CoChat
Sign Up
Cited by 535Advanced Sensor and Energy Harvesting MaterialsNon-Invasive Vital Sign MonitoringNanomaterials and Printing Technologies

The emergence of AI-based wearable sensors for digital health technology: a review

Open access
Shajari, Shaghayegh, Kuruvinashetti, Kirankumar, Komeili, Amin, Sundararaj, Uttandaraman·2023·Sensors
DOI 10.3390/s23239498PMC PMC10708748
Open this paper in CoChat — free accountOpen PDF

Abstract

Disease diagnosis and monitoring using conventional healthcare services is typically expensive and has limited accuracy. Wearable health technology based on flexible electronics has gained tremendous attention in recent years for monitoring patient health owing to attractive features, such as lower medical costs, quick access to patient health data, ability to operate and transmit data in harsh environments, storage at room temperature, non-invasive implementation, mass scaling, etc. This technology provides an opportunity for disease pre-diagnosis and immediate therapy. Wearable sensors have opened a new area of personalized health monitoring by accurately measuring physical states and biochemical signals. Despite the progress to date in the development of wearable sensors, there are still several limitations in the accuracy of the data collected, precise disease diagnosis, and early treatment. This necessitates advances in applied materials and structures and using artificial intelligence (AI)-enabled wearable sensors to extract target signals for accurate clinical decision-making and efficient medical care. In this paper, we review two significant aspects of smart wearable sensors. First, we offer an overview of the most recent progress in improving wearable sensor performance for physical, chemical, and biosensors, focusing on materials, structural configurations, and transduction mechanisms. Next, we review the use of AI technology in combination with wearable technology for big data processing, self-learning, power-efficiency, real-time data acquisition and processing, and personalized health for an intelligent sensing platform. Finally, we present the challenges and future opportunities associated with smart wearable sensors.

AI snapshot

Written from the paper’s full text.

TL;DR

This review surveys AI-based wearable sensors for digital health, covering materials and transduction for physical, chemical, and biosensing plus AI for data processing, diagnosis, and therapy. It concludes that AI can improve accuracy and enable personalized/preventive care, but accuracy, validation, power, standardization, and privacy remain barriers.

Key findings

  • AI improves wearable signal processing and classification: H-LSTM reached 99.15% accuracy in human activity recognition; a multi-kernel CNN classified piezo-yarn motion signals at 99.6%; ANFIS gait-phase accuracy was 93% for toe-off, 94% for mid-swing, 96% for heel-off, and ≈100% for heel-strike.
  • In biosensing and diagnosis, DL-assisted colorimetric sweat analysis matched laboratory measurements at 91.0–99.7%; a combined LSTM/isolation-forest algorithm reached 97.3% accuracy for diabetes and 96.2% for heart disease; prostate cancer urinary multimarker screening with random forest and neural networks exceeded 99% accuracy using surface-enhanced Raman spectroscopy signals from 76 urine specimens.
  • Sleep staging with the SLEEPNET deep recurrent neural network, trained on 10,000 PSG recordings, reached 85.76% overall accuracy (N1: 56%, N2: 88%, N3: 85%, REM: 92%, Wake: 85%).
  • The review identifies persistent challenges in wearable sensors: sensitivity, accuracy, validation against gold standards, battery life, data security, user adoption, standardization, and the need to relate biofluid signals to health parameters.

Methods

  • Narrative review of wearable sensors—physical/physiological, chemical/biosensors, optical, and electrophysiological—and AI integration for signal processing, disease prediction, treatment feedback, and personalized monitoring.
  • Evidence comes from cited literature and case studies; no primary experiments, systematic search strategy, inclusion/exclusion criteria, or meta-analysis are reported.
  • Analysis is thematic, synthesizing materials, structures, transduction mechanisms, AI applications, and challenges/opportunities.

Limitations

  • Stated: wearable sensors still have limitations in sensitivity, accuracy, and technology validation; second-generation wearables must improve accuracy, battery life, data security, user adoption, and standardization; AI/DL require large standardized datasets; optical sensors face ambient light and motion artifacts; biofluid signal–health parameter relationships remain understudied; repeatability, reproducibility, and dynamic stability remain issues.
  • Stated: privacy and security risks arise from sensitive health data and wireless connectivity, and continuous AI processing can strain battery life.
  • Evident: as a narrative review, it does not report a systematic search, quality appraisal, or quantitative synthesis, so its conclusions are not based on a reproducible evidence synthesis.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.3390/s23239498

Cites

252 of 252 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

0 of 208 cited sources checked · 0 supported · 0 flagged

  • Interpretationcitation

    One of the most significant advantages of AI-based wearable health technology is to promote preventive healthcare.

    “One of the most significant advantages of AI-based wearable health technology is to promote preventive healthcare.”
    Quote in the paperNot a substantial claim0 of 4 sources checked
    • cites [12] Harrer, 2019, Artificial Intelligence for Clinical Trial Design, Trends Pharmacol. Sci. source

      Not checked in the free run.

    • cites [13] Jin, X., Liu, C., Xu, T., Su, L., and Zhang, X. (2020). Artificial intelligence biosensors: Challenges and prospects. Biosens. Bioelectron., 165. source

      Not checked in the free run.

    • cites [14] Khan, 2020, Applications of Artificial Intelligence and Big Data Analytics in m-Health: A Healthcare System Perspective, J. Healthc. Eng. source

      Not checked in the free run.

    • cites [15] King, 2017, Application of data fusion techniques and technologies for wearable health monitoring, Med. Eng. Phys. source

      Not checked in the free run.

  • Interpretationcitation

    A deep neural network (DNN) can be employed for multiplex detection of single particles and molecular biomarkers.

    “Moreover, it was shown that a deep neural network (DNN) for the multiplex detection of single particles and molecular biomarkers can be employed [24].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [24] Ganjalizadeh, 2023, Machine learning at the edge for AI-enabled multiplexed pathogen detection, Sci. Rep. source

      Not checked in the free run.

  • Interpretationreasoning

    AI pattern recognition models can be employed for multimodal sensing and cross-sensitivity issues to identify specific patterns associated with each signal and isolate individual signal contributions.

    “Moreover, when it comes to multimodal sensing and cross-sensitivity issues where the measurement of one signal is influenced by the presence of other signals, AI pattern recognition models can be employed. These models are trained to identify specific patterns associated with each signal, enabling the isolation of individual signal contributions even in the presence of cross-sensitivity.”
    Quote in the paperNot a substantial claim
  • Interpretationcitation

    AI filtering techniques can address the cross-sensitivity of multiple biomarker sensing, contributing to more accurate and reliable measurements in healthcare applications.

    “Additionally, AI filtering techniques can address the cross-sensitivity of multiple biomarker sensing, contributing to more accurate and reliable measurements in healthcare applications [23].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [23] Kline, 2022, Multimodal machine learning in precision health: A scoping review, npj Digit. Med. source

      Not checked in the free run.

  • Premisecitation

    An ML-based multimodal electrochemical analytical device utilizing eMoSx-Laser Induced Graphene was used for multiplexed detection of tyrosine and uric acid in sweat and saliva.

    “For instance, an ML-based multimodal electrochemical analytical device, utilizing eMoSx-Laser Induced Graphene, was used for the multiplexed detection of tyrosine and uric acid in sweat and saliva [21].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [21] Kammarchedu, 2022, A machine learning-based multimodal electrochemical analytical device based on eMoSx-LIG for multiplexed detection of tyrosine and uric acid in sweat and saliva, Anal. Chim. Acta source

      Not checked in the free run.

  • May combine two claims. 65
  • Classified as probably not a substantial claim. 38

  • May combine two claims.

