[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
You seem to have javascript disabled. Please note that many of the page functionalities won't work as expected without javascript enabled.
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
7.3
3.4
Journals
Sensors
Special Issues
Sensing-Based Biomedical Communication and Intelligent Identification for Healthcare
sensors-logo
Submit to Sensors Review for Sensors Propose a Special Issue

Journal Menu

► Journal Menu

Journal Browser

► Journal Browser
share announcement

Sensing-Based Biomedical Communication and Intelligent Identification for Healthcare

A special issue of Sensors (ISSN 1424-8220). This special issue belongs to the section "Intelligent Sensors".

Deadline for manuscript submissions: closed (10 September 2023) | Viewed by 41900

Special Issue Editors

Dr. Wenyan Jia

E-Mail Website
Guest Editor
Department of Electrical and Computer Engineering, University of Pittsburgh, Pittsburgh, PA 15260, USA
Interests: biomedical signal and image processing; wearable electronic devices; the implementation of mobile technology in healthcare
Special Issues, Collections and Topics in MDPI journals
Prof. Dr. Yi Gao

E-Mail Website
Guest Editor
Department of Medical Information Engineering, Shenzhen University, Shenzhen 518060, China
Interests: medical image computing; artificial intelligence
Prof. Dr. Zhi-Hong Mao

E-Mail Website
Guest Editor
Department of Electrical and Computer Engineering and Department of Bioengineering at the University of Pittsburgh, Pittsburgh, PA 15260, USA
Interests: human-in-the-loop control systems; control of networked and large-scale dynamical systems
Prof. Dr. Mingui Sun

E-Mail Website
Guest Editor
University of Pittsburgh, Department of Neurosurgery, Pittsburgh, PA, USA
Interests: neurophysiological signals and systems; biosensor designs; brain-computer interface; bioelectronics and bioinformatics

Special Issue Information

Dear Colleagues,

Due to the wide applications of sensors in the field of healthcare, analysis of data acquired by these sensors (e.g., electronic sensor, optical sensor, biosensor, chemical sensor) has been playing an important role in disease screening, diagnosis, monitoring, and communication between sensors/systems and people. This Special Issue is focused on sensing-based biomedical communication and intelligent identification (or recognition) of information in signals/images for healthcare applications. The term intelligent identification includes both deep leaning and conventional machine learning approaches. Potentially valuable information or knowledge for healthcare can be obtained through intelligent identification, and delivered to patients/healthcare providers effectively, efficiently and securely. Topics of interests include, but are not limited to, egocentric image analysis for action/activity recognition, motion/gait analysis using wearable sensors, biometrics identification for security, innovative vital signal (e.g., temperature, blood pressure) monitoring, and human–computer interface via invasive or non-invasive sensors.

We strongly encourage submissions exploring the advances in sensor-based signal/image analysis for health monitoring, disease diagnosis, and human-computer interface. Survey papers and reviews are also welcome.

Dr. Wenyan Jia
Prof. Dr. Yi Gao
Prof. Dr. Zhi-Hong Mao
Prof. Dr. Mingui Sun
Guest Editors

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Sensors is an international peer-reviewed open access semimonthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2600 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

Keywords

  • sensors
  • sensing systems
  • signal or image processing
  • machine learning/deep learning
  • intelligent identification or analysis
  • biomedical communication
  • disease screening, diagnosis and monitoring
  • health systems, healthcare, and wellness

Benefits of Publishing in a Special Issue

  • Ease of navigation: Grouping papers by topic helps scholars navigate broad scope journals more efficiently.
  • Greater discoverability: Special Issues support the reach and impact of scientific research. Articles in Special Issues are more discoverable and cited more frequently.
  • Expansion of research network: Special Issues facilitate connections among authors, fostering scientific collaborations.
  • External promotion: Articles in Special Issues are often promoted through the journal's social media, increasing their visibility.
  • e-Book format: Special Issues with more than 10 articles can be published as dedicated e-books, ensuring wide and rapid dissemination.

Further information on MDPI's Special Issue polices can be found here.

Published Papers (12 papers)

Download All Papers
Order results
Result details
Show export options expand_more Show export options expand_less
Select all
Export citation of selected articles as:

Editorial

Jump to: Research, Review

3 pages, 140 KiB  
Editorial
Editorial for the Special Issue “Sensing-Based Biomedical Communication and Intelligent Identification for Healthcare”
by Wenyan Jia, Yi Gao, Zhi-Hong Mao and Mingui Sun
Sensors 2024, 24(5), 1403; https://doi.org/10.3390/s24051403 - 22 Feb 2024
Cited by 1 | Viewed by 908
Abstract
The integration of sensor technology in healthcare has become crucial for disease diagnosis and treatment [...] Full article
(This article belongs to the Special Issue Sensing-Based Biomedical Communication and Intelligent Identification for Healthcare)

Research

Jump to: Editorial, Review

17 pages, 1978 KiB  
Article
Fast Parabolic Fitting: An R-Peak Detection Algorithm for Wearable ECG Devices
by Ramón A. Félix, Alberto Ochoa-Brust, Walter Mata-López, Rafael Martínez-Peláez, Luis J. Mena and Laura L. Valdez-Velázquez
Sensors 2023, 23(21), 8796; https://doi.org/10.3390/s23218796 - 28 Oct 2023
Cited by 1 | Viewed by 2550
Abstract
Heart diseases rank among the most fatal health concerns globally, with the majority being preventable through early diagnosis and effective treatment. Electrocardiogram (ECG) analysis is critical in detecting heart diseases, as it captures the heart’s electrical activities. For continuous monitoring, wearable electrocardiographic devices [...] Read more.
Heart diseases rank among the most fatal health concerns globally, with the majority being preventable through early diagnosis and effective treatment. Electrocardiogram (ECG) analysis is critical in detecting heart diseases, as it captures the heart’s electrical activities. For continuous monitoring, wearable electrocardiographic devices must ensure user comfort over extended periods, typically 24 to 48 h. These devices demand specialized algorithms with low computational complexity to accommodate memory and power consumption constraints. One of the most crucial aspects of ECG signals is accurately detecting heartbeat intervals, specifically the R peaks. In this study, we introduce a novel algorithm designed for wearable devices, offering two primary attributes: robustness against noise and low computational complexity. Our algorithm entails fitting a least-squares parabola to the ECG signal and adaptively shaping it as it sweeps through the signal. Notably, our proposed algorithm eliminates the need for band-pass filters, which can inadvertently smooth the R peaks, making them more challenging to identify. We compared the algorithm’s performance using two extensive databases: the meta-database QT database and the BIH-MIT database. Importantly, our method does not necessitate the precise localization of the ECG signal’s isoelectric line, contributing to its low computational complexity. In the analysis of the QT database, our algorithm demonstrated a substantial advantage over the classical Pan-Tompkins algorithm and maintained competitiveness with state-of-the-art approaches. In the case of the BIH-MIT database, the performance results were more conservative; they continued to underscore the real-world utility of our algorithm in clinical contexts. Full article
(This article belongs to the Special Issue Sensing-Based Biomedical Communication and Intelligent Identification for Healthcare)
►▼ Show Figures

