareto-based Dynamic Difficulty Adjustment for Cross-domain Co-training in Virtual Reality (CCVR-PDDA) system, integrating psychology paradigm (e.g., Stroop task), upper-limb motor training (e.g., arm lifting and raising), and dynamic visual exercises (e.g., multi-directional eye movements). A 12-week longitudinal study involving 60 older adults (≥65 years) demonstrated significant cognitive improvements, sustained at six-month follow-up. Qualitative feedback further supported the system’s usability, motivation enhancement, and long-term engagement potential. These findings underscore the efficacy and acceptability of CCVR-PDDA in promoting cognitive health in aging populations.

Facial video-based physiological measurement primarily estimates heart rate by analyzing subtle skin color variations or global pixel motion patterns, which is a critical research direction of remote objective signal evaluation in human-computer interaction. However, accurate physiological signal extraction from facial videos remains challenging due to motion artifacts and large-scale motion interference. To address these issues, inspired by physiological synchronization phenomena, we propose a novel dual-domain perception enhanced fusion network for video-based physiological measurement. The model employs a dual-branch architecture that jointly explores the correlation and periodic consistency between intrinsic physiological processes and pixel-level facial variations through synchronized spectral skin reflectance analysis and extemporization motion modeling. The network extracts multi-semantic physiological features from facial video sequences, and introduces a spectral frequency attention module to refine discriminative frequency components in the spectral domain. Additionally, the local shift convolution module is designed to capture fine-grained global features in sparse temporal data via learnable shift operations, enhancing sensitivity to periodic physiological fluctuations. Furthermore, we propose a frequency complementary fusion mechanism, which adaptively retain and fuse high and low frequency components as-sociated with physiological signals. This fusion strategy strength-ens the relationship between latent physiological processes and multi-dimensional frequency-domain representations, improving measurement robustness and accuracy. Extensive experiments on multiple benchmark datasets demonstrate that our method achieves state-of-the-art performance in physiological signal es-timation.

The Efficient Vision-Language Modulation Network (EVLM) aims to achieve accurate polyp segmentation, which is crucial for the early detection and prevention of colorectal cancer. However, the implementation of EVLM remains challenging due to the morphological differences, blurred boundaries, and complex visual background in colonoscopy images. Existing purely visual methods often fail to capture subtle contextual cues and lack the semantic guidance of clinical text, resulting in poor generalization performance across datasets. To overcome these limitations, we propose a novel EVLM that integrates BiomedCLIP text embeddings with SAM2-UNet via progressive, layer-by-layer feature linear modulation (FiLM). An efficient upconvolutional block (EUCB) is introduced to improve upsampling fidelity and boundary reconstruction. Comprehensive evaluation on three benchmark datasets (CVC-ClinicDB, Kvasir-SEG, and CVC-300) demonstrates that EVLM achieves superior performance, achieving mDice scores of 94%, 91.8%, and 89.5 %, respectively, consistently outperforming state-of-the-art methods. These findings highlight the potential of EVLM in real-time clinical deployment to provide robust and accurate polyp segmentation, thereby supporting enhanced colorectal cancer screening and early intervention.

Deploying deep segmentation models in clinical settings is hindered by their excessive computational demands, which exceed the capabilities of mobile and edge devices widely used in low-resource healthcare. We propose RM-SSNet, an ultra-lightweight medical image segmentation framework that achieves 2,612 parameters-a 99.97 % reduction compared to standard UNet-while maintaining competitive accuracy. First, a recursive multi-scale pyramid architecture reuses identical transformation functions across different spatial resolutions, achieving multiresolution modeling capabilities while maintaining efficient parameter sharing. Second, a dynamic state space fusion module captures long-range spatial dependencies with linear computational complexity, replacing the quadratic complexity of traditional attention mechanisms. Third, a boundary-aware knowledge-guided refinement mechanism leverages parameterfree classical edge detection combined with learnable anatomical constraints to enhance segmentation precision without additional computational overhead. This integrated design enables clinically viable accuracy across multiple medical imaging datasets, including skin lesions, breast ultrasound, and retinal images. The framework supports real-time segmentation on mobile devices, making advanced AI-assisted diagnostic screening accessible in low-resource clinical settings.

