Background: Ambiguities and anomalies in the Activity of Daily Living (ADL) patterns indicate deviations from Wellness. The monitoring lifestyles could facilitate remote physicians or caregivers to give insight into symptoms disease provide health improvement advice residents; Objective: This research work aims apply lifestyle an ambient assisted living (AAL) system by diagnosing conduct distinguishing variation norm with slightest conceivable fake alert. In pursuing this aim, main objective...

Automatic sleep staging methods usually extract hand-crafted features or network trained from signals recorded by polysomnography (PSG), and then estimate the stages various classifiers. In this study, we propose a classification approach based on hierarchical neural to process multi-channel PSG for improving performance of automatic five-class staging. The proposed contains two stages: comprehensive feature learning stage sequence stage. first is used obtain matrix fusing features. A...

In recent years, automatic sleep staging methods have achieved competitive performance using electroencephalography (EEG) signals. However, the acquisition of EEG signals is cumbersome and inconvenient. Therefore, we propose a novel approach electrooculogram (EOG) signals, which are more convenient to acquire than EEG. A two-scale convolutional neural network first extracts epoch-wise temporary-equivalent features from raw EOG recurrent then captures long-term sequential information. The...

The electroencephalogram (EEG), for measuring the electrophysiological activity of brain, has been widely applied in automatic detection epilepsy seizures. Various EEG-based seizure algorithms have already yielded high sensitivity, but training those requires a large amount labelled data. Data labelling is often done with lot human efforts, which very time-consuming. In this study, we propose hybrid system integrating an unsupervised learning (UL) module and supervised (SL) module, where UL...

Abstract Automatic and continuous blood pressure monitoring is important for preventing cardiovascular diseases such as hypertension. The evaluation of medication effects the diagnosis clinical hypertension can both benefit from monitoring. current generation wearable monitors frequently encounters limitations with inadequate portability, electrical safety, limited accuracy, precise position alignment. Here, we present an optical fiber sensor-assisted smartwatch A adapter a liquid capsule...

Deep learning methods have become an important tool for automatic sleep staging in recent years. However, most of the existing deep learning-based approaches are sharply constrained by input modalities, where any insertion, substitution, and deletion modalities would directly lead to unusable model or a deterioration performance. To solve modality heterogeneity problems, novel network architecture named MaskSleepNet is proposed. It consists masking module, multi-scale convolutional neural...

Neuromuscular diseases or physical disabilities have the potential to impair hand dexterity, significantly affecting daily life. To date, technologies for gesture recognition based on surface electromyography (sEMG) garnered increasing attention. These aim decode motion intentions, thereby advancing assistive devices such as prosthetic hands in restoring lost function. However, limited generalization capacity across different users has hindered progress towards practical implementation. In...

Heart sounds deliver vital physiological and pathological evidence about health. Wireless cardiac auscultation offers continuous monitoring of an individual without 24*7 manual healthcare care services. In this paper, a novel wireless sensing system to monitor analyze condition is proposed, which sends the information caregiver as well medical practitioner with application Internet Things (IoT). An integrated for heart sound acquisition, storage, asynchronous analysis has been developed,...

Sleep stage classification is a fundamental but cumbersome task in sleep analysis. To score the automatically, this study presents method based on two-stage neural network. The feature learning as first can fuse network trained features with traditional hand-crafted features. A recurrent (RNN) second fully utilized for temporal information between epochs and obtaining results. solve serious sample imbalance problem, novel pre-training process combined data augmentation was introduced....

Objective: Classification of sleep-wake states using multichannel electroencephalography (EEG) data that reliably work for neonates. Methods: A deep multilayer perceptron (MLP) neural network is developed to classify bipolar EEG signals, which takes an input vector size 108 containing the joint features 9 channels. The avoids any post-processing step in order as a full-fledged real-time application. For training and testing model, recordings 3525 30-second segments from 19 neonates...

Sleep posture, as a crucial index for sleep quality assessment, has been widely studied in analysis. In this paper, an unobtrusive smart mat system based on dense flexible sensor array and printed electrodes along with algorithmic framework posture recognition is proposed. With the array, offers comfortable high-resolution solution long-term pressure sensing. Meanwhile, compared to other methods, it reduces production costs computational complexity smaller area of improves portability fewer...

Deep sleep staging networks have reached top performance on large-scale datasets. However, these models perform poorer when training and testing small cohorts due to data inefficiency. Transferring well-trained from datasets (source domain) (target is a promising solution but still remains challenging the domain-shift issue. In this work, an unsupervised domain adaptation approach, statistics alignment (DSA), developed bridge gap between distribution of source target domains. DSA adapts by...

