Deep Learning-Based Electrocardiogram Signal Noise Detection and Screening Model / 대한의료정보학회지

OBJECTIVES: Biosignal data captured by patient monitoring systems could provide key evidence for detecting or predicting critical clinical events; however, noise in these data hinders their use. Because deep learning algorithms can extract features without human annotation, this study hypothesized that they could be used to screen unacceptable electrocardiograms (ECGs) that include noise. To test that, a deep learning-based model for unacceptable ECG screening was developed, and its screening results were compared with the interpretations of a medical expert. METHODS: To develop and apply the screening model, we used a biosignal database comprising 165,142,920 ECG II (10-second lead II electrocardiogram) data gathered between August 31, 2016 and September 30, 2018 from a trauma intensive-care unit. Then, 2,700 and 300 ECGs (ratio of 9:1) were reviewed by a medical expert and used for 9-fold cross-validation (training and validation) and test datasets. A convolutional neural network-based model for unacceptable ECG screening was developed based on the training and validation datasets. The model exhibiting the lowest cross-validation loss was subsequently selected as the final model. Its performance was evaluated through comparison with a test dataset. RESULTS: When the screening results of the proposed model were compared to the test dataset, the area under the receiver operating characteristic curve and the F1-score of the model were 0.93 and 0.80 (sensitivity = 0.88, specificity = 0.89, positive predictive value = 0.74, and negative predictive value = 0.96). CONCLUSIONS: The deep learning-based model developed in this study is capable of detecting and screening unacceptable ECGs efficiently.

Internet of Things Applied in Healthcare Based on Open Hardware with Low-Energy Consumption / 대한의료정보학회지

OBJECTIVES: The Internet of Things (IoT) and its applications are growing simultaneously. These applications need new intelligent devices along heterogeneous networking. Which makes them costly to implement indeed. Platforms and open devices designed for open-source hardware are possible solutions. This research was conducted under an IoT design, implementation, and assessment model for the remote monitoring of pulse oximetry via oxygen partial saturation (SpO2) and heart rate (HR) with low-energy consumption. METHODS: This study focused on the development of SpO2 and HR measurements that will allow the monitoring and estimation in real time of the user's state and health related to the established parameters. Measurements were acquired and recorded using a remote web server that recorded the acquired variables for further processing. The statistical analysis data allows comparison of the registered data measured with theoretical models. RESULTS: The IoT model was developed use Bluetooth low-energy devices, which comply with low-cost and open-hardware solutions operated via ‘HTTP requests’ for data transmission and reception from a cloud server to an edge device. Network performance assessment was conducted to guarantee the availability and integrity of the acquired values and signals. The system measured SpO2 and HR variables. The most significant result was to achieve energy consumption 20% lower than that of devices in the market. CONCLUSIONS: In summary, the acquired data validation based on the IoT model had a transmission error of 0.001% which proves its applicability in healthcare.

Design and preliminary validation of a ubiquitous and wearable physiological monitoring system / 生物医学工程学杂志

To achieve continuously physiological monitoring on hospital inpatients, a ubiquitous and wearable physiological monitoring system SensEcho was developed. The whole system consists of three parts: a wearable physiological monitoring unit, a wireless network and communication unit and a central monitoring system. The wearable physiological monitoring unit is an elastic shirt with respiratory inductive plethysmography sensor and textile electrocardiogram (ECG) electrodes embedded in, to collect physiological signals of ECG, respiration and posture/activity continuously and ubiquitously. The wireless network and communication unit is based on WiFi networking technology to transmit data from each physiological monitoring unit to the central monitoring system. A protocol of multiple data re-transmission and data integrity verification was implemented to reduce packet dropouts during the wireless communication. The central monitoring system displays data collected by the wearable system from each inpatient and monitors the status of each patient. An architecture of data server and algorithm server was established, supporting further data mining and analysis for big medical data. The performance of the whole system was validated. Three kinds of tests were conducted: validation of physiological monitoring algorithms, reliability of the monitoring system on volunteers, and reliability of data transmission. The results show that the whole system can achieve good performance in both physiological monitoring and wireless data transmission. The application of this system in clinical settings has the potential to establish a new model for individualized hospital inpatients monitoring, and provide more precision medicine to the patients with information derived from the continuously collected physiological parameters.

