scholarly journals Detection of COVID-19 in smartphone-based breathing recordings: A pre-screening deep learning tool

PLoS ONE ◽  
2022 ◽  
Vol 17 (1) ◽  
pp. e0262448
Author(s):  
Mohanad Alkhodari ◽  
Ahsan H. Khandoker
Keyword(s):  
Deep Learning ◽  
Healthy Subjects ◽  
Short Term Memory ◽  
P Value ◽  
Mel Frequency Cepstral Coefficients ◽  
Learning Framework ◽  
Significant Difference ◽  
Polymerase Chain ◽  
The Right ◽  
Asymptomatic Subjects

This study was sought to investigate the feasibility of using smartphone-based breathing sounds within a deep learning framework to discriminate between COVID-19, including asymptomatic, and healthy subjects. A total of 480 breathing sounds (240 shallow and 240 deep) were obtained from a publicly available database named Coswara. These sounds were recorded by 120 COVID-19 and 120 healthy subjects via a smartphone microphone through a website application. A deep learning framework was proposed herein that relies on hand-crafted features extracted from the original recordings and from the mel-frequency cepstral coefficients (MFCC) as well as deep-activated features learned by a combination of convolutional neural network and bi-directional long short-term memory units (CNN-BiLSTM). The statistical analysis of patient profiles has shown a significant difference (p-value: 0.041) for ischemic heart disease between COVID-19 and healthy subjects. The Analysis of the normal distribution of the combined MFCC values showed that COVID-19 subjects tended to have a distribution that is skewed more towards the right side of the zero mean (shallow: 0.59±1.74, deep: 0.65±4.35, p-value: <0.001). In addition, the proposed deep learning approach had an overall discrimination accuracy of 94.58% and 92.08% using shallow and deep recordings, respectively. Furthermore, it detected COVID-19 subjects successfully with a maximum sensitivity of 94.21%, specificity of 94.96%, and area under the receiver operating characteristic (AUROC) curves of 0.90. Among the 120 COVID-19 participants, asymptomatic subjects (18 subjects) were successfully detected with 100.00% accuracy using shallow recordings and 88.89% using deep recordings. This study paves the way towards utilizing smartphone-based breathing sounds for the purpose of COVID-19 detection. The observations found in this study were promising to suggest deep learning and smartphone-based breathing sounds as an effective pre-screening tool for COVID-19 alongside the current reverse-transcription polymerase chain reaction (RT-PCR) assay. It can be considered as an early, rapid, easily distributed, time-efficient, and almost no-cost diagnosis technique complying with social distancing restrictions during COVID-19 pandemic.

scholarly journals Detection of COVID-19 in smartphone-based breathing recordings using CNN-BiLSTM: a pre-screening deep learning tool

2021 ◽  
Author(s):  
Mohanad Alkhodari ◽  
Ahsan H. Khandoker
Keyword(s):  
Deep Learning ◽  
Healthy Subjects ◽  
Short Term Memory ◽  
Mel Frequency Cepstral Coefficients ◽  
Learning Framework ◽  
Long Short Term Memory ◽  
Polymerase Chain ◽  
The Right ◽  
Asymptomatic Subjects ◽  
Diagnosis Technique

AbstractThis study was sought to investigate the feasibility of using smartphone-based breathing sounds within a deep learning framework to discriminate between COVID-19, including asymptomatic, and healthy subjects. A total of 480 breathing sounds (240 shallow and 240 deep) were obtained from a publicly available database named Coswara. These sounds were recorded by 120 COVID-19 and 120 healthy subjects via a smartphone microphone through a website application. A deep learning framework was proposed herein the relies on hand-crafted features extracted from the original recordings and from the mel-frequency cepstral coefficients (MFCC) as well as deep-activated features learned by a combination of convolutional neural network and bi-directional long short-term memory units (CNN-BiLSTM). Analysis of the normal distribution of the combined MFCC values showed that COVID-19 subjects tended to have a distribution that is skewed more towards the right side of the zero mean (shallow: 0.59±1.74, deep: 0.65±4.35). In addition, the proposed deep learning approach had an overall discrimination accuracy of 94.58% and 92.08% using shallow and deep recordings, respectively. Furthermore, it detected COVID-19 subjects successfully with a maximum sensitivity of 94.21%, specificity of 94.96%, and area under the receiver operating characteristic (AUROC) curves of 0.90. Among the 120 COVID-19 participants, asymptomatic subjects (18 subjects) were successfully detected with 100.00% accuracy using shallow recordings and 88.89% using deep recordings. This study paves the way towards utilizing smartphone-based breathing sounds for the purpose of COVID-19 detection. The observations found in this study were promising to suggest deep learning and smartphone-based breathing sounds as an effective pre-screening tool for COVID-19 alongside the current reverse-transcription polymerase chain reaction (RT-PCR) assay. It can be considered as an early, rapid, easily distributed, time-efficient, and almost no-cost diagnosis technique complying with social distancing restrictions during COVID-19 pandemic.

