• Journal of Biomedical Engineering Research
• /
• v.35 no.6
• /
• pp.211-218
• /
• 2014

The objective of this research is to develop an automatic algorithm based on electrocardiogram (ECG) to estimate slow-wave sleep (SWS). An algorithm is based on 7 indices extracted from heart rate on ECG which simultaneously recorded with standard full night polysomnography from 31 subjects. Those 7 indices were then applied to independent component analysis to extract a feature that discriminates SWS and other sleep stages. Overall Cohen's kappa, accuracy, sensitivity and specificity of the algorithm to detect 30s epochs of SWS were 0.52, 0.87, 0.70 and 0.90, respectively. The automatic SWS detection algorithm could be useful combining with existing REM and wake estimation technique on unattended home-based sleep monitoring.

• Journal of Electrical Engineering and Technology
• /
• v.13 no.5
• /
• pp.2080-2085
• /
• 2018

There have been numerous studies that extract the R-peak from electrocardiogram (ECG) signals. All of these studies can extract R-peak from ECG. However, these methods are complicated and difficult to implement in a real-time portable ECG device. After filtration choosing a threshold value for R-peak detection is a big challenge. Fixed threshold scheme is sometimes unable to detect low R-peak value and adaptive threshold sometime detect wrong R-peak for more adaptation. In this paper, a simple and robustness algorithm is proposed to detect R-peak with less complexity. This method also solves the problem of threshold value selection. Using the adaptive filter, the baseline drift can be removed from ECG signal. After filtration, an appropriate threshold value is automatically chosen by using the minimum and maximum value of an ECG signals. Then the neighborhood searching scheme is applied under threshold value to detect R-peak from ECG signals. Proposed method improves the detection and accuracy rate of R-peak detection. After R-peak detection, we calculate heart rate to know the heart condition.

• Journal of Biomedical Engineering Research
• /
• v.40 no.2
• /
• pp.62-67
• /
• 2019

The purpose of this study was to evaluate the performance of deep neural network model in order to determine whether there is a risk factor for coronary artery disease based on the cardiac variation parameter. The study used unidentifiable 297 data to evaluate the performance of the model. Input data consists of heart rate parameters, which are SDNN (standard deviation of the N-N intervals), PSI (physical stress index), TP (total power), VLF (very low frequency), LF (low frequency), HF (high frequency), RMSSD (root mean square of successive difference) APEN (approximate entropy) and SRD (successive R-R interval difference), the age group and sex. Output data are divided into normal and patient groups, and the patient group consists of those diagnosed with diabetes, high blood pressure, and hyperlipidemia among the various risk factors that can cause coronary artery disease. Based on this, a binary classification model was applied using Deep Neural Network of deep learning techniques to classify normal and patient groups efficiently. To evaluate the effectiveness of the model used in this study, Kernel SVM (support vector machine), one of the classification models in machine learning, was compared and evaluated using same data. The results showed that the accuracy of the proposed deep neural network was train set 91.79% and test set 85.56% and the specificity was 87.04% and the sensitivity was 83.33% from the point of diagnosis. These results suggest that deep learning is more efficient when classifying these medical data because the train set accuracy in the deep neural network was 7.73% higher than the comparative model Kernel SVM.

• Biomaterials and Biomechanics in Bioengineering
• /
• v.2 no.3
• /
• pp.173-183
• /
• 2015

In obstetrics, cardiotocography is a procedure to record the fetal heartbeat and the uterine contractions usually during the last trimester of pregnancy. It helps to monitor patterns associated with the fetal activity and to detect the pathologies. In this paper, random forest classifier is used to classify normal, suspicious and pathological patterns based on the features extracted from the cardiotocograms. The results showed that random forest classifier can detect these classes successfully with overall classification accuracy of 93.6%. Moreover, important features are identified to reduce the feature space. It is found that using seven important features, similar classification accuracy can be achieved by random forest classifier (93.3%).