Estimation of Cardiac Index: Validation of the Mobil-O-Graph NG in Comparison with the FloTrac/Vigileo

OBJECTIVE: Although the reference value of cardiac index (CI) is derived by pulmonary arterial pressure, the use of pulmonary arterial catheterization is limited by low cost effectiveness and many concerns regarding complications. Therefore, relatively noninvasive indirect measurement is used widely perioperatively. The goal of this study was to determine the accuracy of the CI derived by Mobil-O-Graph NG (cCI) noninvasively in patients undergoing general anesthesia by comparing that measured by FloTrac/Vigileo (fCI), the minimal invasive method. METHODS: The Bland-Altman method was used to quantify agreement. Bias (mean difference between fCI-cCI) represents the systematic error between methods and precision (standard deviation of the bias) represents the random error or variability between techniques. The percentage error was considered clinically acceptable, and the tested method (Mobil-O-Graph NG) was regarded as interchangeable with the reference method (FloTrac/Vigileo), if it was below 30%. RESULTS: One hundred and ninety-five patients were included in this study, and CI, measured in the 121 patients. The Bland-Altman analysis revealed a bias −0.01 and the percentage error of 32.4%. And the difference is inversely increased according the mean CI. CONCLUSION: Results showed that CI measured by Mobil-O-Graph NG had a wide limit of agreement with that measured by FloTrac/Vigileo, therefore regarded as not interchangeable.

Clinical validation study of the SignCare Vital Signs Monitor of Fundación Cardiovascular de Colombia / Estudio de validación clínica del monitor de signos vitales SignCare de la Fundación Cardiovascular de Colombia

Abstract Introduction: In Colombia, due to the difficult access to health services and to geographic conditions, the implementation and innovation of telemedicine technological tools is a priority. Having a validated vital signs monitor (VSM) improves proper medical treatment and diagnosis. Objective: To design and perform clinical trials for the SignCare VSM. Materials and methods: A device for continuous monitoring of electrocardiography, respiration, oxygen saturation, temperature and noninvasive blood pressure (NIBP) was designed. This device was validated in a laboratory in order to ensure a robust prototype, close to the level of commercial medical devices. Clinical trials were performed through a cross -section study with 98 patients, whose vital signs were measured using the SignCare monitor and a commercial monitor. These two measurements were compared using Pearson's correlation coefficients. Results: There were no statistically significant differences between the results obtained with the SignCare VSM and the commercial monitor. The highest correlations were found for the following items: heart rate by electrocardiogram (r=0.844), heart rate by oxymetry (r=0.821), body temperature (r=0.895), systolic blood pressure (r=0.780), and diastolic blood pressure (r=0.811). Conclusions: The SignCare device is as reliable as the commercial monitor in the qualitative detection of morphologic alterations of electrocardiogram records, as well as in breathing, temperature, oxygen saturation and blood pressure parameters, which makes it recommendable for clinical use in adult population.