Figure 1

Figure 1
<p>Parabola fitting at a P peak.</p> Full article ">Figure 2
<p>Parabolic approximations (red parabolas) in different points (green circles) of the ECG signal with different parabolic heights.</p> Full article ">Figure 3
<p>A true R peak (above red asterisk) and a false positive (below red asterisk) with higher parabolic height.</p> Full article ">Figure 4
<p>The proposed algorithm structure.</p> Full article ">Figure 5
<p>Search for the best R-peak candidate on an ECG signal.</p> Full article ">Figure 6
<p>False negative is due to its reduced parabolic height.</p> Full article ">Figure 7
<p>False positive R peaks provoked by high-frequency noise.</p> Full article ">
15 pages, 1969 KiB  
Article
Improving Diagnostics with Deep Forest Applied to Electronic Health Records
by Atieh Khodadadi, Nima Ghanbari Bousejin, Soheila Molaei, Vinod Kumar Chauhan, Tingting Zhu and David A. Clifton
Sensors 2023, 23(14), 6571; https://doi.org/10.3390/s23146571 - 21 Jul 2023
Cited by 3 | Viewed by 2309
Abstract
An electronic health record (EHR) is a vital high-dimensional part of medical concepts. Discovering implicit correlations in the information of this data set and the research and informative aspects can improve the treatment and management process. The challenge of concern is the data [...] Read more.
An electronic health record (EHR) is a vital high-dimensional part of medical concepts. Discovering implicit correlations in the information of this data set and the research and informative aspects can improve the treatment and management process. The challenge of concern is the data sources’ limitations in finding a stable model to relate medical concepts and use these existing connections. This paper presents Patient Forest, a novel end-to-end approach for learning patient representations from tree-structured data for readmission and mortality prediction tasks. By leveraging statistical features, the proposed model is able to provide an accurate and reliable classifier for predicting readmission and mortality. Experiments on MIMIC-III and eICU datasets demonstrate Patient Forest outperforms existing machine learning models, especially when the training data are limited. Additionally, a qualitative evaluation of Patient Forest is conducted by visualising the learnt representations in 2D space using the t-SNE, which further confirms the effectiveness of the proposed model in learning EHR representations. Full article
(This article belongs to the Special Issue Sensing-Based Biomedical Communication and Intelligent Identification for Healthcare)
►▼ Show Figures

Figure 1

Figure 1
<p>Schematic representation of Patient Forest, illustrating the process of patient outcome prediction based on learnt EHR representations from a prepared EHR matrix.</p> Full article ">Figure 2
<p>Readmission and mortality prediction performance on eICU and MIMIC in terms of AUPRC with different data splits: 50%:50% (<b>a</b>), 30%:70% (<b>b</b>), and average of all settings (<b>c</b>).</p> Full article ">Figure 3
<p>t-SNE embeddings of the patients in the MIMIC dataset based on raw (<b>left</b>) and learnt Patient Forest features (<b>right</b>).</p> Full article ">
16 pages, 1542 KiB  
Article
TCU-Net: Transformer Embedded in Convolutional U-Shaped Network for Retinal Vessel Segmentation
by Zidi Shi, Yu Li, Hua Zou and Xuedong Zhang
Sensors 2023, 23(10), 4897; https://doi.org/10.3390/s23104897 - 19 May 2023
Cited by 5 | Viewed by 2152
Abstract
Optical coherence tomography angiography (OCTA) provides a detailed visualization of the vascular system to aid in the detection and diagnosis of ophthalmic disease. However, accurately extracting microvascular details from OCTA images remains a challenging task due to the limitations of pure convolutional networks. [...] Read more.
Optical coherence tomography angiography (OCTA) provides a detailed visualization of the vascular system to aid in the detection and diagnosis of ophthalmic disease. However, accurately extracting microvascular details from OCTA images remains a challenging task due to the limitations of pure convolutional networks. We propose a novel end-to-end transformer-based network architecture called TCU-Net for OCTA retinal vessel segmentation tasks. To address the loss of vascular features of convolutional operations, an efficient cross-fusion transformer module is introduced to replace the original skip connection of U-Net. The transformer module interacts with the encoder’s multiscale vascular features to enrich vascular information and achieve linear computational complexity. Additionally, we design an efficient channel-wise cross attention module to fuse the multiscale features and fine-grained details from the decoding stages, resolving the semantic bias between them and enhancing effective vascular information. This model has been evaluated on the dedicated Retinal OCTA Segmentation (ROSE) dataset. The accuracy values of TCU-Net tested on the ROSE-1 dataset with SVC, DVC, and SVC+DVC are 0.9230, 0.9912, and 0.9042, respectively, and the corresponding AUC values are 0.9512, 0.9823, and 0.9170. For the ROSE-2 dataset, the accuracy and AUC are 0.9454 and 0.8623, respectively. The experiments demonstrate that TCU-Net outperforms state-of-the-art approaches regarding vessel segmentation performance and robustness. Full article
(This article belongs to the Special Issue Sensing-Based Biomedical Communication and Intelligent Identification for Healthcare)
►▼ Show Figures

