A5026640581- Non-Invasive Vital Sign Monitoring
- Heart Rate Variability and Autonomic Control
- ECG Monitoring and Analysis
- Retinal Imaging and Analysis
- Advanced Memory and Neural Computing
- Digital Transformation in Industry
- Neural Networks and Reservoir Computing
- Retinal and Optic Conditions
- Gaze Tracking and Assistive Technology
- Advanced Sensor and Energy Harvesting Materials
- EEG and Brain-Computer Interfaces
- Phonocardiography and Auscultation Techniques
- Neuroscience and Neural Engineering
- Wireless Body Area Networks
- Neural dynamics and brain function
- Lower Extremity Biomechanics and Pathologies
- Diabetic Foot Ulcer Assessment and Management
- Neural Networks and Applications
- Optical Imaging and Spectroscopy Techniques
- Tactile and Sensory Interactions
- Climate Change and Health Impacts
- Hemodynamic Monitoring and Therapy
- Muscle activation and electromyography studies
- Ferroelectric and Negative Capacitance Devices
- Osteoarthritis Treatment and Mechanisms
Imperial College London
2021-2024
Imperial College Healthcare NHS Trust
2023
Abstract This work introduces a silent speech interface (SSI), proposing few-layer graphene (FLG) strain sensing mechanism based on thorough cracks and AI-based self-adaptation capabilities that overcome the limitations of state-of-the-art technologies by simultaneously achieving high accuracy, computational efficiency, fast decoding speed while maintaining excellent user comfort. We demonstrate its application in biocompatible textile-integrated ultrasensitive sensor embedded into smart...
Fueled by the recent proliferation of energy-efficient and energy-autonomous or self-powered nanotechnology-based wearable smart systems, human motion intention prediction (MIP) plays a critical role in wide range applications, such as rehabilitation assistive robotics, to enable more natural, biologically inspired, seamless integrated assistance task execution, including for elders physically impaired patients. With increasing complexity human-machine interactions need personalized...
Abstract Efficient operation of control systems in robotics or autonomous driving targeting real-world navigation scenarios requires perception methods that allow them to understand and adapt unstructured environments with good accuracy, adaptation, generality, similar humans. To address this need, we present a memristor-based differential neuromorphic computing, perceptual signal processing, online adaptation method providing style external sensory stimuli. The ability generality are...
Ear-worn devices offer the opportunity to measure vital signals in a 24/7 fashion, without need of clinician. These are however prone motion artefacts, so that entire epochs artefact-corrupt recordings routinely discarded. This work aims at reducing impact artefacts introduced by series common real life daily activities such as talking, chewing, and walking while recording Electroencephalogram (EEG) from ear canal. The approach used employs multiple external sensors, microphones an...
A human body digital twin (DT) is a virtual representation of an individual's physiological state, created using real-time data from sensors and medical test devices, with the purpose simulating, predicting, optimizing health outcomes through advanced analytics simulations. The DT has potential to revolutionize healthcare wellness, but its responsible effective implementation requires consideration various factors. This article presents comprehensive overview current status future prospects...
Photoplethysmography (PPG) sensors integrated in wearable devices offer the potential to monitor arterial blood pressure (ABP) patients. Such cuffless, non-invasive, and continuous solution is suitable for remote ambulatory monitoring. A machine learning model based on PPG signal can be used detect hypertension, estimate beat-by-beat ABP values, even reconstruct shape of ABP. Overall, models presented literature have shown good performance, but there a gap between research real-world use...
This work aims to classify physiological states using heart rate variability (HRV) features extracted from electrocardiograms recorded in the ears (ear-ECG). The considered this are: (a) normal breathing, (b) controlled slow and (c) mental exercises. Since both cause higher variance heartbeat intervals, breathing-related (SpO
Accurate pulse-oximeter readings are critical for clinical decisions, especially when arterial blood-gas tests - the gold standard determining oxygen saturation levels not available, such as COVID-19 severity. Several studies demonstrate that pulse estimated from photoplethysmography (PPG) introduces a racial bias due to more profound scattering of light in subjects with darker skin increased presence melanin. This leads an overestimation blood those is low and can result patient receiving...
Ambulatory heart rate (HR) monitors that acquire electrocardiogram (ECG) or/and photoplethysmographm (PPG) signals from the torso, wrists, or ears are notably less accurate in tasks associated with high levels of movement compared to clinical measurements. However, reliable estimation HR can be obtained through data fusion different sensors. These methods especially suitable for multimodal hearable devices, where tracked modalities, including electrical ECG, optical PPG, and sounds (heart...
Background: Ambulatory heart rate (HR) monitors that acquire electrocardiogram (ECG) or/and photoplethysmographm (PPG) signals from the torso, wrists, or ears are notably less accurate in tasks associated with high levels of movement compared to clinical measurements. However, a reliable estimation HR can be obtained through data fusion different sensors. These methods especially suitable for multimodal hearable devices, where tracked modalities, including electrical ECG, optical PPG, and...
Machine learning-based fundus image diagnosis technologies trigger worldwide interest owing to their benefits such as reducing medical resource power and providing objective evaluation results. However, current methods are commonly based on supervised methods, bringing in a heavy workload biomedical staff hence suffering expanding effective databases. To address this issue, article, we established label-free method, name 'SSVT',which can automatically analyze un-labeled images generate high...
Fundus diseases are major causes of visual impairment and blindness worldwide, especially in underdeveloped regions, where the shortage ophthalmologists hinders timely diagnosis. AI-assisted fundus image analysis has several advantages, such as high accuracy, reduced workload, improved accessibility, but it requires a large amount expert-annotated data to build reliable models. To address this dilemma, we propose general self-supervised machine learning framework that can handle diverse from...
The cardiac dipole has been shown to propagate the ears, now a common site for consumer wearable electronics, enabling recording of electrocardiogram (ECG) signals. However, in-ear ECG recordings often suffer from significant noise due their small amplitude and presence other physiological signals, such as electroencephalogram (EEG), which complicates extraction cardiovascular features. This study addresses this issue by developing denoising convolutional autoencoder (DCAE) enhance...
Efficient operation of intelligent machines in the real world requires methods that allow them to understand and predict uncertainties presented by unstructured environments with good accuracy, scalability generalization, similar humans. Current rely on pretrained networks instead continuously learning from dynamic signal properties working suffer inherent limitations, such as data-hungry procedures, limited generalization capabilities. Herein, we present a memristor-based differential...
Our research presents a wearable Silent Speech Interface (SSI) technology that excels in device comfort, time-energy efficiency, and speech decoding accuracy for real-world use. We developed biocompatible, durable textile choker with an embedded graphene-based strain sensor, capable of accurately detecting subtle throat movements. This surpassing other sensors sensitivity by 420%, simplifies signal processing compared to traditional voice recognition methods. system uses computationally...
<title>Abstract</title> Efficient operation of intelligent machines in the real world requires methods that allow them to understand and predict uncertainties presented by unstructured environments with good accuracy, scalability generalization, similar humans. Current rely on pretrained networks instead continuously learning from dynamic signal properties working suffer inherent limitations, such as data-hungry procedures, limited generalization capabilities. Herein, we present a...
Foot-care specialists recommend shoes by analysing the patient's gait cycle and looking for any structural or functional problems. Such methods are time consuming, inaccurate unable to identify risk factors that may lead development of foot-related diseases in future. This work presents a footwear recommendation algorithm based on genetic predispositions i.e. profile associated selected Single Nucleotide Polymorphisms (SNPs), individual activity level, addition age, body mass index (BMI)...