    Interpretationcitation

    Data collected by wearable sensors can be analyzed using ML and AI algorithms to provide insights into an individual's health status, enabling early detection of health issues and personalized healthcare.

    “The data collected by wearable sensors can be analyzed using machine learning (ML) and AI algorithms to provide insights into an individual’s health status, enabling early detection of health issues and the provision of personalized healthcare [6,11].”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [6] Yu, 2018, Artificial intelligence in healthcare, Nat. Biomed. Eng. source

      Not checked in the free run.

    • cites [11] Briganti, 2020, Artificial Intelligence in Medicine: Today and Tomorrow, Front. Med. source

      Not checked in the free run.

  • May combine two claims.

    Interpretationnone given

    Despite progress in wearable sensors, limitations remain in the accuracy of collected data, precise disease diagnosis, and early treatment.

    “Despite the progress to date in the development of wearable sensors, there are still several limitations in the accuracy of the data collected, precise disease diagnosis, and early treatment.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationreasoning

    AI-based wearable sensors can collect, analyze, and transmit real-time health data to healthcare providers so they can make efficient decisions based on patient data.

    “This has led to the emergence of artificial intelligence (AI)-based wearable sensors that can collect, analyze, and transmit real-time health data to healthcare providers so that they can make efficient decisions based on patient data.”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites [1] Ha, 2018, Wearable and flexible sensors for user-interactive health-monitoring devices, J. Mater. Chem. B source

      Not checked in the free run.

    • cites [2] Homayounfar, 2020, Wearable Sensors for Monitoring Human Motion: A Review on Mechanisms, Materials, and Challenges, SLAS Technol. source

      Not checked in the free run.

    • cites [3] Huifeng, 2020, Continuous health monitoring of sportsperson using IoT devices based wearable technology, Comput. Commun. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S2 (checkable). May combine two claims.

    Interpretationreasoning

    Advances in applied materials and structures and using AI-enabled wearable sensors are necessary to extract target signals for accurate clinical decision-making and efficient medical care.

    “This necessitates advances in applied materials and structures and using artificial intelligence (AI)-enabled wearable sensors to extract target signals for accurate clinical decision-making and efficient medical care.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    This approach was validated through multi-spot optical excitation of Klebsiella Pneumoniae bacterial nucleic acids flowing through an optofluidic waveguide chip, generating fluorescence signals with varying amplitude, duration, and quality.

    “This approach was validated through the multi-spot optical excitation of Klebsiella Pneumoniae bacterial nucleic acids flowing through an optofluidic waveguide chip, generating fluorescence signals with varying amplitude, duration, and quality.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [24] Ganjalizadeh, 2023, Machine learning at the edge for AI-enabled multiplexed pathogen detection, Sci. Rep. source

      Not checked in the free run.

  • May combine two claims.

    Interpretationcitation

    Wearable technology enables remote monitoring of patients and personalized care.

    “Wearable technology enables remote monitoring of patients and personalized care.”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites [4] Baig, 2017, A Systematic Review of Wearable Patient Monitoring Systems–Current Challenges and Opportunities for Clinical Adoption, J. Med. Syst. source

      Not checked in the free run.

    • cites [14] Khan, 2020, Applications of Artificial Intelligence and Big Data Analytics in m-Health: A Healthcare System Perspective, J. Healthc. Eng. source

      Not checked in the free run.

    • cites [16] Haick, 2021, Artificial Intelligence in Medical Sensors for Clinical Decisions, ACS Nano source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Electrodeposition of MoSx resulted in increased electrochemically active surface area and heterogeneous electron transfer rate constant.

    “The electrodeposition of MoSx resulted in an increased electrochemically active surface area and heterogeneous electron transfer rate constant.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [21] Kammarchedu, 2022, A machine learning-based multimodal electrochemical analytical device based on eMoSx-LIG for multiplexed detection of tyrosine and uric acid in sweat and saliva, Anal. Chim. Acta source

      Not checked in the free run.

  • May combine two claims.

    Interpretationcitation

    AI and ML-based time series analysis can perform comprehensive, time-synchronized data collection from the body sensor network framework and integrate real-time location systems.

    “AI and ML-based time series analysis can perform comprehensive, and time-synchronized data collection from the body sensor network framework and integrate real-time location systems [27].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [27] Phatak, 2021, Artificial Intelligence Based Body Sensor Network Framework-Narrative Review: Proposing an End-to-End Framework using Wearable Sensors, Real-Time Location Systems and Artificial Intelligence/Machine Learning Algorithms for Data Collection, Data Mining and Knowledge Discovery in Sports and Healthcare, Sports Med. Open source

      Not checked in the free run.

  • May combine two claims.

    Interpretationcitation

    AI algorithms can analyze massive amounts of data collected by wearable devices, enabling healthcare providers to identify patterns, predict health outcomes, and make knowledgeable decisions about patient care.

    “AI algorithms can be used to analyze the massive amount of data collected by wearable devices, enabling healthcare providers to identify patterns, predict health outcomes, and make knowledgeable decisions about patient care.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationcitation

    As biosensors are still in development stage for monitoring biochemical signals, the technology is yet to sense biological biomarkers continuously and needs higher sensitivity, selectivity, and accuracy towards specific biomolecule sensing.

    “As the biosensors are still in the development stage in monitoring biochemical signals, the technology is yet to sense biological biomarkers continuously, and they need to achieve higher sensitivity, selectivity, and accuracy towards specific biomolecule sensing [14,17].”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [14] Khan, 2020, Applications of Artificial Intelligence and Big Data Analytics in m-Health: A Healthcare System Perspective, J. Healthc. Eng. source

      Not checked in the free run.

    • cites [17] Vashistha, R., Dangi, A.K., Kumar, A., Chhabra, D., and Shukla, P. (2018). Futuristic biosensors for cardiac health care: An artificial intelligence approach. 3 Biotech, 8. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    A wireless body area sensor network (WBSN) is designed based on one or more sinks to enable biosensors to first transmit data to these sinks for storage, processing, and transmission to another network.

    “In this context, a wireless body area sensor network (WBSN) is designed based on one or more sinks to enable biosensors to first transmit the data to these sinks for storage, processing, and transmission to another network [25].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [25] Nardin, 2015, Towards the fast and robust optimal design of wireless body area networks, Appl. Soft Comput. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    The AI model could integrate rapid wavelet particle detection with short-time Fourier Transform analysis, followed by DNN identification on a specialized AI device.

    “The AI model could integrate rapid wavelet particle detection with short-time Fourier Transform analysis, followed by DNN identification on a specialized AI device.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [24] Ganjalizadeh, 2023, Machine learning at the edge for AI-enabled multiplexed pathogen detection, Sci. Rep. source

      Not checked in the free run.

  • May combine two claims.

    Interpretationreasoning

    There is a risk of inaccurate readings from wearable sensors, which can lead to false alarms or missed opportunities for early detection of health issues.

    “There is still a risk of inaccurate readings, which can lead to false alarms or missed opportunities for early detection of health issues.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    A repulsion-based improved grey wolf optimizer for sensor node localization outperforms the traditional Global Positioning System.

    “Another technique is using a repulsion-based improved grey wolf optimizer for the sensor nodes localization that outperforms the traditional Global Positioning System [26].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [26] Abuaddous, 2022, Repulsion-based grey wolf optimizer with improved exploration and exploitation capabilities to localize sensor nodes in 3D wireless sensor network, Soft Comput. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Wearable tear biosensors have been used for continuous glucose monitoring since 2007 and later for lactate sensing since 2012.