Figure 1

Figure 1
<p>Comparison of color fundus images and fovea-centred (yellow rectangle area) OCTA images: (<b>a</b>) color fundus, (<b>b</b>–<b>d</b>) superficial vascular complexes (SVC), deep vascular complexes (DVC), and the inner retina vascular plexus including both SVC and DVC (SVC+DVC). (<b>e</b>–<b>h</b>) are their corresponding labels. The small vessels usually have low contrast.</p> Full article ">Figure 2
<p>(<b>a</b>) Illustration of the proposed TCU-Net, (<b>b</b>) efficient cross-fusion transformer module, and (<b>c</b>) efficient channel cross-attention module.</p> Full article ">Figure 3
<p>The encoder output is subjected to an interpolation downsampling operation to obtain the cross-scale <math display="inline"><semantics> <mrow> <msub> <mrow> <msup> <mi mathvariant="bold">Q</mi> <mo>′</mo> </msup> </mrow> <mi>i</mi> </msub> <mrow> <mo>(</mo> <mi>i</mi> <mo>=</mo> <mn>1</mn> <mo>,</mo> <mn>2</mn> <mo>,</mo> <mn>3</mn> <mo>,</mo> <mn>4</mn> <mo>)</mo> </mrow> <mo>,</mo> <msup> <mi mathvariant="bold">K</mi> <mo>′</mo> </msup> <mo>,</mo> <msup> <mi mathvariant="bold">V</mi> <mo>′</mo> </msup> </mrow> </semantics></math>.</p> Full article ">Figure 4
<p>Efficient multihead cross-attention.</p> Full article ">Figure 5
<p>Vessel segmentation results from different methods on different layers of ROSE-1 and ROSE-2. From (<b>left</b>) to (<b>right</b>): en face angiograms (original images), manual annotations, and vessel segmentation results obtained by TransFuse, TransUnet, OCTA-Net, and the proposed method (TCU-Net), respectively.</p> Full article ">Figure 6
<p>Effect of size reduction and projection of efficient self-attention on ROSE-1 dataset.</p> Full article ">Figure 7
<p>Effect of size reduction size and projection of efficient self-attention on ROSE-2 dataset.</p> Full article ">
25 pages, 2227 KiB  
Article
Electrical Bioimpedance Analysis for Evaluating the Effect of Pelotherapy on the Human Skin: Methodology and Experiments
by Margus Metshein, Varje-Riin Tuulik, Viiu Tuulik, Monika Kumm, Mart Min and Paul Annus
Sensors 2023, 23(9), 4251; https://doi.org/10.3390/s23094251 - 25 Apr 2023
Cited by 5 | Viewed by 2339
Abstract
Background: Pelotherapy is the traditional procedure of applying curative muds on the skin’s surface—shown to have a positive effect on the human body and cure illnesses. The effect of pelotherapy is complex, functioning through several mechanisms, and depends on the skin’s functional condition. [...] Read more.
Background: Pelotherapy is the traditional procedure of applying curative muds on the skin’s surface—shown to have a positive effect on the human body and cure illnesses. The effect of pelotherapy is complex, functioning through several mechanisms, and depends on the skin’s functional condition. The current research objective was to develop a methodology and electrodes to assess the passage of the chemical and biologically active compounds of curative mud through human skin by performing electrical bioimpedance (EBI) analysis. Methods: The methodology included local area mud pack and simultaneous tap water compress application on the forearms with the comparison to the measurements of the dry skin. A custom-designed small-area gold-plated electrode on a rigid printed circuit board, in a tetrapolar configuration, was designed. A pilot study experiment with ten volunteers was performed. Results: Our results indicated the presence of an effect of pelotherapy, manifested by the varying electrical properties of the skin. Distinguishable difference in the measured real part of impedance (R) emerged, showing a very strong correlation between the dry and tap-water-treated skin (r = 0.941), while a poor correlation between the dry and mud-pack-treated skin (r = 0.166) appeared. The findings emerged exclusively in the frequency interval of 10 kHz …1 MHz and only for R. Conclusions: EBI provides a promising tool for monitoring the variations in the electrical properties of the skin, including the skin barrier. We foresee developing smart devices for promoting the exploitation of spa therapies. Full article
(This article belongs to the Special Issue Sensing-Based Biomedical Communication and Intelligent Identification for Healthcare)
►▼ Show Figures

