• Journal of Information Processing Systems
• /
• 제14권4호
• /
• pp.851-876
• /
• 2018

Dynamic time warping (DTW) is the main algorithms for time series alignment. However, it is unsuitable for quasi-periodic time series. In the current situation, except the recently published the shape exchange algorithm (SEA) method and its derivatives, no other technique is able to handle alignment of this type of very complex time series. In this work, we propose a novel algorithm that combines the advantages of the SEA and the DTW methods. Our main contribution consists in the elevation of the DTW power of alignment from the lowest level (Class A, non-periodic time series) to the highest level (Class C, multiple-periods time series containing different number of periods each), according to the recent classification of time series alignment methods proposed by Boucheham (Int J Mach Learn Cybern, vol. 4, no. 5, pp. 537-550, 2013). The new method (quasi-periodic dynamic time warping [QP-DTW]) was compared to both SEA and DTW methods on electrocardiogram (ECG) time series, selected from the Massachusetts Institute of Technology - Beth Israel Hospital (MIT-BIH) public database and from the PTB Diagnostic ECG Database. Results show that the proposed algorithm is more effective than DTW and SEA in terms of alignment accuracy on both qualitative and quantitative levels. Therefore, QP-DTW would potentially be more suitable for many applications related to time series (e.g., data mining, pattern recognition, search/retrieval, motif discovery, classification, etc.).

• 한국정보통신학회논문지
• /
• 제18권7호
• /
• pp.1531-1539
• /
• 2014

조기심실수축(PVC)은 가장 보편적인 부정맥으로 심실세동, 심실빈맥 등과 같은 위험한 상황을 유발할 수 있는 가능성을 가지고 있기 때문에 이의 조기 검출은 매우 중요하다. 하지만 ECG 신호의 개인 차이가 있음에도 불구하고, 일반적인 신호의 판단 규칙에 따라 진단을 수행함으로써 성능하락이 나타날 수 밖에 없다. 이러한 문제점을 극복하기 위해서는 개인에 따른 이상 신호를 검출한 후 다양한 QRS 패턴을 고려하여 PVC를 분류할 수 있는 알고리즘이 필요하다. 본 연구에서는 개인별 이상신호 검출과 QRS 패턴 변화에 따른 PVC 분류 기법을 제안한다. 이를 위해 전 처리 과정과 차감기법을 통해 R파를 검출하였으며, 개인별 이상신호를 검출하였다. 이후 QRS 패턴에 따른 QS 간격과 R파의 진폭 변화율에 따라 PVC를 분류하였다. 제안한 알고리즘의 이상 신호 검출 및 PVC 분류 성능을 평가하기 위해서 MIT-BIH 부정맥 데이터베이스를 사용하였다. 성능평가 결과, 이상 신호 검출률은 98.33%, PVC는 각각 94.46%의 평균 분류율을 나타내었다.

• 대한의용생체공학회:의공학회지
• /
• 제20권4호
• /
• pp.461-469
• /
• 1999

환자감시장치는 중환자실, 수술실, 응급실 및 병실에서 환자의 상태인 심전도 파형, 맥박수 , 혈압등을 측정하고 감시하는 기본적인 의료장비이다. 본 논문에서 설계한 환자감시장치는 심전도-비관혈식 혈압 감지장치이다. 심전도는 심장 근육의 이완과 수축에 따라 발생하는 전기 현상으로 대부분의 의사들은 환자의 심장 상태를 심전도 신호 패턴으로 진단을 하고 있다. 심전도 감시장치는 장시간동안 환자의 심장상태를 감시하는데 사용된다. 환자의 혈압을 재는 일은 일반화된 임상 측정의 하나로, 진찰실에서나 또는 특별한 수술중에도 시행되고 있다. 본 논문에서는 동맥 내의 혈압을 비관혈적으로 측정하는 오실로메트릭 방식의 간접 측정 방법을 사용하였다. 개발한 환자감시장치를 수술실 장비와 비교, 검토하였다. 매회 심박수의 최대 차이는 1bpm이며,수축기 혈압 최대 차이는 15mmHg, 이완기 혈압 최대차이는 16mmHg, 평균 혈압 최대 차이는 25mmHg를 보였다. 그러나 결과적으로 나타나는 심박수의 평균 오차는 0.15bpm이며, 수축기 혈압의 평균 오차는 5mmHg, 이완기 혈압의 평균 오차는 10mmHg, 그리고 평균 혈압의 평균 오차는 9mmHg로 나타났다. 본 논문의 심전도-비관혈식 혈압 환자감시장치는 심전도 파형, 맥박수, 그리고 혈압을 측정하고 감시하는 의료 장비로서, 설계된 환자감시장치에 산소포화도.호흡.관혈식 혈압감시장치(IPB) 등의 다기능을 모듈로 구성하여 부착 및 제거가 용이하도록 확장 할 수 있다.