    “Wearable tear biosensors have been used for continuous glucose monitoring since 2007 [147,148,149], and later for lactate sensing since 2012 [150].”
    Quote in the paperNot a substantial claim0 of 4 sources checked
    • cites [147] Iguchi, 2007, A flexible and wearable biosensor for tear glucose measurement, Biomed. Microdevices source

      Not checked in the free run.

    • cites [148] Senior, 2014, Novartis signs up for Google smart lens, Nat. Biotechnol. source

      Not checked in the free run.

    • cites [149] Yao, 2012, A contact lens with integrated telecommunication circuit and sensors for wireless and continuous tear glucose monitoring, J. Micromech. Microeng. source

      Not checked in the free run.

    • cites [150] Thomas, 2012, A contact lens with an integrated lactate sensor, Sens. Actuators B Chem. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    The piezoresistive sensitivity in polymer nanocomposite-based sensors is much larger than their piezo-capacitive sensitivities, which are typically less than 1.

    “The piezoresistive sensitivity in polymer nanocomposite-based sensors is much larger than their piezo-capacitive sensitivities, which are typically less than 1 [38,109].”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [38] Amjadi, 2015, Stretchable, Skin-Mountable, and Wearable Strain Sensors and Their Potential Applications: A Review, Adv. Funct. Mater. source

      Not checked in the free run.

    • cites [109] Hu, 2013, Elastomeric transparent capacitive sensors based on an interpenetrating composite of silver nanowires and polyurethane, Appl. Phys. Lett. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Electrochemical detection is highly sensitive and can detect biomarkers in the femtomolar ranges.

    “Electrochemical detection is highly sensitive and can detect biomarkers in the femtomolar ranges [125].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [125] Steckl, 2018, Stress Biomarkers in Biological Fluids and Their Point-of-Use Detection, ACS Sens. source

      Not checked in the free run.

  • May combine two claims.

    Interpretationcitation

    Despite some efforts, several challenges remain for tear biosensors, such as difficulty in sampling tears, small sample volumes, issues with fast evaporation, dependence of tear composition on emotional or mechanical stimulation, and reliable operation of fully integrated wireless sensing platforms.

    “Despite some efforts, several challenges remain, such as difficulty in sampling tears, small sample volumes, issues with fast evaporation, dependence of tear composition on emotional or mechanical stimulation, and reliable operation of fully integrated wireless sensing platforms [126,152].”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [126] Kim, 2019, Wearable biosensors for healthcare monitoring, Nat. Biotechnol. source

      Not checked in the free run.

    • cites [152] Sempionatto, 2019, Eyeglasses-based tear biosensing system: Non-invasive detection of alcohol, vitamins and glucose, Biosens. Bioelectron. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationcitation

    Energy-efficient routing algorithms leverage reinforcement learning to adaptively optimize network pathways based on real-time energy constraints.

    “energy-efficient routing algorithms leverage reinforcement learning to adaptively optimize network pathways based on real-time energy constraints [28].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [28] Zhang, R., and Yu, J. (2020). Energy-Efficient Algorithms and Protocols for Wireless Body Sensor Networks, Springer International Publishing. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Tears were used for monitoring human eye intraocular pressure as an indicator for glaucoma using soft contact lenses for wireless detection of both glucose and intraocular pressure.

    “In addition, tears were used for monitoring human eye intraocular pressure as an indicator for glaucoma, which can cause blindness, using soft contact lenses for the wireless detection of both glucose and intraocular pressure [151].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [151] Kim, 2017, Wearable smart sensor systems integrated on soft contact lenses for wireless ocular diagnostics, Nat. Commun. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Recent progress in flexible devices focusing on electronic skin applications shows a requirement for stretchable and sensitive strain sensors that can tolerate over 30% strain.

    “Recent progress in flexible devices focusing on electronic skin (E-skin) applications shows a requirement for stretchable and sensitive strain sensors that can tolerate over 30% strain [39].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [39] Park, 2015, Material approaches to stretchable strain sensors, Chemphyschem source

      Not checked in the free run.

  • May combine two claims.

    Interpretationcitation

    With the help of wireless systems and AI, wearable smart sensors can be personalized and accessible to patients anywhere and anytime.

    “With the help of wireless systems and AI, these wearable smart sensors can be personalized and be accessible to patients anywhere and anytime [31,32] (Figure 1b).”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [31] Munos, 2016, Mobile health: The power of wearables, sensors, and apps to transform clinical trials, Ann. N. Y. Acad. Sci. source

      Not checked in the free run.

    • cites [32] Sheridan, 2014, Apple moves on health, drug developers shift into smart gear, Nat. Biotechnol. source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Javey group, Gao group and Wang group focused on electrochemical biosensing of sweat using various wearable sweat patches for sweat rate, L-Dopa, C-reactive protein, cortisol, ketone, glucose, and therapeutic drugs.

    “Javey group, Gao group and Wang group also focused on electrochemical biosensing of sweat using various wearable sweat patches for sweat rate [30], L-Dopa [138,139], and C-reactive protein [140], cortisol [141,142], ketone [143], glucose [144], and therapeutic drugs [145].”
    Quote in the paperNot a substantial claim0 of 4 sources checked
    • cites [30] Gao, 2016, Fully integrated wearable sensor arrays for multiplexed in situ perspiration analysis, Nature source

      Not checked in the free run.

    • cites [138] Tai, 2019, Wearable Sweat Band for Noninvasive Levodopa Monitoring, Nano Lett. source

      Not checked in the free run.

    • cites [139] Moon, 2021, Non-Invasive Sweat-Based Tracking of L-Dopa Pharmacokinetic Profiles Following an Oral Tablet Administration, Angew Chem. Int. Ed. Engl. source

      Not checked in the free run.

    • cites [140] Tu, 2023, A wireless patch for the monitoring of C-reactive protein in sweat, Nat. Biomed. Eng. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Conventional metal-based strain sensors showed GF of less than 5, whereas semiconductor-based strain sensors possess much larger GF, up to 200 times more than metals.

    “Conventional metal-based strain sensors showed GF of less than 5, whereas semiconductor-based strain sensors possess much larger GF, up to 200 times more than metals [87].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [87] Qiu, 2019, A Path Beyond Metal and Silicon:Polymer/Nanomaterial Composites for Stretchable Strain Sensors, Adv. Funct. Mater. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    These bio-transferrable sensors exhibited high sensitivity and selectivity and promised pathogenic threats.

    “These bio-transferrable sensors exhibited high sensitivity and selectivity and promised pathogenic threats.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [157] Mannoor, 2012, Graphene-based wireless bacteria detection on tooth enamel, Nat. Commun. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    A wireless mouthguard biosensor used by Kim et al. detects uric acid to prevent diseases such as hyperuricemia, gout, and Lesch-Nyhan syndrome.

    “Wearable saliva biosensors can have different configurations, such as a wireless mouthguard biosensor used by Kim et al. [156] to detect uric acid to prevent diseases such as hyperuricemia, gout, and Lesch–Nyhan syndrome.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [156] Kim, 2015, Wearable salivary uric acid mouthguard biosensor with integrated wireless electronics, Biosens. Bioelectron. source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Saliva is one of the ideal biofluids for detecting body nutrition upon taking foods or supplements, since it shows the quickest response, e.g., ketone bodies or vitamin C and vitamin D can be detected in saliva within 30 min.

    “Saliva is one of the ideal biofluids for detecting body nutrition upon taking foods or supplements, since it shows the quickest response, e.g., ketone bodies or vitamin C and vitamin D [155] can be detected in saliva within 30 min.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [155] Saha, 2023, Non-invasive monitoring of interstitial fluid lactate through an epidermal iontophoretic device, Talanta source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Label-based optical techniques are colorimetric or fluorometric methods that have demonstrated great promise for monitoring metabolites, glucose, and other biomarkers in sweat, tears, or interstitial fluid.