Figure 1

Figure 1
<p>Placement of the custom-designed electrode on the skin for measuring the electrical bioimpedance (EBI).</p> Full article ">Figure 2
<p>Attachment of mud compress (<b>a</b>) and water compress (<b>b</b>) on the forearm of a volunteer.</p> Full article ">Figure 3
<p>Bottom (electrode) (<b>a</b>) and top (connector) (<b>b</b>) side of the custom-designed printed circuit board (PCB).</p> Full article ">Figure 4
<p>Visual representation of structures with layers of different materials in series (<b>a</b>) and in parallel (<b>b</b>) with the corresponding equivalent circuits.</p> Full article ">Figure 5
<p>Measured and calculated curves of impedance (<span class="html-italic">Z</span>) in the frequency scale (<b>a</b>) and the composed equivalent circuits for the verification of the measurement setup (<b>b</b>) in the cases of a single resistor (A), series (B); and parallel connection of the resistor and capacitor (C) (where the arrow in the circle means the excitation alternating current source, the letter V in the circle means the voltage measurement, the boxes with the values denote the resistors, and the two bold lines in parallel denote the capacitor).</p> Full article ">Figure 6
<p>Calculated means of measured <span class="html-italic">Z</span> in the cases of all volunteers before and after the application of mud and water compresses in the frequency domain.</p> Full article ">Figure 7
<p>Calculated means of measured phase angle (<math display="inline"><semantics> <mi>θ</mi> </semantics></math>) in the cases of all volunteers before and after the application of mud and water compresses in the frequency domain.</p> Full article ">Figure 8
<p>Calculated standard deviations (<math display="inline"><semantics> <mi>σ</mi> </semantics></math>) of measured values of <span class="html-italic">Z</span> before the application of mud/water compresses throughout the whole frequency range.</p> Full article ">Figure 9
<p>Means of measured conductance (<span class="html-italic">G</span>) (<b>a</b>) and susceptance (<span class="html-italic">B</span>) (<b>b</b>) before and after the application of mud and water compresses in the frequency domain.</p> Full article ">Figure 10
<p>Means of measured admittance (<span class="html-italic">Y</span>) (by its real (<span class="html-italic">G</span>) and imaginary (<span class="html-italic">B</span>) parts) in the cases of the performed measurements before and after the application of mud and water compresses in the complex plane in the full frequency interval.</p> Full article ">Figure 11
<p>Means of measured real (<span class="html-italic">R</span>) (<b>a</b>) and imaginary (<span class="html-italic">X</span>) (<b>b</b>) parts of <span class="html-italic">Z</span> before and after the application of mud and water compresses in the frequency domain.</p> Full article ">Figure 12
<p>Means of measured <span class="html-italic">Z</span> (by its real (<span class="html-italic">R</span>) and imaginary (<span class="html-italic">X</span>) parts) in the complex plane in the cases of all performed measurements before and after the application of mud and water compresses in the full frequency interval.</p> Full article ">Figure 13
<p>Means of measured <span class="html-italic">Z</span> in the case of measurement after the application of mud and water compresses in the complex plane in the full frequency interval.</p> Full article ">Figure 14
<p>Linear regression graphs for the measured values of <span class="html-italic">R</span> in the 10 kHz …1 MHz frequency interval in the cases of before the mud/water compress to after the mud compress (<math display="inline"><semantics> <mrow> <mi>r</mi> <msub> <mi>R</mi> <mrow> <mi>B</mi> <mi>M</mi> <mi>W</mi> <mi>C</mi> <mo>−</mo> <mi>A</mi> <mi>M</mi> <mi>C</mi> </mrow> </msub> </mrow> </semantics></math>) (<b>a</b>), before the mud/water compress to after the water compress (<math display="inline"><semantics> <mrow> <mi>r</mi> <msub> <mi>R</mi> <mrow> <mi>B</mi> <mi>M</mi> <mi>W</mi> <mi>C</mi> <mo>−</mo> <mi>A</mi> <mi>W</mi> <mi>C</mi> </mrow> </msub> </mrow> </semantics></math>) (<b>b</b>), and after the mud compress to after the water compress (<math display="inline"><semantics> <mrow> <mi>r</mi> <msub> <mi>R</mi> <mrow> <mi>A</mi> <mi>M</mi> <mi>C</mi> <mo>−</mo> <mi>A</mi> <mi>W</mi> <mi>C</mi> </mrow> </msub> </mrow> </semantics></math>) (<b>c</b>), together with the linear regression equations and calculated coefficients of determination (<math display="inline"><semantics> <msup> <mi>R</mi> <mn>2</mn> </msup> </semantics></math>).</p> Full article ">
18 pages, 11131 KiB  
Article
Application of Feedforward and Recurrent Neural Networks for Fusion of Data from Radar and Depth Sensors Applied for Healthcare-Oriented Characterisation of Persons’ Gait
by Paweł Mazurek
Sensors 2023, 23(3), 1457; https://doi.org/10.3390/s23031457 - 28 Jan 2023
Cited by 2 | Viewed by 1649
Abstract
In this paper, the useability of feedforward and recurrent neural networks for fusion of data from impulse-radar sensors and depth sensors, in the context of healthcare-oriented monitoring of elderly persons, is investigated. Two methods of data fusion are considered, viz., one based on [...] Read more.
In this paper, the useability of feedforward and recurrent neural networks for fusion of data from impulse-radar sensors and depth sensors, in the context of healthcare-oriented monitoring of elderly persons, is investigated. Two methods of data fusion are considered, viz., one based on a multilayer perceptron and one based on a nonlinear autoregressive network with exogenous inputs. These two methods are compared with a reference method with respect to their capacity for decreasing the uncertainty of estimation of a monitored person’s position and uncertainty of estimation of several parameters enabling medical personnel to make useful inferences on the health condition of that person, viz., the number of turns made during walking, the travelled distance, and the mean walking speed. Both artificial neural networks were trained on the synthetic data. The numerical experiments show the superiority of the method based on a nonlinear autoregressive network with exogenous inputs. This may be explained by the fact that for this type of network, the prediction of the person’s position at each time instant is based on the position of that person at the previous time instants. Full article
(This article belongs to the Special Issue Sensing-Based Biomedical Communication and Intelligent Identification for Healthcare)
►▼ Show Figures