• 대한의생명과학회지
• /
• 제24권1호
• /
• pp.43-49
• /
• 2018

Electrocardiogram (ECG) is a widely used method to diagnose electrical activity of heart. Although it is a reliable and easy method, ECG could be interfered by electrical signals. One of the interfering signals is electromyogram (EMG) that is caused by muscle contraction in any parts of the body except heart. To avoid the EMG noise, an examinee is advised to be relaxed on supine position while measuring ECG. Sometimes, patients who can't put their arms and legs down on bed due to some reasons such as cast on arms or legs necessarily have the EMG noise. But detailed information about how much of the noise could be induced by positional change of arms and legs has not been reported. Here we examined the noise by analyzing ECG data from 14 candidates, 7 males and 7 females. The ECG data was obtained using the standard 12 lead ECG. EMG noise was induced by raising arms and legs at $90^{\circ}$, $60^{\circ}$ or $30^{\circ}$. Because arms are located close to the heart, noise by the raised arms was analyzed toward left or right arm separately. All of the examinees showed similar pattern of the EMG noise. EMG noise by positional change of left or right arm was clearly monitored in different limb leads. Change of leg positions induced the noise that was monitored in aVF of augmented leads and II and III of limb leads. There was a difference in degree of the noise between male and female examinees. In addition to the EMG noise, decrease of PR interval was monitored in particular positional changes, which was prominent in male examinees. These results will enlarge fundamental understanding about EMG noise in ECG.

• 제어로봇시스템학회논문지
• /
• 제13권8호
• /
• pp.777-782
• /
• 2007

Human Robot Interaction(HRI) through a haptic interface plays an important role in controlling robot systems remotely. The augmented usage of bio-signals in the haptic interface is an emerging research area. To consider operator's state in HRI, we used bio-signals such as ECG and blood pressure in our proposed force reflection interface. The variation of operator's state is checked from the information processing of bio-signals. The statistical standard variation in the R-R intervals and blood pressure were used to adaptively adjust force reflection which is generated from environmental condition. To change the pattern of force reflection according to the state of the human operator is our main idea. A set of experiments show the promising results on our concepts of human adaptive interface.

• 대한전기학회논문지:시스템및제어부문D
• /
• 제52권1호
• /
• pp.55-63
• /
• 2003

In this study was proposed that a new estimating method for investigation of contractile state changes which generated from continuous isometric contraction of skeletal muscle. The physiological changes(EMG, ECG) and the psychological changes by CNS(central nervous system) were measured by experiments, while the muscle of subjects contracted continuously with isometric contraction in constant load. The psychological changes were represented as three-step-change named 'fatigue', 'pain' and 'sick(greatly pain)' from oral test, and the method which compared physiological change with psychological change on basis of these three steps was developed. The result of analyzing the physiological signals, EMG and ECG signal changes were observed at the vicinity of judging point in time of psychological changes. Namely, it is supposed that contractile states have three kind of states pattern (stable, fatigue, pain) instead of two states (stable, fatigue).