Resumen Introducción. Debido a las dificultades geográficas y de acceso a los servicios de salud en Colombia, la implementación e innovación con herramientas de telemedicina se convierte en un tema prioritario; contar con un monitor de signos vitales validado favorece el tratamiento médico oportuno. Objetivos Diseñar y realizar la validación clínica del monitor de signos vitales SignCare. Materiales y métodos. Se diseñó un equipo para el monitoreo constante de las señales de electrocardiografía, respiración, saturación de oxígeno, temperatura y presión arterial no invasiva. El dispositivo fue validado en el laboratorio para asegurar un prototipo robusto a nivel de dispositivos médicos comerciales. La validación clínica se hizo mediante un estudio de corte transversal en 98 pacientes a los que se les realizaron mediciones con el monitor SignCare y con un monitor comercial. Se compararon estas dos mediciones mediante coeficientes de correlación de Pearson. Resultados. No hubo diferencias estadísticamente significativas en cuanto a los resultados obtenidos con el monitor SignCare y con el monitor comercial. Las mayores correlaciones se presentaron en la frecuencia cardíaca por electrocardiograma (r=0.844), frecuencia cardíaca por oximetría (r=0.821), temperatura corporal (r=0.895), tensión arterial sistólica (r=0.780) y tensión arterial diastólica (0.811). Conclusiones. El monitor SignCare es tan confiable como el monitor comercial para la detección cualitativa de alteraciones morfológicas del registro electrocardiográfico, lo que hace posible su recomendación para uso clínico en población adulta.

Costo fisiológico de marcha confortable según el nivel socioeconómico en adultos mayores autovalentes de la comunidad en Talca, Chile / Psychological Cost of Comfortable Gait According to Socioeconomic Status in Self-Reliant Elderly in Talca, Chile / Custo fisiológico de caminhada confortável de acordo com a status socioeconômica em idosos autovalentes comunidade em Talca, Chile

Introducción: la funcionalidad en adultos mayores (AM) se encuentra condicionada por aspectos sociales, económicos y culturales; al respecto, la velocidad de marcha confortable (MC) es considerada su indicador más específico, lo que traduce un costo fisiológico (CF) asociado. En este escenario, el propósito de esta investigación es evaluar el CF de MC, según el nivel socioeconómico (NSE) en AM autovalentes. Materiales y métodos: participaron 75 AM autovalentes, los cuales fueron clasificados según el NSE en medio-bajo y medio-alto. Se solicitó a cada participante que caminara naturalmente durante 3 minutos en una pista de forma elíptica, se registró cada 15 segundos la distancia recorrida y la frecuencia cardiaca de trabajo, esto para el correspondiente cálculo del índice de costo fisiológico (ICF). Además, se estableció como punto crítico fisiológico (PCF) el aumento significativo del ICF desde el estado basal. Resultados: el ICF fue similar según NSE (p = 0,885), siendo mayor en mujeres (≈0,3 lat/m) que en hombres (≈0,2 lat/m; p < 0,001). La distancia recorrida fue alrededor de 200 metros en ambos NSE y géneros, siendo superior en el NSE medio-bajo para el grupo masculino frente al femenino (p = 0,009). En mujeres el PCF aparece 15 segundos antes en el NSE medio-alto, mientras que en hombres el ICF se mantiene estable durante toda la prueba. Conclusión: el NSE no influye en el CF de mc de AM autovalentes, no obstante, se confirma al género como principal factor en su traducción. Se recomienda la inclusión metodológica del PCf, dada su pertinencia para el análisis de la eficiencia de marcha.

Introduction: Elderly functionality is conditioned by social, economic and cultural aspects. The Comfortable Gait Speed (CGS) is considered the most specific indicator of functional capacity, reflecting an associated Physiological Cost (FC). Under this scenario, the purpose of this research is to evaluate the Comfortable Gait Physiologic Cost (CGPC) according to Socioeconomic Status (SES) in self-reliant elderly. Materials and methods: 75 self-reliant elderly were classified into medium-low and medium-high SES levels. Each participant was asked to walk naturally for three minutes on an elliptical track. Traveled distance and heart rate were recorded every 15 seconds for the Physio-logical Cost Index (PCI) calculation. Furthermore, the significant increase from baseline PCI was established as the Critical Physiological Point (CPP). Results: There was no difference in the PCI in relation to the SSE (p = 0.885). However, it was higher in women (≈0,3 lat / m) when compared to men (≈ 0.2 beats/m; p < 0.001). The traveled distance was about 200 meters in both SES and gender, showing higher in the medium-low SES for the male versus the female group (p = 0.009). In women the CPP appears 15 seconds earlier in the medium-high SES, while in men the pci remains stable throughout the test. Conclusions: The SES does not influence the CGPC in self-reliant elderly; however gender is confirmed as the main factor in its translation. Methodological CPP inclusion is recommended given its relevance to the analysis of gait efficiency.