    “Label-based optical techniques are colorimetric or fluorometric methods [172] which have demonstrated great promise for monitoring of metabolites, glucose, and other biomarkers in sweat, tears, or interstitial fluid, providing valuable insights for disease management and preventive healthcare [173,174].”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites [172] Singh, 2023, Optical biosensors: A decade in review, Alex. Eng. J. source

      Not checked in the free run.

    • cites [173] Brunmair, 2021, Finger sweat analysis enables short interval metabolic biomonitoring in humans, Nat. Commun. source

      Not checked in the free run.

    • cites [174] Khemtonglang, K., Chaiyaphet, N., Kumsaen, T., Chaiyachati, C., and Chuchuen, O. (2022). A Smart Wristband Integrated with an IoT-Based Alarming System for Real-Time Sweat Alcohol Monitoring. Sensors, 22. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Recent studies have focused on developing novel nanomaterials for optical transduction in wearables to improve sensitivity and selectivity, such as quantum dots, carbon dots, and plasmonic nanoparticles.

    “Recent studies have focused on developing novel nanomaterials for optical transduction in wearables to improve sensitivity and selectivity, such as quantum dots, carbon dots, and plasmonic nanoparticles with exceptional optical properties [176,177,178].”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites [176] Peng, 2020, Recent Advances in Nanomaterial-Enabled Wearable Sensors: Material Synthesis, Sensor Design, and Personal Health Monitoring, Small source

      Not checked in the free run.

    • cites [177] Yao, 2017, Nanomaterial-Enabled Wearable Sensors for Healthcare, Adv. Healthc. Mater. source

      Not checked in the free run.

    • cites [178] Wu, 2013, Graphene-Based Electrochemical Sensors, Small source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    ISF has the closest composition to blood, with biomarker concentrations and temporal profiles similar to those in blood.

    “While ISF has the closest composition to the blood, with biomarker concentrations and temporal profiles similar to those in blood [159]”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [159] Teymourian, 2021, Lab under the Skin: Microneedle Based Wearable Devices, Adv. Healthc. Mater. source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Photoplethysmography enables continuous and non-invasive measurements of heart rate, blood oxygen saturation, and even blood pressure.

    “Another optical sensing technique is photoplethysmography for vital sign monitoring, which enables continuous and non-invasive measurements of heart rate, blood oxygen saturation, and even blood pressure [20].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [20] Bayoumy, 2021, Smart wearable devices in cardiovascular care: Where we are and how to move forward, Nat. Rev. Cardiol. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    A transdermal patch consisting of graphene microarrays has been developed to monitor ISF glucose through follicles.

    “So far, a transdermal patch consisting of graphene microarrays has been developed to monitor ISF glucose through follicles [169].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [169] Lipani, 2018, Non-invasive, transdermal, path-selective and specific glucose monitoring via a graphene-based platform, Nat. Nanotechnol. source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    The fully integrated wearable closed-loop system (IWCS) could accurately track glucose concentration via electrochemical sensors and accordingly deliver insulin.

    “This fully integrated wearable closed-loop system (IWCS) could accurately track glucose concentration via electrochemical sensors and accordingly deliver insulin.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [164] Li, 2021, A Fully Integrated Closed-Loop System Based on Mesoporous Microneedles-Iontophoresis for Diabetes Treatment, Adv. Sci. source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Limitations of current wearable ISF devices include possible skin irritation due to passage of electric current or needle passage, cross-contamination from sweat, and sampling of extremely low fluid volumes after long operating periods in RI methods.

    “The limitations include possible skin irritation due to the passage of electric current or needle passage, cross-contamination from sweat, and sampling of extremely low fluid volumes after long operating periods in RI methods [170].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [170] Saha, 2023, Wearable Electrochemical Glucose Sensors in Diabetes Management: A Comprehensive Review, Chem. Rev. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Optical sensing platforms often employ nanomaterials, including silver and gold nanoparticles.

    “Optical sensing platforms often employ nanomaterials, including silver and gold nanoparticles [175].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [175] Miranda, B., Rea, I., Dardano, P., De Stefano, L., and Forestiere, C. (2021). Recent Advances in the Fabrication and Functionalization of Flexible Optical Biosensors: Toward Smart Life-Sciences Applications. Biosensors, 11. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Hyaluronic acid was employed to increase osmotic pressure and drive more glucose toward the electrode.

    “They employed hyaluronic acid to increase osmotic pressure and drive more glucose toward the electrode.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [169] Lipani, 2018, Non-invasive, transdermal, path-selective and specific glucose monitoring via a graphene-based platform, Nat. Nanotechnol. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    A fully integrated wearable pacifier-based platform was used for glucose monitoring in newborns' saliva.

    “In another work [158], a fully integrated wearable pacifier-based platform was used for glucose monitoring in newborns’ saliva.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [158] Sempionatto, 2019, Pacifier Biosensor: Toward Noninvasive Saliva Biomarker Monitoring, Anal. Chem. source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    High sensitivity, non-invasiveness, and immunity to electromagnetic interference make optical sensors suitable for a wide range of wearable applications.

    “High sensitivity, non-invasiveness, and immunity to electromagnetic interference make them suitable for a wide range of wearable applications [1,171].”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [1] Ha, 2018, Wearable and flexible sensors for user-interactive health-monitoring devices, J. Mater. Chem. B source

      Not checked in the free run.

    • cites [171] Wu, M., and Luo, J. (2019). Wearable technology applications in healthcare: A literature review. Online J. Nurs. Inf., 23, Available online: https://www.himss.org/resources/wearable-technology-applications-healthcare-literature-review.

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    ML algorithms analyze data from wearable devices such as accelerometers and gyroscopes to monitor motor symptoms and allow healthcare providers to adjust medication dosages.

    “They analyze data from wearable devices, such as accelerometers and gyroscopes, to monitor motor symptoms and allow healthcare providers to adjust medication dosages [189,190].”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [189] Mukhopadhyay, A., Sumner, J., Ling, L.H., Quek, R.H.C., Tan, A.T.H., Teng, G.G., Seetharaman, S.K., Gollamudi, S.P.K., Ho, D., and Motani, M. (2022). Personalised Dosing Using the CURATE.AI Algorithm: Protocol for a Feasibility Study in Patients with Hypertension and Type II Diabetes Mellitus. Int. J. Environ. Res. Public Healthc., 19. source

      Not checked in the free run.

    • cites [190] Zhu, 2021, A machine learning approach to personalized dose adjustment of lamotrigine using noninvasive clinical parameters, Sci. Rep. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisenone given

    The need to recharge electrical devices frequently can limit their effectiveness in monitoring health status over an extended period, which is particularly problematic for those with chronic health conditions.

    “The need to recharge electrical devices frequently can limit their effectiveness in monitoring health status over an extended period, which is particularly problematic for those with chronic health conditions.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    Wearable ECG sensors continuously monitor heart activity and facilitate early detection of arrhythmias and cardiovascular abnormalities.

    “In this regard, wearable ECG sensors continuously monitor heart activity and facilitate early detection of arrhythmias and cardiovascular abnormalities [181].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [181] Camps, 2019, Machine learning in the electrocardiogram, J. Electrocardiol. source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Wearable EEG sensors can monitor brain activities and are essential in tracking encephalopathies such as stroke, brain tumor, schizophrenia, and epilepsy.

