Kyoung Joung Lee

Adaptive Noise Canceling of Electrocardiogram Artifacts in Single Channel Electroencephalogram

A new method for estimating and eliminating electrocardiogram (ECG) artifacts from single channel scalp electroencephalogram (EEG) is proposed. The proposed method consists of emphasis of QRS complex from EEG using least squares acceleration (LSA) filter, generation of synchronized pulse with R-peak and ECG artifacts estimation and elimination using adaptive filter. The performance of the proposed method was evaluated using simulated and real EEG recordings, we found that the ECG artifacts were successfully estimated and eliminated in comparison with the conventional multi-channel techniques, which are independent component analysis (ICA) and ensemble average (EA) method. From this we can conclude that the proposed method is useful for the detecting and eliminating the ECG artifacts from single channel EEG and simple to use for ambulatory/portable EEG monitoring system.

Motion Artifact Reduction in Blood Pressure Signals Using Adaptive Digital Filter with a Capacitive Sensor

Motion artifacts can degrade the quality of data measured. In this paper, we suggest a capacitive sensor application which is able to be fabricated easily and employed for adaptive filtering. When the motion artifacts occur, the capacitance varies as well. Thereby, the capacitance data coincide with the motion artifacts. The capacitance data synchronous with the motion artifacts and the corrupted blood pressure signal were applied to the adaptive digital filter so that motion artifacts in blood pressure signal managed to be removed successfully.

A New Acupuncture Point Detection Using the Impedance Measurement System Based on ANF and Phase-Space-Method

In this study, a new method for acupuncture point detection using the impedance measurement system based on the PSM (phase space method) is presented. The developed device has been developed as detectors for acupuncture points which are used for diagnosis and treatment in acupuncture. In this system, multi-frequency current injection and voltage measurements are both performed by the surface electrodes, which are controlled by a microcontroller. Also, the microcontroller process continuous time demodulation of the modulated signal by multi frequency components using the adaptive notch filter. After that, PSM is applied about each frequency using an acupuncture equivalent model which is proposed in the pre-study.

Optical polarimetry probe system for glucose concentration monitoring

Over the last two decades, optical polarimetry method has been applied for glucose concentration monitoring in biological media such as aqueous humor and cell culture media as a non-invasive alternative method. Compared to glucose analyzer and electronic analyte sensor, the advantages of polaimetry method are non-ionizing radiation to interrogate the sample, non-interference with the sample, needless of consumable reagents, use of readily available sources, and prospect of miniaturizing the optics. Commercial polarimeter is widely used to measure the concentration of chemical compounds with optical activity. However, this device was based on off-line measurement so that it needs sample extraction process to measure the concentration of sample. This process does not reflect the real-time status of sample concentration and sample contamination can be occurred during sample extraction when applied to cell culture process. In polarimetry method, the measurement sensitivity can be controlled by varying optical path length. However, in current polarimeter, the sample cell should be exchanged to vary the optical path length. This process is a time consuming and might cause sample contamination in cell culture process. Therefore, it is necessary to develop a new polarimetry method which can measure the real-time status of sample concentration without sample extraction. In this paper, we introduce a new polarimetry probe system which might be utilized to monitor glucose concentration during cell culture process. It was designed to have variable optical path lengths to control the optical rotation angle of polarized light. We describe the feasibility of the system and the preliminary results.

Minimization of artifact using adaptive digital filter during the oscillometric blood pressure measurement

Artifacts caused during the oscillometric blood pressure measurement can degrade the quality of data measured. In this paper, the stepped deflation artifact(SDA) and motion artifact have been dealt with. The former and the latter are due to the bleed valve switching and cuff movements, respectively. The adaptive impulse correlated filter (AICF) was applied to remove the stepped deflation artifact. This AICF is performed using a parametric method which use Finite Impulse Response(FIR) structure. A capacitive sensor was utilized in order to minimize the motion artifact caused by cuff movements. A control signal formed from the capacitance data with respect to the motion artifact was applied to the adaptive digital filter. The removal of artifact was worked almost perfectly without reducing the information in the blood pressure pulsation signal.

The Novel Method for the Fetal Electrocardiogram Extraction from the Abdominal Signal

In this paper, we have proposed a new method to extract the fetal ECG from a pregnant womans abdominal signal using least square acceleration (LSA) filter and adaptive impulse correlation (AIC) filter. To evaluate the performance, the proposed method and other fetal ECG extraction techniques were processed using the synthetic and real ECG data and then the results were compared. According to comparative results, the proposed method is powerful and successful for extracting the fetal ECG. It was able to separate perfectly even though the fetal beats overlap with the QRS wave of the maternal beats and to extract fetal ECG using any single-channel abdominal signal measured from pregnant womans abdominal surface. Also, it could be implemented easily by fast computation time and simple structure. It is sure that our method could be useful for portable fetal monitoring system.