Figure 1

Figure 1
<p>Artificial neural networks used for fusion of data: the nonlinear autoregressive network with exogenous inputs (<b>a</b>) and the multilayer perceptron (<b>b</b>); <math display="inline"><semantics> <mrow> <msub> <mi>x</mi> <mi mathvariant="normal">r</mi> </msub> </mrow> </semantics></math> and <math display="inline"><semantics> <mrow> <msub> <mi>y</mi> <mi mathvariant="normal">r</mi> </msub> </mrow> </semantics></math> denote the coordinates acquired by means of the impulse-radar sensor; <math display="inline"><semantics> <mrow> <msub> <mi>x</mi> <mi mathvariant="normal">d</mi> </msub> </mrow> </semantics></math> and <math display="inline"><semantics> <mrow> <msub> <mi>y</mi> <mi mathvariant="normal">d</mi> </msub> </mrow> </semantics></math> denote the coordinates acquired by means of the depth sensor; <math display="inline"><semantics> <mrow> <msub> <mi>x</mi> <mi mathvariant="normal">f</mi> </msub> </mrow> </semantics></math> and <math display="inline"><semantics> <mrow> <msub> <mi>y</mi> <mi mathvariant="normal">f</mi> </msub> </mrow> </semantics></math> denote the fused coordinates; <math display="inline"><semantics> <mrow> <msub> <mi mathvariant="normal">h</mi> <mn>1</mn> </msub> <mo>,</mo> <mtext> </mtext> <mo>…</mo> <mo>,</mo> <mtext> </mtext> <msub> <mi mathvariant="normal">h</mi> <mn>8</mn> </msub> </mrow> </semantics></math> denote the neurons in the hidden layers; <math display="inline"><semantics> <mrow> <msub> <mi mathvariant="normal">o</mi> <mn>1</mn> </msub> </mrow> </semantics></math> and <math display="inline"><semantics> <mrow> <msub> <mi mathvariant="normal">o</mi> <mn>2</mn> </msub> </mrow> </semantics></math> denote the neurons in the output layers.</p> Full article ">Figure 2
<p>Reference trajectories used for generation of the synthetic data.</p> Full article ">Figure 3
<p>Synthetic data used for training the neural networks: the radar data (<b>left column</b>), the depth data without occlusion (<b>middle column</b>) and the depth data with occlusion (<b>right column</b>). Each graph for radar data depicts 40 superimposed sequences of synthetic data, while each graph for depth data depicts 20 superimposed sequences of synthetic data.</p> Full article ">Figure 3 Cont.
<p>Synthetic data used for training the neural networks: the radar data (<b>left column</b>), the depth data without occlusion (<b>middle column</b>) and the depth data with occlusion (<b>right column</b>). Each graph for radar data depicts 40 superimposed sequences of synthetic data, while each graph for depth data depicts 20 superimposed sequences of synthetic data.</p> Full article ">Figure 4
<p>Experimental setup (<b>a</b>) and the movement scenarios considered in the experiments: the serpentine trajectory used in EXP#1 (<b>b</b>), and the rectangle-shaped trajectory used in EXP#2 (<b>c</b>). The reference points, i.e., the points where marks have been placed on the floor, are indicated with the crosses.</p> Full article ">Figure 5
<p>Estimates of the trajectories obtained in experiment EXP#1, on the basis of the radar data (<b>a</b>), the depth data (<b>b</b>), and the fused data (<b>c</b>–<b>e</b>); the black dashed lines denote the reference trajectories.</p> Full article ">Figure 6
<p>Zoom on the empirical cumulative distribution functions characterising the position errors, obtained in EXP#1.</p> Full article ">Figure 7
<p>Estimates of the trajectories obtained in experiment EXP#2, on the basis of the radar data (<b>a</b>), the depth data (<b>b</b>), and the fused data (<b>c</b>–<b>e</b>); the black dashed lines denote the reference trajectories.</p> Full article ">Figure 7 Cont.
<p>Estimates of the trajectories obtained in experiment EXP#2, on the basis of the radar data (<b>a</b>), the depth data (<b>b</b>), and the fused data (<b>c</b>–<b>e</b>); the black dashed lines denote the reference trajectories.</p> Full article ">Figure 8
<p>Zoom on the empirical cumulative distribution functions characterising the position errors, obtained in EXP#2.</p> Full article ">Figure 9
<p>Estimates of the trajectories obtained in EXP#1, by means of the NARX method (<b>a</b>), and the WINFNS method described in [<a href="#B37-sensors-23-01457" class="html-bibr">37</a>] (<b>b</b>); for the NARX method, the value of the <math display="inline"><semantics> <mrow> <msub> <mi mathvariant="normal">A</mi> <mrow> <mi>ECDF</mi> </mrow> </msub> </mrow> </semantics></math> indicator is 0.85, while for the WINFNS method, it is 0.83.</p> Full article ">Figure 10
<p>Uncertainty indicators characterising the estimates of the numbers of turns, obtained in EXP#2.</p> Full article ">Figure 11
<p>Uncertainty indicators characterising the estimates of the travelled distance, obtained in EXP#2.</p> Full article ">Figure 12
<p>Uncertainty indicators characterising the estimates of the walking speed, obtained in EXP#2.</p> Full article ">
17 pages, 11586 KiB  
Article
A Comprehensive Landscape of Imaging Feature-Associated RNA Expression Profiles in Human Breast Tissue
by Tian Mou, Jianwen Liang, Trung Nghia Vu, Mu Tian and Yi Gao
Sensors 2023, 23(3), 1432; https://doi.org/10.3390/s23031432 - 28 Jan 2023
Cited by 3 | Viewed by 2287
Abstract
The expression abundance of transcripts in nondiseased breast tissue varies among individuals. The association study of genotypes and imaging phenotypes may help us to understand this individual variation. Since existing reports mainly focus on tumors or lesion areas, the heterogeneity of pathological image [...] Read more.
The expression abundance of transcripts in nondiseased breast tissue varies among individuals. The association study of genotypes and imaging phenotypes may help us to understand this individual variation. Since existing reports mainly focus on tumors or lesion areas, the heterogeneity of pathological image features and their correlations with RNA expression profiles for nondiseased tissue are not clear. The aim of this study is to discover the association between the nucleus features and the transcriptome-wide RNAs. We analyzed both microscopic histology images and RNA-sequencing data of 456 breast tissues from the Genotype-Tissue Expression (GTEx) project and constructed an automatic computational framework. We classified all samples into four clusters based on their nucleus morphological features and discovered feature-specific gene sets. The biological pathway analysis was performed on each gene set. The proposed framework evaluates the morphological characteristics of the cell nucleus quantitatively and identifies the associated genes. We found image features that capture population variation in breast tissue associated with RNA expressions, suggesting that the variation in expression pattern affects population variation in the morphological traits of breast tissue. This study provides a comprehensive transcriptome-wide view of imaging-feature-specific RNA expression for healthy breast tissue. Such a framework could also be used for understanding the connection between RNA expression and morphology in other tissues and organs. Pathway analysis indicated that the gene sets we identified were involved in specific biological processes, such as immune processes. Full article
(This article belongs to the Special Issue Sensing-Based Biomedical Communication and Intelligent Identification for Healthcare)
►▼ Show Figures