Early diagnosis of mild cognitive impairment (MCI) is pivotal in mitigating the risk of cognitive impairment and the onset of dementia. However, prevailing clinical cognitive screening tests and biomarker assessment approaches often suffer from drawbacks such as high cost, invasiveness, time consumption, or subjectivity. In China, existing digital cognitive assessment tools face multiple challenges due to their distinctive cultural, language, and healthcare landscape. Therefore, we embarked on a study to develop a digital cognitive assessment tool, evaluate its efficacy in distinguishing between healthy individuals and MCI patients, and examine its acceptability among Chinese older adults. Through a series of symposiums, programming, and interviews with stakeholders, we iteratively designed the "BrainNursing" mobile application. The system consists of eleven single tasks and three dual tasks, each taking only 1–3 minutes. Subsequently, we conducted statistical comparisons of movement kinetics and physiological signals recorded during cognitive testing in 181 older adults to investigate which parameters could serve as effective digital biomarkers for MCI screening. Leveraging machine learning classifiers and a majority voting principle, we evaluated the classification performance of the BrainNursing system in detecting MCI, yielding an accuracy rate of 90.3%. Furthermore, our analysis of movement kinetics revealed that time, score, and sequence features are crucial in cognitive assessment. Contrasting with their healthy counterparts, MCI patients exhibited decreased heart rate variability, increased sympathetic nervous system activity, and weakened autonomic nervous system regulation in response to stimuli during cognitive testing. Finally, user experience feedback indicated that participants universally perceived BrainNursing as a convenient and user-friendly cognitive screening tool and provided valuable insights for further improvement.

Deep neural network models generally enhance super-resolution (SR) reconstruction of remote sensing images but may distort feature edge details. Recovering low-resolution (LR) remote sensing images with clear texture and high-fidelity edge details is challenging. Recent approaches improve feature fusion in the frequency domain, yielding promising results. We propose a dual-domain optimization network (DDOM) based on discrete Fourier transform (DFT) and frequency-domain complex-valued neural networks. Unlike end-to-end approaches in the image domain, DDOM incorporates frequency-domain information via DFT transformation operators, preserving high-level semantics (phase) and low-level statistical information (magnitude). Using lightweight complex neural networks and Swin Transformer architecture, the frequency-domain and image-domain subnetworks are designed. The dual-domain data consistency constraints ensure the positivity of model optimization. Extensive experiments show superior performance over existing methods in quantitative and qualitative evaluations. The proposed scheme’s robustness is verified on additional datasets. Code and model configurations are available athttp://github.com/YB-Cheng/DDOM.

Despite the similar global structures in Chest X-ray (CXR) images, the same anatomy exhibits varying appearances across images, including differences in local textures, shapes, colors, etc. Learning consistent representations for anatomical semantics through these diverse appearances poses a great challenge for self-supervised pre-training in CXR images. To address this challenge, we propose two new pre-training tasks: inner-image anatomy localization (IIAL) and cross-image anatomy localization (CIAL). Leveraging the relatively stable positions of identical anatomy across images, we utilize position information directly as supervision to learn consistent semantic representations. Specifically, IIAL adopts a coarse-to-fine heatmap localization approach to correlate anatomical semantics with positions, while CIAL leverages feature affine alignment and heatmap localization to establish a correspondence between identical anatomical semantics across varying images, despite their appearance diversity. Furthermore, we introduce a unified end-to-end pre-training framework, anatomy-aware representation learning (AARL), integrating IIAL, CIAL, and a pixel restoration task. The advantages of AARL are: 1) preserving the appearance diversity and 2) training in a simple end-to-end way avoiding complicated preprocessing. Extensive experiments on six downstream tasks, including classification and segmentation tasks in various application scenarios, demonstrate that our AARL: 1) has more powerful representation and transferring ability; 2) is annotation-efficient, reducing the demand for labeled data and 3) improves the sensitivity to detecting various pathological and anatomical patterns.

Given the prevalence of cognitive impairment in older adults and its frequent misdiagnosis or delayed diagnosis, non-pharmacological interventions have been proposed as solutions, which include cognitive training, prevention or risk reduction of dementia using virtual reality (VR) technology. This study aimed to investigate the effects of a virtual reality cognitive-motor training intervention (VRCMTI) on improving cognitive and physical function in older adults with cognitive impairment. We co-designed the VRCMTI system with multiple stakeholders by organizing symposiums and conducting pilot evaluations. The VRCMTI consisted of three virtual cognitive tasks and three upper limb movement tasks focused on improving working memory, spatial cognition, attention shifting, executive control, joint flexibility, and coordination. One hour of brain cognition and upper limb motor training was performed each week during the 12-week intervention. Sixty older adults were included in the study and randomly assigned to either the VR group or the control group. Participants in the control group received usual care and would not undergo any intervention training. Task independence, accuracy and time were measured during each session. The results showed that the VR group significantly improved global cognitive ability scores compared to the control group, especially in attention and verbal cognition. In addition, older adults in the VR group also showed significant improvements in upper limb motor skills, driven primarily by movement quality and processing speed. The findings suggest that VRCMTI could improve cognitive function and enhance motor performance. When implemented with routine health care for older adults, this practical and effective intervention may be an appropriate complementary strategy for maintaining cognitive health and preventing motor deterioration.