    “Wearable EEG sensors can monitor brain activities and are essential in tracking encephalopathies, such as stroke, brain tumor, schizophrenia, and epilepsy [1].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [1] Ha, 2018, Wearable and flexible sensors for user-interactive health-monitoring devices, J. Mater. Chem. B source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisenone given

    One notable advantage of ML algorithms is their ability to discern subtle changes in health metrics, which could indicate early signs of disease or other health issues.

    “One of the most notable advantages of these algorithms is their ability to discern subtle changes in health metrics, which could indicate early signs of disease or other health issues.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Wearable EMG and EOG sensors measure muscle activity and have advanced fields such as rehabilitation, robotics, and sports medicine.

    “Wearable EMG and EOG sensors measure muscle activity and have advanced fields, such as rehabilitation, robotics, and sports medicine [173,182].”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [173] Brunmair, 2021, Finger sweat analysis enables short interval metabolic biomonitoring in humans, Nat. Commun. source

      Not checked in the free run.

    • cites [182] Li, 2020, Wearable biochemical sensors for human health monitoring: Sensing materials and manufacturing technologies, J. Mater. Chem. B source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Premisenone given

    Lack of standardization with gold standard methods in data analysis makes interpretation difficult and unreliable.

    “The lack of standardization with gold standard methods in data analysis makes interpretation difficult and unreliable.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    ML algorithms are valuable in analyzing extensive data gathered by wearable devices, empowering healthcare providers to identify patterns, predict outcomes, and make suitable decisions regarding patient care.

    “ML algorithms are valuable in analyzing the extensive data gathered by wearable devices, empowering healthcare providers to identify patterns, predict outcomes, and make suitable decisions regarding patient care.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [185] Choi, 2020, Introduction to Machine Learning, Neural Networks, and Deep Learning, Transl. Vis. Sci. Technol.

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Optical transduction mechanisms in wearable biosensors face challenges of interference from ambient light and motion artifacts, which affect measurement accuracy and require advanced signal processing algorithms and motion compensation techniques.

    “The significant challenges are interference from ambient light and motion artifacts, which affect measurement accuracy and require advanced signal processing algorithms and motion compensation techniques [180].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [180] Pillai, S., Upadhyay, A., Sayson, D., Nguyen, B.H., and Tran, S.D. (2021). Advances in Medical Wearable Biosensors: Design, Fabrication and Materials Strategies in Healthcare Monitoring. Molecules, 27. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    The APT nerve can detect mechanical stimulation, discern the location of the stimulation, and transmit mechanosensitive signals.

    “The APT nerve can detect mechanical stimulation, discern the location of the stimulation, and transmit mechanosensitive signals.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [191] Liao, 2020, A bioinspired analogous nerve towards artificial intelligence, Nat. Commun. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    AROMA identifies basic activities without relying on human-made rules.

    “Human activity recognition using deep multi-task learning (AROMA), a convolutional neural network, identifies basic activities without relying on human-made rules.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [192] Peng, 2018, AROMA: A Deep MultiTask Learning Based Simple and Complex Human Activity Recognition Method Using Wearable Sensors, ACM Intereactive Mob. Wearable Ubiquitous Technol. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing). May combine two claims.

    Premisecitation

    Second-generation wearables must improve accuracy, battery life, data security, user adoption, and standardization.

    “The second-generation wearables must improve accuracy, battery life, data security, user adoption, and standardization [14,16].”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [14] Khan, 2020, Applications of Artificial Intelligence and Big Data Analytics in m-Health: A Healthcare System Perspective, J. Healthc. Eng. source

      Not checked in the free run.

    • cites [16] Haick, 2021, Artificial Intelligence in Medical Sensors for Clinical Decisions, ACS Nano source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    For more complex activities, AROMA utilizes a long-term memory network to understand the order and context of actions, demonstrating impressive recognition performance comparable to other methods.

    “For more complex activities, AROMA utilizes a “long-term memory network” to understand the order and context of actions, demonstrating impressive recognition performance comparable to other methods in this domain [192].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [192] Peng, 2018, AROMA: A Deep MultiTask Learning Based Simple and Complex Human Activity Recognition Method Using Wearable Sensors, ACM Intereactive Mob. Wearable Ubiquitous Technol. source

      Not checked in the free run.

  • May combine two claims.

    Premisenone given

    The most significant limitations of wearable sensor technology for early detection of health are sensitivity, accuracy, and technology validation.

    “The most significant limitations of wearable sensor technology for early detection of health are sensitivity, accuracy, and technology validation.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    The APT nerve was employed for playing music, controlling positioning in a two-dimensional plane, and handling rotation.

    “The APT nerve was employed for playing music, controlling positioning in a two-dimensional plane, and handling rotation [191].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [191] Liao, 2020, A bioinspired analogous nerve towards artificial intelligence, Nat. Commun. source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    The combined algorithm obtained maximum accuracies of 97.3% and 96.2% for diabetes and heart disease diagnosis, respectively.

    “This combined algorithm obtained maximum accuracies of 97.3% and 96.2% for diabetes and heart disease diagnosis, respectively, which is very promising for healthcare wearables.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    An ANN accurately predicted vertical and anterior-posterior knee joint force values for most movements, showcasing the potential of wearable sensors combined with ANN for estimating joint reactions in sports.

    “The ANN accurately predicted vertical and anterior-posterior KJF values for most movements, showcasing the potential of wearable sensors combined with ANN for estimating joint reactions in sports [199].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [199] Stetter, B.J., Ringhof, S., Krafft, F.C., Sell, S., and Stein, T. (2019). Estimation of Knee Joint Forces in Sport Movements Using Wearable Sensors and Machine Learning. Sensors, 19. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

    Premisenone given

    DL algorithms require substantial datasets for practical training, necessitating standardized data collection and analysis procedures to ensure high-quality data.

    “These algorithms require substantial datasets for practical training, necessitating standardized data collection and analysis procedures to ensure high-quality data.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    The H-LSTM method outperformed other DL methods, achieving 99.15% accuracy in recognizing human activities.

    “The results demonstrated that H-LSTM outperformed other DL methods, achieving an impressive 99.15% accuracy in recognizing human activities.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [193] Ali, 2021, An intelligent healthcare monitoring framework using wearable sensors and social networking data, Futur. Gener. Comput. Syst. source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Comparing data across a wide range of wearable devices becomes challenging, compromising the efficacy of ML algorithms in classifying health conditions.

    “Comparing data across a wide range of wearable devices becomes challenging, compromising the efficacy of ML algorithms in classifying health conditions [20,194].”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [20] Bayoumy, 2021, Smart wearable devices in cardiovascular care: Where we are and how to move forward, Nat. Rev. Cardiol. source

      Not checked in the free run.

    • cites [194] Nweke, 2018, Deep learning algorithms for human activity recognition using mobile and wearable sensor networks: State of the art and research challenges, Expert Syst. Appl. source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    A smart wristband equipped with a non-invasive sweat alcohol sensor and an IoT-based alarm system known as "Drunk Mate" could detect alcohol levels in real-time and send alerts to the user's smartphone when their blood alcohol concentration reaches an illegal limit.

    “For instance, researchers designed a smart wristband equipped with a non-invasive sweat alcohol sensor and an IoT-based alarm system known as “Drunk Mate.” This wristband could detect alcohol levels in real-time and send alerts to the user’s smartphone when their blood alcohol concentration reaches an illegal limit [174].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [174] Khemtonglang, K., Chaiyaphet, N., Kumsaen, T., Chaiyachati, C., and Chuchuen, O. (2022). A Smart Wristband Integrated with an IoT-Based Alarming System for Real-Time Sweat Alcohol Monitoring. Sensors, 22. source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Keum et al. used a smart contact lens with an electrochemical biosensor to continuously monitor glucose levels in tears and perform diabetic diagnosis and diabetic retinopathy therapy.