Introdução: A capacidade funcional em idosos é condicionada por aspectos sociais, econômicos e culturais. A Velocidade de Caminhada Confortável (VCC) é considerada o indicador mais específico da capacidade funcional, traduzindo um Custo Fisiológico (CF) associado. Neste cenário, o objetivo deste estudo é avaliar o CF da VCC relacionada ao Status Socioeconômico (SSE) de idosos autovalentes. Materiais e métodos: Fizeram parte deste estudo 75 idosos independentes, os quais foram classificados de acordo com o SSE em média-baixa e média-alta. Cada participante foi motivado a caminhar naturalmente por 3 minutos em uma pista elíptica. Foram registrados a cada 15 segundos distância percorrida e da freqüência cardíaca, para o cálculo do Índice de Custo Fisiológico (ICF). Além disso, estabeleceu-se como um Ponto Crítico Fisiológico (PCF) em ICF aumento significativo a partir da linha de base. Resultados: Não houve diferença no ICF em relação ao SSE (p = 0,885), porém, o mesmo foi maior em mulheres (≈0,3 lat/m) comparado aos homens (≈0,2 lat/m; p < 0,001). A distância percorrida foi de cerca de 200 metros em ambos SSE e gêneros, sendo maior no SSE médio-baixo para o sexo masculino em relação o grupo feminino (p = 0,009). Nas mulheres, o PCF aparece 15 segundos antes no SSE médio-alto, nos homens o ICF permanece estável ao longo do teste. Conclusão: O SSE não influencia o CF da VCC de idosos independentes, no entanto, o gênero influência de forma importante na tradução do CF. É recomendada a inclusão metodológica do PCF devido à sua relevância para a análise da eficiência da marcha.

Multimodality Monitoring in the Neurointensive Care Unit: A Special Perspective for Patients with Stroke / 대한뇌졸중학회지

Multimodality monitoring (MMM) is a recently developed method that aids in understanding real-time brain physiology. Early detection of physiological disturbances is possible with the help of MMM, which allows identification of underlying causes of deterioration and minimization of secondary brain injury (SBI). MMM is especially helpful in comatose patients with severe brain injury because neurological examinations are not sensitive enough to detect SBI. The variables frequently examined in MMM are hemodynamic parameters such as intracranial pressure, cerebral perfusion pressure, and mean arterial pressure; brainspecific oxygen tension; markers for brain metabolism including glucose, lactate, and pyruvate levels in brain tissue; and cerebral blood flow. Continuous electroencephalography can be performed, if needed. The majority of SBIs stem from brain tissue hypoxia, brain ischemia, and seizures, which lead to a disturbance in brain oxygen levels, cerebral blood flow, and electrical discharges, all of which are easily detected by MMM. In this review, we discuss the clinical importance of physiological variables as well as the practical applicability of MMM in patients with stroke.

Measurement of nursing students' sleep using by an automated wireless sleep tracker and subjective sleep scales / 기본간호학회지

PURPOSE: This study was conducted to investigate sleep patterns in nursing students as measured by an automated wireless sleep tracker and self-reported sleep scales. METHODS: Data were collected from October, 2011 to May, 2012. ZQ (a symbolic number that measures the quality and quantity of sleep) was calculated from the automated wireless sleep tracker. Pittsburgh Sleep Quality Index-Korean version, Sleep scale A, and sleep satisfaction were used as self-report sleep scales. Female nursing college students (n=37) were recruited by convenience sampling. Before going to bed, they were asked to put on a head-band type sleeping sensor and to move it the next morning. Every student slept 2~4 nights with the sleep tracker. Data were collected for 86 (91.4%) of 94 nights' sleep. The students were also requested to fill in the PSQI-K, sleep scale A, and sleep satisfaction. RESULTS: The only statistically significant correlation was between ZQ and sleep satisfaction (r= .869, p=.026). The biggest obstacle in using this device was putting on the head-band type sensor. CONCLUSION: Validity of the wireless sleep tracker was verified in part. The usefulness of this device in nursing research and practice was identified. Further evaluation need to be conducted with various populations.