Mild cognitive impairment is associated with many neurodegenerative diseases. It is essential to detect mild cognitive impairment on time to reduce the prevalence of such disorders. Nevertheless, present clinically employed test scales and biomarkers are time-consuming, user-unfriendly, and expensive. Hence, we developed a neuropsychological test system based on virtual reality in this study, the Virtual Reality Neuropsychological Mild Cognitive Impairment Test (VRNPT). The diagnosis and classification of MCI were achieved by effectively combining digital cognitive parameters and EEG signal features obtained during the VRNPT cognitive task. The VRNPT contains three head-mounted display-based cognitive tasks that assess participants’ attention, memory, spatial perception, working memory, and visuospatial executive ability across multiple cognitive domains of functioning. We investigated how to design and optimize these tasks. We conducted a field study by recruiting 80 participants (40 MCI patients and 40 normal older adults). The results showed that the classification accuracy of combining digitized cognitive parameters and EEG signals during VRNPT was 91.3%, higher than using only digitized parameters from VRNPT and applying EEG signals alone, demonstrating the validity and feasibility of this method for diagnosing MCI. The user satisfaction survey showed that the subjects were satisfied with VRNPT.

Abstract Safety is an important issue in the mining industry and the Internet of Things (IoT) plays an important role to enhance the safety of the underground working environment. The IoT is used to transfer data generated by underground sensors to cloud storage for further processing. However, third‐party platforms are often a target for cyber attacks. Serious mining accidents might occur if the data were tampered with. In the overground scenario, the security of trading data is also an important issue. The Mine Consortium Blockchain (MCB) is proposed to solve the above problems. The MCB avoids the risk of centralized storage and enables data security, provenance and transparency by taking advantage of blockchain technology. The MCB platform ensures that only designated participants can process mineral data. Any violation is immutably recorded in the MCB and is easily traced back by other participants. Classical consensus mechanisms as the core technology of the blockchain cannot be directly and appropriately applied to the mining industry. A Block Alliance Consensus (BAC) mechanism, which is suitable for all consortium blockchain scenarios, is proposed to improve the performance of the MCB. In addition, the block structure of the underground sensor data is optimized: the blocks only contain a hash of the sensor data and the data being stored in the cloud. The efficiency of the BAC is demonstrated by simulation experiments where the performance of the BAC consensus mechanism is compared with with the performance of classical consensus mechanisms. The MCB and the BAC consensus mechanism were also implemented on Hyperledger Fabric. Finally the Hyperledger Caliper evaluation tool was used to evaluate the performance of the system.

Summary The diagnosis of the disease requires a variety of data and indicators. In order to make the computer perform intelligent computation and diagnosis from a doctor's perspective, complementary information between different modal data needs to be taken into account. Meanwhile, the redundancy features with key information should be selected in order to reduce the complexity of calculation. In this study, an adaptive dynamic loss function is proposed to weight different scales in the multi‐scale expansion network of pathological images according to the doctor's diagnosis process. And an ant colony algorithm based on maximum information coefficient correlation was designed for unsupervised feature selection of fusion features combined image feature and patient differential genes. Experimental results show that the addition of pathological image information and genetic information plays an important role in the classification of lung cancer subtypes. Compared with other feature selection methods, the proposed algorithm can quickly converge. Combining pathological image and gene expression matrix for cancer diagnosis can improve the diagnostic accuracy of specific patients, with an accuracy of 95.62 and AUC achieves 0.897. The proposed method has high effectiveness and superior performance in the classification of lung cancer.

Abstract The diagnosis of pulmonary nodules by clinicians depends not only on radiological imaging but also on the patient's own clinical record information and other factors. However, exploring the guiding role of clinical information is a major challenge. In this article, an intelligent personalized diagnosis decision‐making model is proposed, which combines radiology images with patient information. First, the 3D image cube of the pulmonary nodule is constructed. Then, a 3D multi‐level fusion ResNet is designed to extract the features of the nodule by making full use of the spatial context information. Finally, a kind of classification model based on feature‐related analysis was proposed, which fused clinical information features and image features and realized a nonlinear radial basis feature mapping. We tested this method on the public dataset and a cooperation hospital dataset. Experiments show that this method can effectively improve the classification accuracy of unstable nodules at the classification boundary. Our model showed significant improvements in sensitivity, specificity, and accuracy. Meanwhile, compared with other deep learning diagnosis methods, our method achieves better discriminative results and is highly suited to be used for pulmonary nodule diagnosis.