    “In this regard, Keum et al. [219] used a smart contact lens with an electrochemical biosensor to continuously monitor glucose levels in tears and, thus, perform diabetic diagnosis and diabetic retinopathy therapy.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [219] Keum, 2020, Wireless smart contact lens for diabetic diagnosis and therapy, Sci. Adv. source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Signal processing algorithms for high-resolution ECG sensors significantly improved the fidelity of ECG measurements, enabling early detection of arrhythmias, ischemic events, and other cardiac abnormalities.

    “For instance, signal processing algorithms for high-resolution ECG sensors significantly improved the fidelity of ECG measurements, enabling early detection of arrhythmias, ischemic events, and other cardiac abnormalities [20].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [20] Bayoumy, 2021, Smart wearable devices in cardiovascular care: Where we are and how to move forward, Nat. Rev. Cardiol. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    An ANN for these important physiology signatures could make almost individual differences negligible.

    “Thus, an ANN for these important physiology signatures could make almost individual differences negligible [237].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [237] Zhang, 2019, Skin-Friendly Electronics for Acquiring Human Physiological Signatures, Adv. Mater. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    The proposed ANSFIS model reached a mean performance accuracy of 93% for the TO phase, 94% for the MS phase, 96% for the HO phase, and approximately 100% for the HS phase.

    “The proposed ANSFIS model reached a mean performance accuracy of 93% (for the TO phase), 94% (for the MS phase), 96% (for the HO phase), and ≈100% (for the HS phase).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [227] Vu, 2020, Highly elastic capacitive pressure sensor based on smart textiles for full-range human motion monitoring, Sens. Actuators A Phys. source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Action instructions from silent speech recognition to control an intelligent car, emotion instructions to control a Bluetooth speaker, and virtual interaction in real time could be recognized with high accuracy using wavelet decomposition and machine-learning algorithms.

    “The action instructions from silent speech recognition to control an intelligent car, emotion instructions to control a Bluetooth speaker, and virtual interaction in real time could be recognized with high accuracy using wavelet decomposition and machine-learning algorithms.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [225] Liu, 2020, An epidermal sEMG tattoo-like patch as a new human–machine interface for patients with loss of voice, Microsyst. Nanoeng. source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Lee et al. performed in vitro tests and therapy on diabetic mice, with feedback treatment when hyperglycemia was detected based on the results of glucose level monitoring.

    “They performed in vitro tests and therapy on diabetic mice, with feedback treatment when hyperglycemia was detected based on the results of glucose level monitoring.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [220] Lee, 2016, A graphene-based electrochemical device with thermoresponsive microneedles for diabetes monitoring and therapy, Nat. Nanotechnol. source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Liu et al. fabricated an epidermal sEMG tattoo-like patch that could be laminated on the jaw and face and monitored the sEMG from a patient who had lost their voice.

    “Liu et al. [225] fabricated an epidermal sEMG tattoo-like patch that could be laminated on the jaw and face and monitored the sEMG from a patient who had lost their voice.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [225] Liu, 2020, An epidermal sEMG tattoo-like patch as a new human–machine interface for patients with loss of voice, Microsyst. Nanoeng. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Sleep monitoring wearable devices have gained significant popularity in consumer health and sleep research in recent years as a non-invasive and convenient tool for tracking sleep patterns and assessing sleep quality.

    “Sleep monitoring wearable devices have gained significant popularity in consumer health and sleep research in recent years as a non-invasive and convenient tool for tracking sleep patterns and assessing sleep quality [197,239] (Figure 7d).”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [197] Mostafa, S.S., Mendonça, F., Ravelo-García, A.G., and Morgado-Dias, F. (2019). A Systematic Review of Detecting Sleep Apnea Using Deep Learning. Sensors, 19. source

      Not checked in the free run.

    • cites [239] Cellini, 2019, Wearable Sleep Technology in Clinical and Research Settings, Med. Sci. Sports Exerc. source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Integration with smartphones and mobile devices enhances the accessibility of wearable ECG monitors for real-time monitoring, data storage, and cloud-based analytics, enabling seamless integration into telemedicine and remote patient monitoring systems.

    “Integration with smartphones and mobile devices enhances the accessibility of wearable ECG monitors for real-time monitoring, data storage, and cloud-based analytics, enabling seamless integration into telemedicine and remote patient monitoring systems [4,16,17].”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites [4] Baig, 2017, A Systematic Review of Wearable Patient Monitoring Systems–Current Challenges and Opportunities for Clinical Adoption, J. Med. Syst. source

      Not checked in the free run.

    • cites [16] Haick, 2021, Artificial Intelligence in Medical Sensors for Clinical Decisions, ACS Nano source

      Not checked in the free run.

    • cites [17] Vashistha, R., Dangi, A.K., Kumar, A., Chhabra, D., and Shukla, P. (2018). Futuristic biosensors for cardiac health care: An artificial intelligence approach. 3 Biotech, 8. source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    AI helped sensors with throat state identification at an accuracy above 85% and the early diagnosis of viral throat illness.

    “The AI helped sensors with throat state identification at an accuracy above 85% and the early diagnosis of viral throat illness.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [222] Lin, 2022, An ultralight, flexible, and biocompatible all-fiber motion sensor for artificial intelligence wearable electronics, npj Flex. Electron. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    The results indicate that 90.12% of displayed images were correctly identified, and 90.51% of emotionally responsive individuals accurately answered questions about time and image type.

    “The results indicate that 90.12% of displayed images were correctly identified, and 90.51% of emotionally responsive individuals accurately answered questions about time and image type [201].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [201] Swapna, M., Viswanadhula, U.M., Aluvalu, R., Vardharajan, V., and Kotecha, K. (2022). Bio-Signals in Medical Applications and Challenges Using Artificial Intelligence. J. Sens. Actuator Netw., 11. source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    With AI and ML algorithms, wearable devices provide users with personalized health recommendations, early detection of potential issues, and even life-saving interventions.

    “With cutting-edge AI and ML algorithms, these devices provide users with personalized health recommendations, early detection of potential issues, and even life-saving interventions [177,183,184,229,230,231].”
    Quote in the paperNot a substantial claim0 of 6 sources checked
    • cites [177] Yao, 2017, Nanomaterial-Enabled Wearable Sensors for Healthcare, Adv. Healthc. Mater. source

      Not checked in the free run.

    • cites [183] Qiao, 2019, Graphene-based wearable sensors, Nanoscale source

      Not checked in the free run.

    • cites [184] Yu, 2018, All VN-graphene architecture derived self-powered wearable sensors for ultrasensitive health monitoring, Nano Res. source

      Not checked in the free run.

    • cites [229] Xia, 2019, A flexible, adhesive and self-healable hydrogel-based wearable strain sensor for human motion and physiological signal monitoring, J. Mater. Chem. B source

      Not checked in the free run.

    • cites [230] Wang, 2020, Multifunctional conductive hydrogel-based flexible wearable sensors, TrAC Trends Anal. Chem. source

      Not checked in the free run.

    • cites [231] Xin, 2020, MXenes and Their Applications in Wearable Sensors, Front. Chem. source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Glucose, pH, and lactate in human sweat could be easily and accurately classified and quantified by the CNN and DL algorithm, and the testing results of actual sweat via the DL-assisted colorimetric approach matched 91.0−99.7% with the laboratory measurements.