Over-expression of IL-6 coding gene in the peripheral blood of migraine with aura patients

2021 ◽  
pp. 1-5
Author(s):  
Mahdi Ramezani ◽  
Alireza Komaki ◽  
Mohammad Mahdi Eftekharian ◽  
Mehrdokht Mazdeh ◽  
Soudeh Ghafouri-Fard
Keyword(s):  
Migraine With Aura ◽  
Peripheral Blood ◽  
Healthy Subjects ◽  
Migraine Without Aura ◽  
P Value ◽  
Healthy Controls ◽  
Over Expression ◽  
Significant Difference ◽  
Male Patients ◽  
Years Lived With Disability

Migraine is a common disorder which is placed among the top ten reasons of years lived with disability. Cytokines are among the molecules that contribute in the pathophysiology of migraine. In the current study, we evaluated expression levels of IL-6 coding gene in the peripheral blood of 120 migraine patients (54 migraine without aura and 66 migraine with aura patients) and 40 healthy subjects. No significant difference was detected in expression of IL-6 between total migraine patients and healthy controls (Posterior beta = 0.253, P value = 0.199). The interaction effect between gender and group was significant (Posterior beta =-1.274, P value = 0.011), therefore, we conducted subgroup analysis within gender group. Such analysis revealed that while expression of this gene is not different between male patients and male controls (Posterior beta =-0.371, P value > 0.999), it was significantly over-expressed in female patients compared with female controls (Posterior beta = 0.86, P= 0.002). Expression of IL-6 was significantly higher in patients with aura compared with controls (Posterior beta = 0.63, adjusted P value = 0.019). However, expression of this cytokine coding gene was not different between patients without aura and healthy subjects (Posterior beta = 0.193, adjusted P value = 0.281). Therefore, IL-6 might be involved in the pathophysiology of migraine among females and migraine with aura among both sexes.

Surrounding Vehicles’ Contribution to Car-Following Models: Deep-Learning-Based Analysis

2021 ◽  
pp. 036119812110186
Author(s):  
Saeed Vasebi ◽  
Yeganeh M. Hayeri ◽  
Peter J. Jin
Keyword(s):  
Deep Learning ◽  
Traffic Flow ◽  
Short Term Memory ◽  
Data Availability ◽  
Car Following ◽  
Preceding Vehicle ◽  
Long Short Term Memory ◽  
The Right ◽  
Mean Square Errors ◽  
Deep Learning Model

Relatively recent increased computational power and extensive traffic data availability have provided a unique opportunity to re-investigate drivers’ car-following (CF) behavior. Classic CF models assume drivers’ behavior is only influenced by their preceding vehicle. Recent studies have indicated that considering surrounding vehicles’ information (e.g., multiple preceding vehicles) could affect CF models’ performance. An in-depth investigation of surrounding vehicles’ contribution to CF modeling performance has not been reported in the literature. This study uses a deep-learning model with long short-term memory (LSTM) to investigate to what extent considering surrounding vehicles could improve CF models’ performance. This investigation helps to select the right inputs for traffic flow modeling. Five CF models are compared in this study (i.e., classic, multi-anticipative, adjacent-lanes, following-vehicle, and all-surrounding-vehicles CF models). Performance of the CF models is compared in relation to accuracy, stability, and smoothness of traffic flow. The CF models are trained, validated, and tested by a large publicly available dataset. The average mean square errors (MSEs) for the classic, multi-anticipative, adjacent-lanes, following-vehicle, and all-surrounding-vehicles CF models are 1.58 × 10−3, 1.54 × 10−3, 1.56 × 10−3, 1.61 × 10−3, and 1.73 × 10−3, respectively. However, the results show insignificant performance differences between the classic CF model and multi-anticipative model or adjacent-lanes model in relation to accuracy, stability, or smoothness. The following-vehicle CF model shows similar performance to the multi-anticipative model. The all-surrounding-vehicles CF model has underperformed all the other models.