Figure 1

Figure 1
<p>Pipeline of the image processing. Two parts are consisted in our method: representative features are extracted by multilayer perceptron, and the clustering is performed by k-means. The nucleus features were firstly used to produce the pseudo label by k-means for training the MLP Network, and the representative features were used to produce the pseudo label again. The framework iterated 50 epochs until convergence.</p> Full article ">Figure 2
<p>Pipeline of the association analysis between image-based features and RNA-expression profiles of healthy breast tissue. A total of 456 H&amp;E images and the corresponding RNA-seq data from the GTEx database were included in the analysis. Sixty-five intensity and texture features of nuclei in glandular tissue were computed and then classified into four clusters. We discovered 1447 genes specific to single clusters, and the top 5 genes of each cluster are shown in the output panel. The circles in the boxplot represent outliers in the data.</p> Full article ">Figure 3
<p>One example of glandular tissue segmentation. (<b>A</b>) The standard image at its lowest resolution to show the global view. (<b>B</b>) The corresponding global mask image in which the white part represents the glandular tissue. (<b>C</b>,<b>D</b>) The corresponding zoomed-in version.</p> Full article ">Figure 4
<p>Two example results of three nucleus segmentation methods for (<b>A</b>) challenge data and (<b>B</b>) GTEx data. Contour colors: red (ground truth), green (algorithm).</p> Full article ">Figure 5
<p>Heat map of clustering results for nucleus features. Each row represents a sample and each column represents a nucleus feature. Feature scores in this heat map were normalized.</p> Full article ">Figure 6
<p>Illustration of images sampled from those with the most extreme image features in each cluster. Correspondingly, the top five feature-specific genes that are most highly expressed in each cluster are also shown.</p> Full article ">Figure 7
<p>The gene-level expression distribution of the <span class="html-italic">SCTR</span> gene across four clusters, where the value of the ordinate is log2(TPM+1). The circles in the boxplot represent outliers in the data.</p> Full article ">Figure 8
<p>Color-map of the top 15 feature-specific genes from each cluster. Each row represents a gene, and each column represents a sample. Red and green indicate a gene’s mRNA expression level above and below its median expression level across all samples, respectively. The genes in each cluster were ordered by <span class="html-italic">p</span>-value from bottom to top.</p> Full article ">Figure 9
<p>Visualization of enrichment analysis. (<b>A</b>) Bar graph of significant pathways in KEGG analysis. (<b>B</b>) Bar graph of significant pathways in GO analysis.</p> Full article ">
16 pages, 803 KiB  
Article
Identifying Biomarkers for Accurate Detection of Stress
by Kiran Jambhale, Smridhi Mahajan, Benjamin Rieland, Nilanjan Banerjee, Abhijit Dutt, Sai Praveen Kadiyala and Ramana Vinjamuri
Sensors 2022, 22(22), 8703; https://doi.org/10.3390/s22228703 - 11 Nov 2022
Cited by 5 | Viewed by 2261
Abstract
Substance use disorder (SUD) is a dangerous epidemic that develops out of recurrent use of alcohol and/or drugs and has the capability to severely damage one’s brain and behaviour. Stress is an established risk factor in SUD’s development of addiction and in reinstating [...] Read more.
Substance use disorder (SUD) is a dangerous epidemic that develops out of recurrent use of alcohol and/or drugs and has the capability to severely damage one’s brain and behaviour. Stress is an established risk factor in SUD’s development of addiction and in reinstating drug seeking. Despite this expanding epidemic and the potential for its grave consequences, there are limited options available for management and treatment, as well as pharmacotherapies and psychosocial treatments. To this end, there is a need for new and improved devices dedicated to the detection, management, and treatment of SUD. In this paper, the negative effects of SUD-related stress were discussed, and based on that, a few significant biomarkers were selected from a set of eight features collected by a chest-worn device, RespiBAN Professional, on fifteen individuals. We used three machine learning classifiers on these optimal biomarkers to detect stress. Based on the accuracies, the best biomarkers to detect stress and those considered as features for classification were determined to be electrodermal activity (EDA), body temperature, and a chest-worn accelerometer. Additionally, the differences between mental stress and physical stress, as well as different administrations of meditation during the study, were identified and analysed. Challenges, implications, and applications were also discussed. In the near future, we aim to replicate the proposed methods in individuals with SUD. Full article
(This article belongs to the Special Issue Sensing-Based Biomedical Communication and Intelligent Identification for Healthcare)
►▼ Show Figures

Figure 1

Figure 1
<p>RespiBAN Professional’s placement of electrodes. 1. RespiBAN Professional with temperature, EDA, and control module. 2. Three ECG electrodes. 3. Two EMG electrodes on the back where the shoulder meets the neck. Adapted from [<a href="#B19-sensors-22-08703" class="html-bibr">19</a>].</p> Full article ">Figure 2
<p>The two protocols tested under this study. The blue bars indicate the times when the study participants filled out questionnaires for self-report. Adapted from [<a href="#B18-sensors-22-08703" class="html-bibr">18</a>].</p> Full article ">Figure 3
<p>Significance of the features based on how frequently they were used in three different classifiers for best accuracies. EDA, temperature, and accelerometer Z (and Y) stand out as the important features.</p> Full article ">
13 pages, 5819 KiB  
Article
Control of Brushless Direct-Current Motors Using Bioelectric EMG Signals
by Sebastian Glowinski, Sebastian Pecolt, Andrzej Błażejewski and Bartłomiej Młyński
Sensors 2022, 22(18), 6829; https://doi.org/10.3390/s22186829 - 9 Sep 2022
Cited by 8 | Viewed by 2405
Abstract
(1) Background: The purpose of this study was to evaluate the analysis of measurements of bioelectric signals obtained from electromyographic sensors. A system that controls the speed and direction of rotation of a brushless DC motor (BLDC) was developed; (2) Methods: The system [...] Read more.
(1) Background: The purpose of this study was to evaluate the analysis of measurements of bioelectric signals obtained from electromyographic sensors. A system that controls the speed and direction of rotation of a brushless DC motor (BLDC) was developed; (2) Methods: The system was designed and constructed for the acquisition and processing of differential muscle signals. Basic information for the development of the EMG signal processing system was also provided. A controller system implementing the algorithm necessary to control the speed and direction of rotation of the drive rotor was proposed; (3) Results: Using two muscle groups (biceps brachii and triceps), it was possible to control the direction and speed of rotation of the drive unit. The control system changed the rotational speed of the brushless motor with a delay of about 0.5 s in relation to the registered EMG signal amplitude change; (4) Conclusions: The prepared system meets all the design assumptions. In addition, it is scalable and allows users to adjust the signal level. Our designed system can be implemented for rehabilitation, and in exoskeletons or prostheses. Full article
(This article belongs to the Special Issue Sensing-Based Biomedical Communication and Intelligent Identification for Healthcare)
►▼ Show Figures