Wearable Physiologic Monitoring System in Health Promotion

With the development of body sensor technology, wearable health monitoring systems have been an emerging information technology in the 'ubiquitous health care' system. Wearable sensors enable long-term continuous physiological monitoring important in health promotion and management of many chronic disorders. In this paper, we present several wearable health monitoring systems developed by different countries and discuss emerging opportunities, as well as, existing challenges such as standardization, administration, validation, and discrepancy issues between technology and clinical implication.

Análise contínua de medidas de cateter de artéria pulmonar volumétrico, ecotransesofágico, variações da pressão arterial sistêmica e marcadores de hipoperfusão tissular no choque hemorrágico em suínos / Continuous analyses of pulmonary, volumetric artery catheter parameters, transesophageal echocardiography, pressure pulse variation, and biomarkers of tissue hypoperfusion during hemorrhagic shock in swine: experimental study in swines

INTRODUÇÃO: Diferentes parâmetros hemodinâmicos, incluindo os indicadores estáticos de pré-carga cardíaca como o índice de volume diastólico final ventrículo direito (IVDFVD) e parâmetros dinâmicos como a variação de pressão de pulso (VPP) têm sido usados na tomada de decisão para considerar o processo da expansão volêmica em pacientes em estado grave. O objetivo deste estudo foi comparar a reanimação por fluidos guiados tanto por VPP ou IVDFVD após choque hemorrágico induzido experimentalmente. MÉTODO: vinte e seis suínos anestesiados e ventilados mecanicamente foram alocados em três grupos: controle (Grupo I), VPP (Grupo II) e IVDFVD (Grupo III). Foi induzido choque hemorrágico por retirada de sangue até atingir a pressão arterial média de 40mmhg, que foi mantida por 60 minutos. Parâmetros foram medidos no tempo basal (B), no tempo do choque (Choque 0), sessenta minutos depois do choque (Choque 60), imediatamente depois da ressuscitação com hidroxietilamido 6% (130/0. 4) (R0), uma hora (R60) e duas horas (R120) depois ressuscitação. Os pontos de avaliação da reanimação por fluidos foram determinados pelo retorno aos valores basais iniciais de VPP e IVDFVD. A análise estatística dos dados foi baseada em ANOVA para medidas repetidas seguidos pelo teste de Bonferroni (P<0.05%). RESULTADOS: O volume e tempo para ressuscitação foram maiores no grupo III do que no grupo II (Grupo III = 1305±331ml e Grupo II = 965±245ml; p<0.05 e Grupo III = 24.8± 4.7min e Grupo II = 8.8 ± 1.3 min, p<0.01, respectivamente). Todos os parâmetros estáticos e dinâmicos, bem como os biomarcadores de oxigenação tecidual foram afetados pelo choque hemorrágico e quase todos os parâmetros foram totalmente restaurados após a reanimação em ambos os grupos. CONCLUSÃO: Neste estudo em modelo de choque hemorrágico, a reanimação guiada pelo VPP utilizou menor quantidade de fluido e menor quantidade de tempo do que quando guiado por IVDFVD derivado de cateter de artéria pulmonar.