The prediction of lung tumour growth is the key to early treatment of lung cancer. However, the lack of intuitive and clear judgments about the future development of the tumour often leads patients to miss the best treatment opportunities. Combining the characteristics of the variational autoencoder and recurrent neural networks, this study proposes a tumour growth prediction via a conditional recurrent variational autoencoder. The proposed model uses a variational autoencoder to reconstruct tumour images at different times. Meanwhile, the recurrent units are proposed to infer the relationship between tumour images according to the chronological order. The different tumour development varies in different patients, patients' condition is adopted to achieve personalised prediction. To solve the problem of blurred results, the authors add the total variation regularisation term into the object function. The proposed method was tested on longitudinal studies, National Lung Screening Trial and cooperative hospital dataset, with three points on lung tumours. The precision, recall, and dice similarity coefficient reach 82.22, 79.89 and 82.49%, respectively. Both quantitative and qualitative experimental results show that the proposed method can produce realistic tumour images.

BACKGROUND: Lung cancer is one of the most common types of cancer, among which lung adenocarcinoma accounts for the largest proportion. Currently, accurate staging is a prerequisite for effective diagnosis and treatment of lung adenocarcinoma. Previous research has used mainly single-modal data, such as gene expression data, for classification and prediction. Integrating multi-modal genetic data (gene expression RNA-seq, methylation data and copy number variation) from the same patient provides the possibility of using multi-modal genetic data for cancer prediction. A new machine learning method called gcForest has recently been proposed. This method has been proven to be suitable for classification in some fields. However, the model may face challenges when applied to small samples and high-dimensional genetic data. RESULTS: In this paper, we propose a multi-weighted gcForest algorithm (MLW-gcForest) to construct a lung adenocarcinoma staging model using multi-modal genetic data. The new algorithm is based on the standard gcForest algorithm. First, different weights are assigned to different random forests according to the classification performance of these forests in the standard gcForest model. Second, because the feature vectors generated under different scanning granularities have a diverse influence on the final classification result, the feature vectors are given weights according to the proposed sorting optimization algorithm. Then, we train three MLW-gcForest models based on three single-modal datasets (gene expression RNA-seq, methylation data, and copy number variation) and then perform decision fusion to stage lung adenocarcinoma. Experimental results suggest that the MLW-gcForest model is superior to the standard gcForest model in constructing a staging model of lung adenocarcinoma and is better than the traditional classification methods. The accuracy, precision, recall, and AUC reached 0.908, 0.896, 0.882, and 0.96, respectively. CONCLUSIONS: The MLW-gcForest model has great potential in lung adenocarcinoma staging, which is helpful for the diagnosis and personalized treatment of lung adenocarcinoma. The results suggest that the MLW-gcForest algorithm is effective on multi-modal genetic data, which consist of small samples and are high dimensional.

To solve the energy limitation problems of wireless sensor networks used in forest fire detection applications, a new algorithm based on a fire probability threshold switch is proposed. The method uses the weighted average method to obtain the weights of each sensor node, and then calculates the fire probability for each node based on a logistic regression to obtain the fire probability threshold, based on this, the numbers of sensor nodes sending data to the sink node are obtained. The proposed algorithm is applied to detection of a wood fire in a simulated situation. The results show that the proposed method reduced the transmission energy used by 34%. These results indicate that the method can reduce the unnecessary energy usage effectively while guaranteeing the reliability of the transmission data.

Wildfires are highly destructive disasters that spread quickly. The use of advanced technology to achieve early warnings of wildfires is essential for the protection of wilderness resources. Nowadays, the method of using wireless sensor networks for wildfire warning has been extensively studied by many researchers. In this paper, we propose and have implemented a multi-sensor data fusion algorithm for wildfire monitoring and warning based on adaptive weighted fusion algorithm (AWFA) and Dempster -- Shafer theory (DST) of evidence. At the same time, we also have put forward some auxiliary algorithms for fire warning, including heterogeneous sensor data homogenization methods, a judgment algorithm for sensor numerical errors, and an evidence conflict solution of Dempster -- Shafer theory of evidence. Experimental results show that this algorithm can ensure the timeliness and accuracy of the wildfire warning, effectively reduce the amount of data transmission of sensor nodes and the whole network, and reduce the energy consumption, thus prolonging the network lifetime.