Audio-visual speech recognition (AVSR) has gained remarkable success for ameliorating the noise-robustness of recognition. Mainstream methods focus on fusing audio and visual inputs to obtain modality-invariant representations. However, such representations are prone over-reliance modality as it is much easier recognize than video in clean conditions. As a result, AVSR model underestimates importance stream face noise corruption. To this end, we leverage modality-specific provide stable...

Most existing neonatal sleep staging appro- aches applied multiple EEG channels to obtain good performance. However, it potentially increased the computational complexity and led an risk of skin disruption neonates during data acquisition. In this paper, a multi-scale hierarchical neural network (MS-HNN) with squeeze excitation (SE) block for is presented in study on basis single channel. MS-HNN composes convolutional (MSCNN), temporal information learning (TIL) module, block. MSCNN can...

Neonatal sleep staging is crucial for understanding infant brain development and assessing neurological health. This study explores the optimal electrode configuration to reduce technical complexities potential risks of causing skin irritation neonates during data collection. A Multi-Branch Convolutional Neural Network (CNN) used categorize neonatal states based on single-channel Electroencephalography (EEG) data. The proposed model was trained tested 16803 30-second segments from 64...

The efficient-coding hypothesis asserts that neural and perceptual sensitivity evolved to faithfully represent biologically relevant sensory signals. Here we characterized the spectrotemporal modulation statistics of several natural sound ensembles examined how neurons encode these in central nucleus inferior colliculus (CNIC) cats. We report modulation-tuning CNIC is matched equalize power sounds. Specifically, sounds exhibited a tradeoff between spectral temporal modulations, which...

Astrocytes, once considered passive support cells, are increasingly appreciated as dynamic regulators of neuronal development and function, in part via secreted factors. The extent to which they similarly regulate oligodendrocytes or proliferation differentiation oligodendrocyte progenitor cells (OPCs) is less understood. Here, we generated astrocytes from human pluripotent stem (hiPSC-Astros) demonstrated that immature astrocytes, opposed mature ones, promote oligodendrogenesis vitro. In...

Abstract Objective. Automatic sleep staging models suffer from an inherent class imbalance problem (CIP), which hinders the classifiers achieving a better performance. To address this issue, we systematically studied electroencephalogram data augmentation (DA) approaches. Furthermore, modified and transferred novel DA approaches related research fields, yielding new efficient ways to enhance datasets. Approach. This study covers five methods, including repeating minority classes,...

Accurate individual exposure assessment is crucial for evaluating the health effects of particulate matter (PM). Various portable monitors built upon low-cost optical sensors have emerged. However, main challenge their application to guarantee accuracy measurements. To assess performance a newly developed PM sensor, and develop methods post-hoc data calibration optimize its quality. We conducted series laboratory experiments field evaluations quantify reproducibility within Plantower 7003...

Speech enhancement (SE) aims to suppress the additive noise from noisy speech signals improve speech's perceptual quality and intelligibility. However, over-suppression phenomenon in enhanced might degrade performance of downstream automatic recognition (ASR) task due missing latent information. To alleviate such problem, we propose an interactive feature fusion network (IFF-Net) for noise-robust learn complementary information original feature. Experimental results show that proposed method...

Surface electromyogram (sEMG) based hand gesture recognition for prosthesis or armband is an important application of the human-machine interface. However, measurement location sensors greatly influences performance, especially with inter-day inter-subject validation protocols. Therefore, we acquired two-day data 41 subjects a 256 (16×16) channel high-density sEMG electrode array. With data, initially compared support vector machine (SVM) and other four state-of-art classifiers under three...

Speech enhancement (SE) is proved effective in reducing noise from noisy speech signals for downstream automatic recognition (ASR), where multi-task learning strategy employed to jointly optimize these two tasks. However, the enhanced learned by SE objective may not always yield good ASR results. From optimization view, there sometimes exists interference between gradients of and tasks, which could hinder finally lead sub-optimal performance. In this paper, we propose a simple yet approach...

With recent advances of diffusion model, generative speech enhancement (SE) has attracted a surge research interest due to its great potential for unseen testing noises. However, existing efforts mainly focus on inherent properties clean speech, underexploiting the varying noise information in real world. In this paper, we propose noise-aware (NASE) approach that extracts noise-specific guide reverse process model. Specifically, design classification (NC) model produce acoustic embedding as...