Amplitude based beat detection for atrial fibrillation in pacemaker

Bradycardia is defined as a sinus rhythm of less than 60 beats per minute and atrial tachyarrhythmia including atrial fibrillation (AF) is frequently associated with bradycardia. Pacemaker is the only effective treatment for symptomatic bradycardia and automatic mode switching (AMS) function is built in pacemaker to switch mode in the presence of atrial tachyarrhythmia. AMS algorithms consider appropriate mode switching in case of undersensing or oversensing and this consideration makes their onset time and resynchronization time late. Current pacemakers have onset time from 2.5 seconds to 26 seconds and resynchronization time from 3.4 seconds to 143 seconds according to manufacturers. In this work, we proposed beat detection algorithm based on amplitude difference between peak and trough for accurate extraction of atrial rate achieving faster mode switching. Evaluation of beat detection algorithm was conducted with six canine AF electrogram (EGM) data. Result showed 96.64% sensitivity, 95.5% positive predictive value in average. With this, transition from AF to normal sinus rhythm could be detected faster than existing AMS algorithms. In conclusion, proposed algorithm can efficiently detect beats in EGM during AF and from this, we can implement faster AMS algorithm.Bradycardia is defined as a sinus rhythm of less than 60 beats per minute and atrial tachyarrhythmia including atrial fibrillation (AF) is frequently associated with bradycardia. Pacemaker is the only effective treatment for symptomatic bradycardia and automatic mode switching (AMS) function is built in pacemaker to switch mode in the presence of atrial tachyarrhythmia. AMS algorithms consider appropriate mode switching in case of undersensing or oversensing and this consideration makes their onset time and resynchronization time late. Current pacemakers have onset time from 2.5 seconds to 26 seconds and resynchronization time from 3.4 seconds to 143 seconds according to manufacturers. In this work, we proposed beat detection algorithm based on amplitude difference between peak and trough for accurate extraction of atrial rate achieving faster mode switching. Evaluation of beat detection algorithm was conducted with six canine AF electrogram (EGM) data. Result showed 96.64% sensitivity, 95.5% positive predictive value in average. With this, transition from AF to normal sinus rhythm could be detected faster than existing AMS algorithms. In conclusion, proposed algorithm can efficiently detect beats in EGM during AF and from this, we can implement faster AMS algorithm.

A pressure controlled low-level laser probe to enhance photon density in soft tissue

Noninvasive low-level laser devices have been introduced for therapeutic purpose in medicine. However, low-level laser cannot deliver enough photon density to expect positive therapeutic results in deep tissue layer due to light scattering property in tissue. In order to overcome the limitation, this study was aimed to develop a negative pressure applied low-level laser probe to enhance laser transmission and, therefore, photon density in soft tissue. In order to evaluate clinical feasibility of developed laser probe, ex-vivo experiments were performed with porcine skin samples and laser transmissions were quantitatively measured as a function of tissue compression. The laser probe has an air suction hole to apply negative pressure to skin, a transparent plastic body to observe tissue deformation, and a small metallic optical fiber guide to support the optical fiber when negative pressure was applied. By applying negative pressure to the laser probe, the porcine skin under the fiber guide is compressed down and, low-level laser is emitted into the skin. Diffusion images of laser in the skin samples were acquired with a CCD camera and analyzed. Compared to the intensity without compression, the peak intensity of laser beam profiles increased about 2~2.5 times and FWHM (Full Width at Half Maximum) decreased about 1.67~2.85 times. In addition, the peak intensity was linearly increased as a function of compression. In conclusion, we verified the enhancement of laser transmission and therefore, photon density in tissue by applying negative pressure to the developed low-level laser probe and its potential for clinical usefulness.

Contrast enhancement of laser speckle skin image: use of optical clearing agent in conjunction with micro-needling

Laser speckle imaging modality is one of widely used methods to evaluate blood flow because of its simplicity. However, laser speckle image has a limitation in the evaluation of subcutaneous blood flow due to its low contrast perfusion image. Various methods have been tried to enhance the perfusion image contrast. Such methods presented positive results in some degree. However, it could not be fundamental solutions due to low penetration depth of lasers restricted by optical tissue scattering property. This study suggests a method to enhance the perfusion image contrast of laser speckle imaging modality by increasing the penetration depth of lasers. An optical clearing agent (glycerol) was topically applied on skin treated with micro-needle roller in order to reduce the time period of optical tissue clearing and therefore, enhance the penetration depth of laser. In this study, we investigated the effect of glycerol and micro-needling methods in the contrast enhancement of laser speckle perfusion skin image and presented the results of in-vitro and in-vivo animal experiment.

Discriminant analysis of HRV's autoregressive model coefficient for cardiac arrhythmia

An attempt is made to obtain an autoregressive (AR) model coefficient for heart rate variability (HRV) and then to classify the symptom groups of the coefficient. As a result of analyzing 104 sets of clinical data, the following conclusions are drawn. (1) HRV obtained from clinical data can be considered time-series data. (2) The optimal condition for classification is to assume that the coefficient is at 110 R-R intervals with 11th order. This fixed order shows a tendency to approach the largest one among the representative optimal orders of each symptom.<<ETX>>