Comparison of Bone-Conducted Vibration for Eliciting Ocular Vestibular-Evoked Myogenic Potentials

2011 ◽  
Vol 146 (2) ◽  
pp. 289-294 ◽  
Author(s):  
Chia-Chen Tseng ◽  
Shou-Jen Wang ◽  
Yi-Ho Young
Keyword(s):  
Prospective Study ◽  
Study Design ◽  
Healthy Subjects ◽  
University Hospital ◽  
Vestibular Evoked Myogenic Potentials ◽  
Ménière’S Disease ◽  
Asymmetry Ratio ◽  
Significant Difference ◽  
The Asymmetry ◽  
The Right

Objective. This study compared bone-conducted vibration (BCV) stimuli at forehead (Fz) and mastoid sites for eliciting ocular vestibular-evoked myogenic potentials (oVEMPs). Study Design. Prospective study. Setting. University hospital. Methods. Twenty healthy subjects underwent oVEMP testing via BCV stimuli at Fz and mastoid sites. Another 50 patients with unilateral Meniere’s disease also underwent oVEMP testing. Results. All healthy subjects showed clear oVEMPs via BCV stimulation regardless of the tapping sites. The right oVEMPs stimulated by tapping at the right mastoid had earlier nI and pI latencies and a larger nI-pI amplitude compared with those stimulated by tapping at the Fz and left mastoid. Similar trends were also observed in left oVEMPs. However, the asymmetry ratio did not differ significantly between the ipsilateral mastoid and Fz sites. Clinically, tapping at the Fz revealed absent oVEMPs in 28% of Meniere’s ears, which decreased to 16% when tapping at the ipsilesional (hydropic) mastoid site, exhibiting a significant difference. Conclusion. Tapping at the ipsilateral mastoid site elicits earlier oVEMP latencies and larger oVEMP amplitudes when compared with tapping at the Fz site. Thus, tapping at the Fz site is suggested to screen for the otolithic function, whereas tapping at the ipsilesional mastoid site is suitable for evaluating residual otolithic function.

scholarly journals Pediatric triage variations among nurses, pediatric and emergency residents using the Canadian triage and acuity scale

2021 ◽  
Vol 21 (1) ◽  
Author(s):  
Saleh Alshaibi ◽  
Tala AlBassri ◽  
Suliman AlQeuflie ◽  
Winnie Philip ◽  
Nesrin Alharthy
Keyword(s):  
Clinical Signs ◽  
Sampling Technique ◽  
Weighted Kappa ◽  
P Value ◽  
Kappa Statistics ◽  
Cross Sectional ◽  
Pediatric Residents ◽  
Significant Difference ◽  
The Right ◽  
Triage Level

Abstract Background Emergency care continues to be a challenge, since patients’ arrival is unscheduled and could occur at the same time which may fill the Emergency Department with non-urgent patients. Triaging is an integral part of every busy ED. The Canadian Triage and Acuity Scale (CTAS) is considered an accurate tool to be used outside Canada. This study aims to identify the chosen triage level and compare the variation between registered nurses, pediatric and adult emergency residents by using CTAS cases. Method This study was conducted at King Abdulaziz Medical City,Saudi Arabia. A cross-sectional self-administered questionnaire was used, and which contains 15 case scenarios with different triage levels. All cases were adopted from a Canadian triage course after receiving permission. Each case provides the patient’s symptoms, clinical signs and mode of arrival to the ED. The participants were instructed to assign a triage level using the following scale. A non-random sampling technique was used for this study. The rates of agreement between residents were calculated using kappa statistics (weighted-kappa) (95%CI). Result A total of 151 participants completed the study questionnaire which include 15 case scenarios. 73 were nurses and 78 were residents. The results showed 51.3, 56.6, and 59.9% mis-triaged the cases among the nurses, emergency residents, and pediatric residents respectively. Triage scores were compared using the Kruskal Wallis test and were statistically significant with a p value of 0.006. The mean ranks for nurses, emergency residents and pediatric residents were 86.41, 73.6 and 59.96, respectively. The Kruskal Wallis Post-Hoc test was performed to see which groups were statistically significant, and it was found that there was a significant difference between nurses and pediatrics residents (P value = 0.005). Moreover, there were no significant differences found between nurses and ER residents (P value> 0.05). Conclusion The triaging system was found to be a very important tool to prioritize patients based on their complaints. The results showed that nurses had the greatest experience in implementing patients on the right triage level. On the other hand, ER and pediatric residents need to develop more knowledge about CTAS and become exposed more to the triaging system during their training.