Figure 1

Figure 1
<p>General scheme of the measurement system construction (<b>a</b>), diagram of the actuating system (<b>b</b>).</p> Full article ">Figure 2
<p>Design of printed circuit boards of data acquisition circuit (KiCad) (<b>a</b>), design of PCBs of BLDC motor controller (<b>b</b>), ready system of data acquisition without Bluetooth communication module (<b>c</b>), ready circuit of BLDC controller without connected Bluetooth communication module (<b>d</b>), suction-cup probes measuring and BLDC motor without integrated Hall sensors, (<b>e</b>) measuring card by National Instruments model NI USB-6211, (<b>f</b>) schematic diagram of the connection state or rest state of the entire system (<b>g</b>).</p> Full article ">Figure 3
<p>Scheme of connection of the instrumental amplifier circuit model AD620 (<b>a</b>), EMG biceps brachii signal amplified 495 times by the instrumental amplifier AD620ARZ (<b>b</b>).</p> Full article ">Figure 4
<p>Connection diagram of one of the LM324 operational amplifier circuits in the inverting amplifier configuration (<b>a</b>), comparison diagram of the signal with the gain G = 495 (blue) and <span class="html-italic">G</span> = −990 (red) (<b>b</b>).</p> Full article ">Figure 5
<p>Connection diagram of the second LM324 operational amplifier circuit in the configuration of the first-order high-pass filter (<b>a</b>), frequency-phase characteristic of the high-pass filter (<b>b</b>).</p> Full article ">Figure 6
<p>Diagram of a precise full-wave amplifier (<b>a</b>), equivalent circuit for simulation of a full-wave rectifier for condition V<sub>0</sub> &gt; 0 (<b>b</b>), equivalent circuit for simulation of a full-wave rectifier for the condition V<sub>0</sub> &lt; 0 (<b>c</b>).</p> Full article ">Figure 7
<p>Connection diagram of LM324 operational amplifier in configuration of the first-order low-pass filter (<b>a</b>), time and frequency diagram of signal after low-pass filtering (<b>b</b>).</p> Full article ">Figure 8
<p>Connection diagram of one of the LM324 operational amplifier circuits in an inverting amplifier configuration with adjustable gain (<b>a</b>), time and frequency diagram of the signal after low-pass filtering (<b>b</b>).</p> Full article ">Figure 9
<p>EMG signal as a function of time and frequency after amplification using instrument amplifier (<b>a</b>), EMG signal as a function of time and frequency after amplification using inverting amplifier (<b>b</b>).</p> Full article ">Figure 10
<p>EMG signal as a function of time and frequency after high-pass filtering (<b>a</b>), EMG signal as a function of time and frequency after full-wave rectification (<b>b</b>).</p> Full article ">Figure 11
<p>EMG signal as a function of time and frequency after low-pass filtering (<b>a</b>), EMG signal as a function of time after amplification using an inverting amplifier with adjustable gain (<b>b</b>).</p> Full article ">Figure 12
<p>Reaction of the engine to a given EMG control signal (color: black—biceps signal; blue—triceps signal; red—analog interpretation of engine; speed, green—first direction of rotation, blue—second direction of rotation) (<b>a</b>), normalized correlation between the processed triceps and biceps signals and motor reactions (<b>b</b>).</p> Full article ">
10 pages, 913 KiB  
Article
Intra-Rater Reliability of Shear Wave Elastography for the Quantification of Respiratory Muscles in Adolescent Athletes
by Małgorzata Pałac and Paweł Linek
Sensors 2022, 22(17), 6622; https://doi.org/10.3390/s22176622 - 1 Sep 2022
Cited by 4 | Viewed by 1956
Abstract
The aim of this study was to assess the intra-rater reliability and agreement of diaphragm and intercostal muscle elasticity and thickness during tidal breathing. The diaphragm and intercostal muscle parameters were measured using shear wave elastography in adolescent athletes. To calculate intra-rater reliability, [...] Read more.
The aim of this study was to assess the intra-rater reliability and agreement of diaphragm and intercostal muscle elasticity and thickness during tidal breathing. The diaphragm and intercostal muscle parameters were measured using shear wave elastography in adolescent athletes. To calculate intra-rater reliability, intraclass correlation coefficient (ICC) and Bland–Altman statistics were used. The reliability/agreement for one-day both muscle measurements (regardless of probe orientation) were at least moderate. During the seven-day interval between measurements, the reliability of a single measurement depended on the measured parameter, transducer orientation, respiratory phase, and muscle. Excellent reliability was found for diaphragm shear modulus at the peak of tidal expiration in transverse probe position (ICC3.1 = 0.91–0.96; ICC3.2 = 0.95), and from poor to excellent reliability for the intercostal muscle thickness at the peak of tidal inspiration with the longitudinal probe position (ICC3.1 = 0.26–0.95; ICC3.2 = 0.15). The overall reliability/agreement of the analysed data was higher for the diaphragm measurements (than the intercostal muscles) regardless of the respiratory phase and probe position. It is difficult to identify a more appropriate probe position to examine these muscles. The shear modulus/thickness of the diaphragm and intercostal muscles demonstrated good reliability/agreement so this appears to be a promising technique for their examination in athletes. Full article
(This article belongs to the Special Issue Sensing-Based Biomedical Communication and Intelligent Identification for Healthcare)
►▼ Show Figures