INTRODUCTION: Different hemodynamic parameters, including static indicators of cardiac preload as right ventricular end-diastolic volume index (RVEDVI) and dynamic parameters as pulse pressure variation (PPV) have been used in the decision-making process regarding volume expansion in critically ill patients. The objective of this study was to compare fluid resuscitation guided by either PPV or RVEDVI after experimentally-induced hemorrhagic shock. METHODS: 26 anesthetized and mechanically ventilated pigs were allocated into control (Group-I), PPV (Group-II) and RVEDVI (Group- III). Hemorrhagic shock was induced by blood withdrawal to target mean arterial pressure of 40mmHg, maintained for 60 minutes. Parameters were measured at baseline, time of shock, sixty minutes after shock, immediately after resuscitation with hydroxyethyl starch 6% (130/0.4), one hour and two hours thereafter. The endpoint of fluid resuscitation was determined as the baseline values of PPV and RVEDVI. Statistical analysis of data was based on ANOVA for repeated measures followed by the Bonferroni test (P<0.05). RESULTS: Volume and time to resuscitation were higher in Group-III than in Group-II (Group-III = 1305±331ml and Group-II = 965±245ml; p<0.05 and Group-IIII = 24.8±4.7min and Group-II = 8.8±1.3 min, p<0.05, respectively). All static and dynamic parameters and biomarkers of tissue oxygenation were affected by hemorrhagic shock and nearly all parameters were restored after resuscitation in both groups. CONCLUSION: In the proposed model of hemorrhagic shock, resuscitation to the established endpoints was achieved within a smaller amount of time and with less volume when guided by PPV than when guided by pulmonary artery catheter-derived RVEDVI.

Validation of a Monitoring System for CPR Quality in a Manikin Model

PURPOSE: The 2005 resuscitation guidelines stipulate the need for monitoring CPR (cardiopulmonary resuscitation) quality. Recently, several clinical investigations have shown that a real time monitoring and feedback system is effective for improving the quality of chest compressions during resuscitation. However little data exists regarding the accuracy of the monitoring system using an accelerometer sensor and a pressure sensor for the measuring of compression rate and depth. Our goal for this study was to investigate how well chest compression rate and depth can be estimated using the monitoring system. METHODS: Thirty seconds of continuous chest compressions were delivered on a standard skillmeter manikin lying on the floor with the monitoring system. The chest compressions were delivered with variations in compression rate (67~142 /min) and with variations in compression depth (22~61 mm). A total of 120 sets of compressions were delivered for validation of rate and depth. RESULTS: The correlation coefficient for compression rate between the monitoring system and the standard method was 0.999 (p<0.001), and Bland-Altman analysis showed a mean bias of -0.10+/-0.77/min, with limits of agreement ranging from -1.60 to 1.40 /min. The correlation coefficient for compression depth between two methods was 0.983 (p<0.001), and Bland-Altman analysis showed a mean bias of 4.2+/-2.0 mm, with limits of agreement ranging from 0.24 to 8.10 mm. CONCLUSION: Compared with a skillmeter manikin, a monitoring system for the quality of CPR estimates chest compression rate precisely, but overestimates chest compression depth by an average of 10.3%.

Analysis of Research Trends in Physiological Variables in Complementary and Alterative Therapy(CAT) in Korean Nursing / 기본간호학회지

PURPOSE: The purpose of this study was to analyze the trends in physiological variables in CAT research in nursing in Korea. METHOD: Of studies published in Korea between January, 2000 and December, 2005, 227 studies were analyzed according to the criteria of type, theme, and patterns in physiological variables related to CAT. RESULTS: There were 72 articles on CAT research in which physiological variables were examined. The most frequently researched type of CAT was massage and in particular, foot massage. The most frequently used physiological variables in CAT research were blood pressure, pulse, and body temperature. Patients with high blood pressure were the most frequent subjects for CAT research. As to the effect of physiological variable by CAT type, foot massage showed a decline in blood pressure in all six research studies involving patients with high blood pressure. CONCLUSION: There is a need to describe accurately the mechanism by which CAT affects physiological variables. There is also a need for repetitive analysis to verify the effect, and meta-analysis for the effect on physiological variables according to type of CAT.