• 한국지능시스템학회논문지
• /
• 제14권1호
• /
• pp.82-87
• /
• 2004

심전도 데이터는 심장의 전기적인 신호의 다양한 파형으로 이루어지며, 이와 같은 파형을 분석하고 분류하기 위하여 데이터마이닝 기법을 이용할 수 있다. 심전도신호를 분류하기 위한 기존의 연구들은 왜곡된 특징추출과 과적합 등 문제점을 가지고 있다. 본 연구에서는 이와 같은 문제점들을 해결하기 위하여 BP 알고리즘과 SVM을 이용하여 심전도 신호를 분류해 보았다 그 결과 SVM이 신경망에서 발생하는 과적합을 효과적으로 방지하고, 유일한 전역해를 보장함으로써, 일반화 성능에서 우수함을 보이고 있다는 사실을 확인하였다.

• 센서학회지
• /
• 제31권1호
• /
• pp.16-23
• /
• 2022

Cardiac disease is the most common cause of death worldwide. Therefore, detection and classification of electrocardiogram (ECG) signals are crucial to extend life expectancy. In this study, we aimed to implement an artificial intelligence signal recognition system in field programmable gate array (FPGA), which can recognize patterns of bio-signals such as ECG in edge devices that require batteries. Despite the increment in classification accuracy, deep learning models require exorbitant computational resources and power, which makes the mapping of deep neural networks slow and implementation on wearable devices challenging. To overcome these limitations, spiking neural networks (SNNs) have been applied. SNNs are biologically inspired, event-driven neural networks that compute and transfer information using discrete spikes, which require fewer operations and less complex hardware resources. Thus, they are more energy-efficient compared to other artificial neural networks algorithms.

• Progress in Superconductivity
• /
• 제9권1호
• /
• pp.35-39
• /
• 2007

Myocardial ischemia causes heterogeneity of ventricular repolarization and sometimes produces changes of the ST-T wave in ECG. Therefore, morphological changes of ST-T waveform in ECG have a clinical significance in diagnosing myocardial ischemia. In this study, we investigated the ST-T changes caused by myocardial ischemia in magnetocardiography (MCG). We analyzed MCG patterns of biphasic T, ST segment deviations from baseline, main current angle of $T_{peak}$ and $T_{peak}$ dispersion in 300 CAD patients without ST elevation in ECG, 122 symptomatic patients and 48 normal subjects. MCGs were recorded by multichannel SQUID system in a magnetically shielded room. As results, we found that appearances of the abnormality were strongly correlated with the severity of myocardial ischemia. Also we found that the percentage of the patients showing MCG changes were higher than those in ECG. These results show that morphological changes of ST-T waveform in MCG can be used as a marker of myocardial ischemia.

• 대한임상검사과학회지
• /
• 제44권3호
• /
• pp.136-141
• /
• 2012

The hearts of highly trained athletes show morphologic and electrocardiographic (ECG) changes that suggest the presence of cardiovascular disease, including sinus bradycardia, a striking increase in precordial R-wave or S-wave voltages, ST segment depression, and T-wave inversions. Despite a number of previous observational surveys, the determinants of abnormal ECG patterns in trained athletes remain largely unresolved. In this study, we compared the electrocardiographic characteristics of athletes to determine any sensitive indicators. Comparison between ECG patterns and cardiac physiology was performed in 21 junior athletes and 25 untrained subjects with no signs of cardiac disease. Sinus bradycardia was detected in a subset of athletes but not statistically significant between the athletes ($69.9{\pm}11.1bpm$) and the control ($72.7{\pm}9.9bpm$) group. The mean values of the PR and QTc intervals in the athletes' group were $149.2{\pm}15.4ms$ and $402.3{\pm}28.8ms$, respectively. Also, there were no significantly differences between control group and the athletes' group. In addition, the athletes demonstrated a spectrum of alterations in the 12-lead ECG pattern, including marked increase in precordial R-wave or S-wave voltages ($$SV_1+RV_5{\geq_-}35mm$$, 23.8%), QRS duration ($${\geq_-}90ms$$, 90.5%), suggestive of left ventricular hypertrophy. However, left axis deviation, ST segment depression, and T-wave changes in V5, V6 were not observed in either the athletes or control group. Our findings suggest that sinus bradycardia, precordial R-wave or S-wave voltages, and QRS duration seem to be more sensitively detected in athletes than in control group. Further researches on the electrocardiographic patterns of athletes should be carried out to improve the sensitivity and specificity of diagnostic criteria.