Figure 1

Figure 1
<p>Illustration showing the placement of the transversally (<b>A</b>) and parallel to the ribs (<b>B</b>) ultrasound probe position.</p> Full article ">Figure 2
<p>Data extraction from the images collected in SWE mode in transverse (<b>A</b>) and longitudinal (<b>B</b>) probe view. D—diaphragm; IC—intercostal muscle.</p> Full article ">

Review

Jump to: Editorial, Research

18 pages, 983 KiB  
Review
Intelligent Health: Progress and Benefit of Artificial Intelligence in Sensing-Based Monitoring and Disease Diagnosis
by Gabriela Palavicini
Sensors 2023, 23(22), 9053; https://doi.org/10.3390/s23229053 - 8 Nov 2023
Cited by 2 | Viewed by 3266
Abstract
Technology has progressed and allows people to go further in multiple fields related to social issues. Medicine cannot be the exception, especially nowadays, when the COVID-19 pandemic has accelerated the use of technology to continue living meaningfully, but mainly in giving consideration to [...] Read more.
Technology has progressed and allows people to go further in multiple fields related to social issues. Medicine cannot be the exception, especially nowadays, when the COVID-19 pandemic has accelerated the use of technology to continue living meaningfully, but mainly in giving consideration to people who remain confined at home with health issues. Our research question is: how can artificial intelligence (AI) translated into technological devices be used to identify health issues, improve people’s health, or prevent severe patient damage? Our work hypothesis is that technology has improved so much during the last decades that Medicine cannot remain apart from this progress. It must integrate technology into treatments so proper communication between intelligent devices and human bodies could better prevent health issues and even correct those already manifested. Consequently, we will answer: what has been the progress of Medicine using intelligent sensor-based devices? Which of those devices are the most used in medical practices? Which is the most benefited population, and what do physicians currently use this technology for? Could sensor-based monitoring and disease diagnosis represent a difference in how the medical praxis takes place nowadays, favouring prevention as opposed to healing? Full article
(This article belongs to the Special Issue Sensing-Based Biomedical Communication and Intelligent Identification for Healthcare)
►▼ Show Figures

Figure 1

Figure 1
<p>Schematic representation of a biosensor.</p> Full article ">Figure 2
<p>Different smart wearable devices and their cardiovascular applications.</p> Full article ">
21 pages, 1757 KiB  
Review
A Comprehensive Review on Machine Learning in Healthcare Industry: Classification, Restrictions, Opportunities and Challenges
by Qi An, Saifur Rahman, Jingwen Zhou and James Jin Kang
Sensors 2023, 23(9), 4178; https://doi.org/10.3390/s23094178 - 22 Apr 2023
Cited by 86 | Viewed by 16269
Abstract
Recently, various sophisticated methods, including machine learning and artificial intelligence, have been employed to examine health-related data. Medical professionals are acquiring enhanced diagnostic and treatment abilities by utilizing machine learning applications in the healthcare domain. Medical data have been used by many researchers [...] Read more.
Recently, various sophisticated methods, including machine learning and artificial intelligence, have been employed to examine health-related data. Medical professionals are acquiring enhanced diagnostic and treatment abilities by utilizing machine learning applications in the healthcare domain. Medical data have been used by many researchers to detect diseases and identify patterns. In the current literature, there are very few studies that address machine learning algorithms to improve healthcare data accuracy and efficiency. We examined the effectiveness of machine learning algorithms in improving time series healthcare metrics for heart rate data transmission (accuracy and efficiency). In this paper, we reviewed several machine learning algorithms in healthcare applications. After a comprehensive overview and investigation of supervised and unsupervised machine learning algorithms, we also demonstrated time series tasks based on past values (along with reviewing their feasibility for both small and large datasets). Full article
(This article belongs to the Special Issue Sensing-Based Biomedical Communication and Intelligent Identification for Healthcare)
►▼ Show Figures

Figure 1

Figure 1
<p>Concept of machine learning in healthcare area.</p> Full article ">Figure 2
<p>Types of machine learning such as supervised and unsupervised learning.</p> Full article ">Figure 3
<p>The general architecture of machine learning with requires steps such as data to feature extraction and training to prediction using different machine learning models.</p> Full article ">Figure 4
<p>Demonstration of decision trees with the root, decision, and Leaf nodes. Start from the root node, then move to the decision node using the leaf node information.</p> Full article ">Figure 5
<p>Demonstration of support vector machine. The solid red line indicates the separating hyperplane and the distance between two dotted lines is the maximum margin for separating different classes.</p> Full article ">Figure 6
<p>Demonstration of naïve Bayes with the distribution of different classes.</p> Full article ">Figure 7
<p>Demonstration of K-NN identifying unknown pattern by assigning a value to the K, where the nearest neighbor category of the K training sample is considered the same as the classification.</p> Full article ">Figure 8
<p>Demonstration of linear regression with best-fit line.</p> Full article ">Figure 9
<p>Demonstration of logistic regression with s-curve line.</p> Full article ">Figure 10
<p>Demonstration of ensemble methods which combine the different machine learning algorithms.</p> Full article ">Figure 11
<p>Demonstration of support vector regression. The solid black line indicates the separating hyperplane, and the distance between two dotted lines is the boundary line for separating different classes.</p> Full article ">Figure 12
<p>Demonstration of K-means algorithm by the partition of data points into k clusters by minimizing the sum of the squared distance between the point.</p> Full article ">Figure 13
<p>Demonstration of K-medoids through finding the most central object within the cluster and assigning the nearest object to the medoids to create a cluster as a reference point.</p> Full article ">Figure 14
<p>Demonstration of hierarchical clustering by analysing similarities of the characteristics in clusters.</p> Full article ">Figure 15
<p>Demonstration of fuzzy c-means by grouping the data into N clusters when clusters overlap.</p> Full article ">Figure 16
<p>Demonstration of Gaussian mixture model involves representing the probability density function as a blend of several Gaussian distributions, with each distribution corresponding to a cluster present in the data.</p> Full article ">Figure 17
<p>Demonstration of hidden Markov model for a sequence of hidden states over time.</p> Full article ">
Show export options expand_more Show export options expand_less
Select all
Export citation of selected articles as:
 
Displaying articles 1-12
Back to TopTop