• Science of Emotion and Sensibility
• /
• v.11 no.1
• /
• pp.105-112
• /
• 2008

This study investigated the effect of 40% oxygen administration on the addition task performance in three levels of difficulties and physiological signals. Ten male and female college students were selected as the subjects for this study. The experiment consisted of two runs: one was a addition task, with normal air (21% oxygen) administered and the other was with hyperoxic air (40% oxygen) administered. The experimental sequence in each run consisted of Rest1 (3 min), Task1 (1 min, one digit addition task), Task2 (1 min, two digit addition task), Task3 (1 min, three digit addition task), and Rest2 (4 min). Blood oxygen saturation and heart rate were measured throughout the five phases. The accuracy rates of the addition task were enhanced with 40% oxygen administration compared to 21% oxygen. Difference in the accuracy rates grew higher with the rise of difficulty. When 40% concentration oxygen is supplied, blood oxygen saturation increased and heart rate was decreased comparing to 21%. This study showed that the supply of high concentration oxygen increases blood oxygen saturation, which in turns accelerates brain activation resulting from cognitive process and enhances arithmetic abilities. Particularly when difficulty is high, demand for oxygen increases and, as a result, the effect of high concentration oxygen becomes more significant.

• Biomedical Science Letters
• /
• v.21 no.3
• /
• pp.135-143
• /
• 2015

Electrocardiogram measures the electric impulses generated by the heart during its cycle. Recently genome-wide association studies on electrocardiogram traits revealed many relevant genetic loci. Therefore, these findings need to be validated and investigated to determine the underlying mechanisms using mouse models. Invasive radiotelemetry has been widely used to record the electrocardiogram in mice because it has several advantages over non-invasive measurements. However, radiotelemetry is expensive and requires complicated surgery. On the other hand, a non-invasive method using 3 electrodes (one for earth) for lead II is easy to establish and allows for rapid measurement. In this study, eleven mice were measured with this non-invasive method and no statistical difference among them was found in any ECG measurements. In addition, repeat measurement in the same mouse was performed in 9 sets of experiment and the results indicated that non-invasive method was reliable for reproducibility. Further it was shown that measurements for 1, 5, 10, and 15 minutes were not different so that a short recording such as 5 minutes was enough to estimate the ECG values including heart rate. Further this method was validated by measuring the ECG of Balb/c and FVB that were previously shown to differ in ECG values by radiotelemetry. Significant differences were found in heart rate, PR interval and corrected QT interval between these mouse strains. This study partially proved that non-invasive method also could provide the accuracy and reproducibility. Based on these results, the non-invasive ECG recordings of lead II is recommended as a useful method for quick test in mouse model.

• Korean Journal of Artificial Intelligence
• /
• v.11 no.3
• /
• pp.29-34
• /
• 2023

Recently, the rapid development of artificial intelligence technology, many studies are being conducted to predict the risk of heart disease in order to lower the mortality rate of cardiovascular diseases worldwide. This study presents exercise or dietary improvement contents in the form of a software app or web to patients with cardiovascular disease, and cardiovascular disease through digital devices such as mobile phones and PCs. LR, LDA, SVM, XGBoost for the purpose of developing "Life style Improvement Contents (Digital Therapy)" for cardiovascular disease care to help with management or treatment We compared and analyzed cardiovascular disease prediction models using machine learning algorithms. Research Results XGBoost. The algorithm model showed the best predictive model performance with overall accuracy of 80% before and after. Overall, accuracy was 80.0%, F1 Score was 0.77~0.79, and ROC-AUC was 80%~84%, resulting in predictive model performance. Therefore, it was found that the algorithm used in this study can be used as a reference model necessary to verify the validity and accuracy of cardiovascular disease prediction. A cardiovascular disease prediction analysis algorithm that can enter accurate biometric data collected in future clinical trials, add lifestyle management (exercise, eating habits, etc.) elements, and verify the effect and efficacy on cardiovascular-related bio-signals and disease risk. development, ultimately suggesting that it is possible to develop lifestyle improvement contents (Digital Therapy).