scholarly journals Signal Analysis For Cardiac Electrical Activation Studies

2021 ◽  
Author(s):  
Elnaz Shokrollahi
Keyword(s):  
Transfer Function ◽  
Electrical Activation ◽  
P Value ◽  
Canine Heart ◽  
Linear Discriminant ◽  
Noncontact Mapping ◽  
Mapping System ◽  
Significant Difference ◽  
The Right ◽  
Unipolar Electrograms

The aim of this study is to determine if some of the characteristics of reconstructed unipolar electrograms from the noncontact mapping system can be used to detect epicardial and to differentiate it from endocardial electrical activation in a canine heart. This would help the electrophysiologist know where exactly the origin of ventricular tachycardia or the critical point in tissue is located. Following this, arrhythmia can be successfully treated by ablating that part of the tissue of the heart. Virtual electrograms were recorded while pacing the right ventricle of an open-chest dog at multiple endocardial and epicardial sites using the commercially available noncontact mapping system (EnSite Array™ Catheter 3000). The endocardial and epicardial paced virtual electrograms from the juxtaposing sites allow for analyzing systematically the differences in their morphologies. Maximal dV / dt, area under the depolarization curve and latency extracted from unipolar electrograms demonstrated significant difference between epicardial and endocardial pacing sites with a p-value of less than 0.01 in all three cases. The above features were fed to a linear discriminant analysis based classifier and high classification accuracy was achieved. Therefore, reliable criteria can be proposed to allow for discrimination of an endocardial versus epicardial origin of electrical activation. And also the endocardial and epicardial paced virtual electrograms from the juxtaposing sites allows for an estimate of the transfer function of the myocardium in different positions of the right ventricles of a canine heart. The transfer function estimation will aid in better mathematical modeling of myocardium and could be a sensitive measure of myocardial homogeneity and arrhythmic foci localization.Another study was done on a human heart. This study was to evaluate the ability of virtual electrograms to predict abnormal bipolar electrograms. We tested the hypothesis of maxdV/dt, filtering and optimized DSM threshold. This allows better identification of abnormal myocardial substrate traditionally defined by contact bipolar mapping in human RVOT.

scholarly journals Short-Term Memory Comparison Of Students Of Faculty Of Medicine Pelita Harapan University Batch 2015 Between The Handwriting And Typing Method

Medicinus ◽  
2020 ◽  
Vol 7 (4) ◽  
pp. 108
Author(s):  
Caroline Widjaja ◽  
Stefanus Satria Sumali
Keyword(s):  
Experimental Study ◽  
Short Term Memory ◽  
P Value ◽  
Short Term ◽  
Typing Method ◽  
Term Memory ◽  
New Information ◽  
Study Population ◽  
The Right ◽  
Experimental Study Design

<p>Introduction : A lot of research has been done to determine if handwriting or typing note influenced short-term memory, however, the results obtained are still controversial. Therefore this study is structured to see the effect of note taking methods by handwriting and typing on short-term memory.</p><p>Aim : The aims of this study were to increase the performance of students in Faculty of Medicine Pelita Harapan University as well as providing the right and effective method of taking notes.<strong></strong></p><p>Method : Experimental study design was chosen in this study. Study population is students of faculty of medicine Pelita Harapan University batch 2015. 40 samples will be divide randomly into two, one group will take a note by handwriting and another by typing. Each group is required to watch a video about 15 minutes long.  The results were analyzed statistically using T-test.</p><p>Result : The average of  new information that can be remembered by group that take a note by handwriting significantly (p&lt;0,05) higher than group than take a note by typing with a p-value of 0,009.</p><p>Conclusion : Take a note by handwriting allows people to remember more new information than typing.</p>

scholarly journals ASSOCIATION BETWEEN MATRIX METALLOPROTEINASE-1 −1607 1G/2G POLYMORPHISM AND CHRONIC PERIODONTITIS IN INDONESIAN POPULATION

2020 ◽  
pp. 15-17
Author(s):  
RAHMATUL HAYATI ◽  
ANTONIUS WINOTO SUHARTONO ◽  
SRI LELYATI MASULILI ◽  
CHRISTOPHER TALBOT ◽  
ELZA IBRAHIM AUERKARI
Keyword(s):  
Matrix Metalloproteinase ◽  
Chronic Periodontitis ◽  
Healthy Subjects ◽  
Genomic Dna ◽  
Chain Reaction ◽  
Chi Square ◽  
Indonesian Population ◽  
Matrix Metalloproteinase 1 ◽  
Significant Difference ◽  
Polymerase Chain