    “Figure 7a shows glucose, pH, and lactate in human sweat could be easily and accurately classified and quantified by the CNN and DL algorithm and the testing results of actual sweat via the DL-assisted colorimetric approach matched 91.0−99.7% with the laboratory measurements [233].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [233] Liu, 2022, Explainable Deep-Learning-Assisted Sweat Assessment via a Programmable Colorimetric Chip, Anal. Chem. source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    The prevalent use of Bluetooth, NFC, or Wi-Fi for connecting wearable devices to smartphones poses security challenges, emphasizing the importance of secure wireless connectivity to mitigate data breaches.

    “The prevalent use of Bluetooth, NFC, or Wi-Fi for connecting wearable devices to smartphones poses security challenges, emphasizing the importance of secure wireless connectivity to mitigate the risk of data breaches [250].”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [250] Cilliers, 2019, Wearable devices in healthcare: Privacy and information security issues, Healthc. Inf. Manag. J.

      Not checked in the free run.

    • Rests on an assumption: Privacy and security considerations are paramount in AI-based data collection for human physiological information.

  • May combine two claims.

    Premisecitation

    These innovative devices can continuously track various health parameters, giving users personalized insights and feedback.

    “These innovative devices can continuously track various health parameters, giving users personalized insights and feedback [6,12,13,16,188].”
    Quote in the paperSubstantial claim0 of 5 sources checked
    • cites [6] Yu, 2018, Artificial intelligence in healthcare, Nat. Biomed. Eng. source

      Not checked in the free run.

    • cites [12] Harrer, 2019, Artificial Intelligence for Clinical Trial Design, Trends Pharmacol. Sci. source

      Not checked in the free run.

    • cites [13] Jin, X., Liu, C., Xu, T., Su, L., and Zhang, X. (2020). Artificial intelligence biosensors: Challenges and prospects. Biosens. Bioelectron., 165. source

      Not checked in the free run.

    • cites [16] Haick, 2021, Artificial Intelligence in Medical Sensors for Clinical Decisions, ACS Nano source

      Not checked in the free run.

    • cites [188] Raghavendra, 2019, Artificial Intelligence Techniques for Automated Diagnosis of Neurological Disorders, Eur. Neurol. source

      Not checked in the free run.

    • Rests on an assumption: AI and Machine learning enable multiple aspects of health status assessment, such as disease prediction and diagnosis, and provide effective treatment.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Chatbots are highly effective tools for delivering mental health services through digital platforms.

    “For instance, Chatbots are highly effective tools for delivering mental health services through digital platforms [186].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [186] Mody, V., and Mody, V. (2019, January 29–31). Mental Health Monitoring System using Artificial Intelligence A Review. Proceedings of the IEEE 5th International Conference for Convergence in Technology (I2CT), Bombay, India. source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    The SUN interface could differentiate tactile from proprioceptive stimuli and classify the location of the stimulus with high spatial resolution.

    “They could differentiate tactile from proprioceptive stimuli and classify the location of the stimulus with a high spatial resolution (Figure 7f).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [247] Wang, 2017, A Highly Selective 3D Spiked Ultraflexible Neural (SUN) Interface for Decoding Peripheral Nerve Sensory Information, Adv. Healthc. Mater. source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    AI can help with automated sleep stage classifiers and EEG pattern detectors with easy access to huge sleep datasets such as the National Sleep Research Resource.

    “In addition, AI can help with automated sleep stage classifiers and EEG pattern detectors with easy access to huge sleep datasets, such as the National Sleep Research Resource [243].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [243] Arnal, 2022, The Dreem Headband as an Alternative to Polysomnography for EEG Signal Acquisition and Sleep Staging, Sleep source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    The SUN interface enabled deformation following nerve movements and penetrating spiked electrodes for intra-fascicular recordings.

    “This enabled the SUN interface to deform following nerve movements and penetrating spiked electrodes for intra-fascicular recordings.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [247] Wang, 2017, A Highly Selective 3D Spiked Ultraflexible Neural (SUN) Interface for Decoding Peripheral Nerve Sensory Information, Adv. Healthc. Mater. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based). May combine two claims.

    Interpretationreasoning

    By improving user experience and wearability, with researchers designing lightweight and comfortable devices and providing feedback using AI-enabled sleep monitoring devices, positive and healthy sleep habits can be expected.

    “By improving user experience and wearability, with researchers designing lightweight and comfortable devices and providing feedback using AI-enabled sleep monitoring devices, positive and healthy sleep habits can be expected.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    Integrating AI technologies has transformed mental health care, enabling new possibilities for identifying and intervening in mental health issues via immediate and personalized support.

    “Integrating AI technologies has transformed mental health care, enabling new possibilities for identifying and intervening in mental health issues via immediate and personalized support [186].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [186] Mody, V., and Mody, V. (2019, January 29–31). Mental Health Monitoring System using Artificial Intelligence A Review. Proceedings of the IEEE 5th International Conference for Convergence in Technology (I2CT), Bombay, India. source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    The persistent operation of AI algorithms for data processing can strain battery life, necessitating a balance between sophisticated processing needs and practical battery constraints.

    “Yet the persistent operation of AI algorithms for data processing can strain battery life, necessitating a delicate balance between sophisticated processing needs and practical battery constraints [248].”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [248] Vijayan, V., Connolly, J.P., Condell, J., McKelvey, N., and Gardiner, P. (2021). Review of Wearable Devices and Data Collection Considerations for Connected Health. Sensors, 21. source

      Not checked in the free run.

    • Rests on an assumption: Continuous data reading is imperative for wearables to comprehensively capture an individual’s physiological information over an extended period.
    • Rests on an assumption: Electronic readers for wearable devices must possess long-lasting battery power to sustain continuous data reading.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Implantable sensors are designed to interact with biological components for support and treatment of bodily functions and mental health monitoring.

    “Implantable sensors represent a frontier in healthcare, designed to interact with biological components for support and treatment of bodily functions and mental health monitoring [245,246].”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [245] Lee, 2019, From flexible electronics technology in the era of IoT and artificial intelligence toward future implanted body sensor networks, APL Mater. source

      Not checked in the free run.

    • cites [246] Wang, C., He, T., Zhou, H., Zhang, Z., and Lee, C. (2023). Artificial intelligence enhanced sensors—Enabling technologies to next-generation healthcare and biomedical platform. Bioelectron. Med., 9. source

      Not checked in the free run.

    • Rests on an assumption: Lost functions resulting from neural damage can be monitored using neural prostheses.

  • May combine two claims.

    Interpretationreasoning

    Using AI-enabled wearable sensor technology improves health management, disease diagnostics, and speed of treatments.

    “This improves health management, disease diagnostics, and speed of treatments.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: AI and Machine learning enable multiple aspects of health status assessment, such as disease prediction and diagnosis, and provide effective treatment.

  • May combine two claims.

    Premisecitation

    Lower cost, time, and fewer visits are key benefits of using AI-enabled wearable sensor technology in healthcare systems.

    “Lower cost, time, and fewer visits are some of the key benefits of using AI-enabled wearable sensor technology in healthcare systems [254].”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [254] Junaid, S.B., Imam, A.A., Abdulkarim, M., Surakat, Y.A., Balogun, A.O., Kumar, G., Shuaibu, A.N., Garba, A., Sahalu, Y., and Mohammed, A. (2022). Recent Advances in Artificial Intelligence and Wearable Sensors in Healthcare Delivery. Appl. Sci., 12. source

      Not checked in the free run.

    • Rests on an assumption: AI and Machine learning enable multiple aspects of health status assessment, such as disease prediction and diagnosis, and provide effective treatment.
84 of 84

  • The paper relies on this without support.

    Wearable sensors can accurately measure physical states and biochemical signals.

    “Wearable sensors have opened a new area of personalized health monitoring by accurately measuring physical states and biochemical signals.”