• Journal of Korean Medical Science
• /
• v.33 no.53
• /
• pp.343.1-343.8
• /
• 2018

Background: Linkage of public healthcare data is useful in stroke research because patients may visit different sectors of the health system before, during, and after stroke. Therefore, we aimed to establish high-quality big data on stroke in Korea by linking acute stroke registry and national health claim databases. Methods: Acute stroke patients (n = 65,311) with claim data suitable for linkage were included in the Clinical Research Center for Stroke (CRCS) registry during 2006-2014. We linked the CRCS registry with national health claim databases in the Health Insurance Review and Assessment Service (HIRA). Linkage was performed using 6 common variables: birth date, gender, provider identification, receiving year and number, and statement serial number in the benefit claim statement. For matched records, linkage accuracy was evaluated using differences between hospital visiting date in the CRCS registry and the commencement date for health insurance care in HIRA. Results: Of 65,311 CRCS cases, 64,634 were matched to HIRA cases (match rate, 99.0%). The proportion of true matches was 94.4% (n = 61,017) in the matched data. Among true matches (mean age 66.4 years; men 58.4%), the median National Institutes of Health Stroke Scale score was 3 (interquartile range 1-7). When comparing baseline characteristics between true matches and false matches, no substantial difference was observed for any variable. Conclusion: We could establish big data on stroke by linking CRCS registry and HIRA records, using claims data without personal identifiers. We plan to conduct national stroke research and improve stroke care using the linked big database.

• Journal of the Korean Institute of Electrical and Electronic Material Engineers
• /
• v.23 no.1
• /
• pp.76-85
• /
• 2010

In general, anesthetic depth is evaluated by experience of anesthesiologist based on the changes of blood pressure and pulse rate. So it is difficult to guarantee the accuracy in evaluation of anesthetic depth. The efforts to develop the objective index for evaluation of anesthetic depth were continued but there was few progression in this area. Heart rate variability provides much information of autonomic activity of cardiovascular system and almost all anesthetics depress the autonomic activity. Novel monitoring system which can simply and exactly analyze the autonomic activity of cardiovascular system will provide important information for evaluation of anesthetic depth. We investigated the anesthetic depth as following 7 stages. These are pre-anesthesia, induction, skin incision, before extubation, after extubation, Post-anesthesia. In this study, temporal, frequency and chaos analysis method were used to analyze the HRV time series from electrocardiogram signal. There were NN10-NN50, mean, SDNN and RMS parameter in the temporal method. In the frequency method, there are LF and HF and LF/HF ratio, 1/f noise, alphal and alpha2 of DFA analysis parameter. In the chaos analysis, there are CD, entropy and LPE. Chaos analysis method was valuable to estimate the anesthetic depth compared with temporal and frequency method. Because human body was involved the choastic character.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.26 no.10
• /
• pp.1490-1499
• /
• 2022

Arrhythmia is a condition in which the heart has an irregular rhythm or abnormal heart rate, early diagnosis and management is very important because it can cause stroke, cardiac arrest, or even death. In this paper, we propose arrhythmia classification using hybrid combination model of CNN-BLSTM. For this purpose, the QRS features are detected from noise removed signal through pre-processing and a single bit segment was extracted. In this case, the GAN oversampling technique is applied to solve the data imbalance problem. It consisted of CNN layers to extract the patterns of the arrhythmia precisely, used them as the input of the BLSTM. The weights were learned through deep learning and the learning model was evaluated by the validation data. To evaluate the performance of the proposed method, classification accuracy, precision, recall, and F1-score were compared by using the MIT-BIH arrhythmia database. The achieved scores indicate 99.30%, 98.70%, 97.50%, 98.06% in terms of the accuracy, precision, recall, F1 score, respectively.