Objective: This study aimed to identify the distribution of matrix metalloproteinase (MMP)-1 −1607 1G/2G alleles and genotypes in subjects withchronic periodontitis and healthy subjects in a sample of Indonesian population and assess the possible association of this polymorphism withsusceptibility to chronic periodontitis.Methods: Genomic DNA samples were obtained from 200 Indonesian males aged 33–78 years old, comprising 100 chronic periodontitis patientsand 100 healthy controls. DNA fragments were amplified by a polymerase chain reaction and analyzed by restriction fragment length polymorphism.Results were analyzed by Chi-square test.Results: The frequency of the 2G allele was high both in subjects with periodontitis (87%) and in controls (91%). Analysis of MMP-1 genotype(−1607 1G/2G) showed no significant difference between the chronic periodontitis and healthy groups (p>0.05).Conclusion: The result found no association between MMP-1 −1607 1G/2G polymorphism and susceptibility to chronic periodontitis in Indonesiansubjects.

scholarly journals Incidental finding of bronchial diverticula in a non-smoker population: evaluation on thin-section CT

2016 ◽  
Vol 81 (1-2) ◽  
Author(s):  
Federica Ciccarese ◽  
Giorgio Garzillo ◽  
Anna Maria Chiesa ◽  
Antonio Poerio ◽  
Serena Baroncini ◽  
...  
Keyword(s):  
Thin Section ◽  
Healthy Subjects ◽  
Respiratory Symptoms ◽  
Incidental Finding ◽  
Radiological Finding ◽  
Chest Ct ◽  
Lobe Bronchus ◽  
Significant Difference ◽  
The Right

<p>Bronchial diverticula have been described as a common radiological finding in smoker patients with COPD, but the specificity of this sign should be further investigated. Thus, the aim of our study was to evaluate the prevalence of diverticula in a cohort of non-smoker subjects. Between February and July 2012, 2438 patients were admitted to our Radiology Unit to undergo a chest CT. Among them, we enrolled 121 non-smoking patients (78/121-64.5% females, 43/121-35.5% males), of different age (57.0±20.7 years-range: 12-88), without any respiratory symptoms, submitted to chest CT for several reasons (oncologic evaluation: 59/121-48.8%; follow up of lung nodules: 27/121-22.3%; screening in connectivitis: 12/121-9.9%; others: 23/121-19.0%). We considered thin-section CT scan on axial, coronal and sagittal plans to evaluate prevalence, numbers and level of bronchial diverticula. Diverticula were found in 41/121-33.9% patients, with a slight major prevalence in males (p=0.048), but no significant difference on age. In 31/41-75.6% the number was &lt;3, whereof 17/31-54.8% with just one diverticulum assessed. Regarding the level, in 30/41-73.2% they were subcarinal, but they were also detected in mainstem (2/41-4.9%) and lobar bronchi (with the right upper lobe bronchus most frequently involved- 12/41-29.3%). Bronchial diverticula can be observed in non-smokers, as well as in smoker patients with COPD. However, their prevalence seems to be lower than in smokers and they tend to be isolated and subcarinal. The age of patients does not influence their finding. More studies should be proposed to better define a cut-off between smokers and healthy subjects.</p>

scholarly journals Foot Strengthening Exercise on Postural Balance and Functional Ability of Foot on Children 6-10 Years Old with Flexible Flatfoot

2020 ◽  
Vol 2 (3) ◽  
pp. 99-107
Author(s):  
Zahra Sativani ◽  
Riza Pahlawi
Keyword(s):  
Functional Ability ◽  
Postural Balance ◽  
Control Group ◽  
P Value ◽  
Strengthening Exercise ◽  
Flexible Flatfoot ◽  
Significant Difference ◽  
Post Test ◽  
The Right ◽  
And Control

The activities of children more involve the foot. One of the common problems in the foot is flexible flatfoot. A disturbance in the process of the formation of the arch foot could result in a deformation of the foot and increases the risk of an injury due to postural balance change. Normally, the arch of the foot formed the first five years for the age range of 2-6 years. The right choices of the intrinsic muscle exercises of the foot can prevent deformation and improve postural balance. This study aimed to discuss the effectiveness of foot strengthening exercise to improving postural balance and functional ability of foot on a flexible flatfoot 6-10 years old. This study was pre-experimental research with two groups of pre-post test design. The subjects of this research were 30 students that had been divided into two groups, case, and control. Each group consists of 15 students selected used purposive sampling method based on the criteria of inclusion that had been set. There was a significant difference after foot strengthening exercise between the case and control group, p-value = 0,000. The foot strengthening exercise could improve the postural balance and functional ability of the foot on a flexible flatfoot 6-10 years old.

Sign in / Sign up

Export Citation Format

Share Document