  • The paper relies on this without support.

    Continuous data reading is imperative for wearables to comprehensively capture an individual’s physiological information over an extended period.

    “Continuous data reading is imperative for these devices to comprehensively capture an individual’s physiological information over an extended period.”

  • The paper relies on this without support.

    Electronic readers for wearable devices must possess long-lasting battery power to sustain continuous data reading.

    “Electronic readers for wearable devices must possess long-lasting battery power to sustain continuous data reading.”

  • The paper relies on this without support.

    The accuracy and reliability of data collected by wearables are critical, demanding precise sensor calibration, interference minimization, and accounting for variations in user activities and physiology.

    “The accuracy and reliability of data collected by these wearables are critical, demanding precise sensor calibration, interference minimization, and accounting for variations in user activities and physiology.”

  • The paper relies on this without support.

    Privacy and security considerations are paramount in AI-based data collection for human physiological information.

    “In the realm of AI-based data collection for human physiological information, privacy and security considerations are paramount.”

  • The paper relies on this without support.

    The repeatability, reproducibility, and dynamic stability of sensors remain an issue.

    “However, the repeatability, reproducibility, and dynamic stability of sensors remain an issue.”

  • The paper relies on this without support.

    Current sweat stimulation methods, including iontophoresis, sauna, and exercise for on-body access or collecting tears, are still uncomfortable for everyone.

    “For instance, current sweat stimulation methods, including iontophoresis, sauna, and exercise for on-body access or collecting tears, are still uncomfortable for everyone.”

  • The paper relies on this without support.

    AI technologies cannot replace the essential human element of empathy and emotional connection in mental health care.

    “However, it is crucial to recognize that AI technologies cannot replace the essential human element of empathy and emotional connection in mental health care.”

  • The paper relies on this without support.

    Precision therapy and diagnostic systems have contributed to early detection and individualized treatment, leading to more efficient interventions.

    “Precision therapy and diagnostic systems have contributed to early detection and individualized treatment, leading to more efficient interventions.”

  • The paper relies on this without support.

    Lost functions resulting from neural damage can be monitored using neural prostheses.

    “For instance, lost functions resulting from neural damage can be monitored using neural prostheses.”

  • The paper relies on this without support.

    AI and Machine learning enable multiple aspects of health status assessment, such as disease prediction and diagnosis, and provide effective treatment.

    “AI and Machine learning enable multiple aspects of health status assessment, such as disease prediction and diagnosis, and provide effective treatment.”

  • The paper relies on this without support.

    These devices accumulate a wealth of sensitive data, including heart rate, sleep patterns, and location information.

    “These devices accumulate a wealth of sensitive data, including heart rate, sleep patterns, and location information.”

  • The paper relies on this without support.

    Wearable sensors are the interface to body organs and therefore must match the mechanical flexibility of human tissues to reduce negative long-term effects and enable sensing transduction with high signal-to-noise ratio.

    “Wearable sensors are the interface to body organs (such as skin, muscles, and eyes) and therefore they must match the mechanical flexibility of human tissues to reduce any negative long-term effects and enable sensing transduction mechanism with high signal-to-noise ratio.”

  • The paper relies on this without support.

    Different nanomaterials and fabrication technologies have been used to improve sensor platforms’ sensitivity, selectivity, and physical and mechanical performance.

    “In addition, different nanomaterials and fabrication technologies have been used to improve the sensor platforms’ sensitivity, selectivity, and physical and mechanical performance.”

  • The paper relies on this without support.

    The current research focuses on using AI technology in biosensing sweat and blood biomarkers.

    “The current research focuses on using AI technology in biosensing sweat and blood biomarkers.”

  • The paper relies on this without support.

    Machine learning and big data processing can help to predict general and complex health conditions.

    “Machine learning and big data processing can help to predict general and complex health conditions.”

  • The paper relies on this without support.

    Miniaturization and lower power consumption can help AI-sensors function more efficiently.

    “Miniaturization and lower power consumption can help AI-sensors function more efficiently.”
17 of 17

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • Self-collected data such as those obtained from wearable devices or POCT results have been suggested to be useful in improving patient care, particularly in future applications using machine learning or AI tools.

    “Recently, self-collected data such as those obtained from wearable devices or POCT results, have been suggested to be useful in improving patient care, particularly in future applications using machine learning (ML) or AI tools [59].”

    From Point-of-care testing: state-of-the art and perspectives (2025) · cites this paper as [59]

    The Introduction says AI and ML can analyze wearable-sensor data to inform patient care, and Section 5.2 describes future uses for improving diagnosis and treatment. The cited passages do not establish the same claim for POCT results as a distinct source of self-collected data. (LLM, conf=0.85)

    Experimental — accuracy not yet measured

  • In online monitoring, variations in data collected from different types of wearable devices may minimize the accuracy of predictive models in healthcare applications.

    “In online monitoring, variations in data collected from different types of wearable devices may minimize the accuracy of predictive models in healthcare applications [114].”

    From Federated Learning in Smart Healthcare: A Comprehensive Review on Privacy, Security, and Predictive Analytics with IoT Integration (2024) · cites this paper as [114]

  • Wearable sensors integrated with AI can facilitate continuous patient monitoring, signaling subtle physiological changes before they manifest as complications.

    “For instance, wearable sensors integrated with AI can facilitate continuous patient monitoring, signaling subtle physiological changes before they manifest as complications (17).”

    From Artificial intelligence in nursing: an integrative review of clinical and operational impacts (2025) · cites this paper as [17]

Cites

2 retracted

Flagged first, then by citations

Research works printed but not linked: 2

  1. Wu, M., and Luo, J. (2019). Wearable technology applications in healthcare: A literature review. Online J. Nurs. Inf., 23, Available online: https://www.himss.org/resources/wearable-technology-applications-healthcare-literature-review.
  2. Biswal, S., Kulas, J., Sun, H., Goparaju, B., Westover, M.B., Bianchi, M.T., and Sun, J. (2017). SLEEPNET Automated Sleep Staging System via Deep Learning. arXiv.

Cited by

Sorted by citations

Related

Authors
Shajari, S., Kuruvinashetti, K., Komeili, A., & Sundararaj, U.
Affiliations
  • Northwestern University (US)
  • University of Calgary (CA)
Corresponding author
Sundararaj, Uttandaraman
ORCID
Venue
Sensors
Publisher
MDPI AG
Published
2023-11-29
Volume
23
Issue
23
Pages
9498
ISSN
1424-8220 (ISSN-L)
DOI
10.3390/s23239498
PMC
PMC10708748
OpenAlex
W4389203612
Semantic Scholar
384e8788685cbb444314e7447e9844398d189858
Cited by
535
References
252
Citations by year
2026: 197 · 2025: 267 · 2024: 71
FWCI
11.99
Type
journal article
Language
en
Open access
Licence
Locations
  • Sensors (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Topics
  • Physical Sciences › Engineering › Biomedical Engineering › Advanced Sensor and Energy Harvesting Materials
  • Physical Sciences › Engineering › Biomedical Engineering › Non-Invasive Vital Sign Monitoring
  • Physical Sciences › Engineering › Electrical and Electronic Engineering › Nanomaterials and Printing Technologies
Keywords
wearable sensors, artificial intelligence, flexible electronics, digital health, biosensors, physical sensors, chemical sensors, transduction mechanisms, personalized health monitoring, disease diagnosis, self-learning, power efficiency, real-time data processing
MeSH terms
Wearable Electronic Devices, Artificial Intelligence, Humans, Monitoring, Physiologic, Biosensing Techniques

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-09.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-09.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review