PubMed Health⌕ Search

Biomedical subjects

W J Tompkins

Publications and source records attributed to W J Tompkins.

At least 19 recordsLinked to original sources
PubMed1999

Classification of premature ventricular complexes using filter bank features, induction of decision trees and a fuzzy rule-based system.

The classification of heart beats is important for automated arrhythmia monitoring devices. The study describes two different classifiers for the identification of premature ventricular complexes (PVCs) in surface ECGs. A decision-tree algorithm based on inductive learning from a training set and a fuzzy rule-based classifier are explained in detail. Traditional features for the classification task are extracted by analysing the heart rate and morphology of the heart beats from a single lead. In addition, a novel set of features based on the use of a filter bank is presented. Filter banks allow for time-frequency-dependent signal processing with low computational effort. The performance of the classifiers is evaluated on the MIT-BIH database following the AAMI recommendations. The decision-tree algorithm has a gross sensitivity of 85.3% and a positive predictivity of 85.2%, whereas the gross sensitivity of the fuzzy rule-based system is 81.3%, and the positive predictivity is 80.6%.

Decision Trees↗
PubMed1999

ECG beat detection using filter banks.

We have designed a multirate digital signal processing algorithm to detect heart beats in the electrocardiogram (ECG). The algorithm incorporates a filter bank (FB) which decomposes the ECG into subbands with uniform frequency bandwidths. The FB-based algorithm enables independent time and frequency analysis to be performed on a signal. Features computed from a set of the subbands and a heuristic detection strategy are used to fuse decisions from multiple one-channel beat detection algorithms. The overall beat detection algorithm has a sensitivity of 99.59% and a positive predictivity of 99.56% against the MIT/BIH database. Furthermore this is a real-time algorithm since its beat detection latency is minimal. The FB-based beat detection algorithm also inherently lends itself to a computationally efficient structure since the detection logic operates at the subband rate. The FB-based structure is potentially useful for performing multiple ECG processing tasks using one set of preprocessing filters.

Algorithms↗
PubMed1997

Comparison of impedance and inductance ventilation sensors on adults during breathing, motion, and simulated airway obstruction.

The goal of this study was to compare the relative performance of two noninvasive ventilation sensing technologies on adults during artifacts. We recorded changes in transthoracic impedance and cross-sectional area of the abdomen (abd) and rib cage (rc) using impedance pneumography (IP) and respiratory inductance plethysmography (RIP) on ten adult subjects during natural breathing, motion artifact, simulated airway obstruction, yawning, snoring, apnea, and coughing. We used a pneumotachometer to measure air flow and tidal volume as the standard. We calibrated all sensors during natural breathing, and performed measurements during all maneuvers without changing the calibration parameters. No sensor provided the most-accurate measure of tidal volume for all maneuvers. Overall, the combination of inductance sensors [RIP(sum)] calibrated during an isovolume maneuver had a bias (weighted mean difference) as low or lower than all individual sensors and all combinations of sensors. The IP(rc) sensor had a bias as low or lower than any individual sensor. The cross-correlation coefficient between sensors was high during natural breathing, but decreased during artifacts. The cross correlation between sensor pairs was lower during artifacts without breathing than it was during maneuvers with breathing for four different sensor combinations. We tested a simple breath-detection algorithm on all sensors and found that RIP(sum) resulted in the fewest number of false breath detections, with sensitivity of 90.8% and positive predictivity of 93.6%.

Adult↗
PubMed1997

A patient-adaptable ECG beat classifier using a mixture of experts approach.

We present a "mixture-of-experts" (MOE) approach to develop customized electrocardiogram (ECG) beat classifier in an effort to further improve the performance of ECG processing and to offer individualized health care. A small customized classifier is developed based on brief, patient-specific ECG data. It is then combined with a global classifier, which is tuned to a large ECG database of many patients, to form a MOE classifier structure. Tested with MIT/BIH arrhythmia database, we observe significant performance enhancement using this approach.

Adaptation, Physiological↗
PubMed1995

Optimisation of transcutaneous cardiac pacing by three-dimensional finite element modelling of the human thorax.

The goal of the study is to determine by finite element analysis (FE) the optimal electrode placement, size and electrolyte resistivity that minimise the pain experienced by patients during successful transcutaneous cardiac pacing (TCP). The three-dimensional FE model generated for this purpose has 55,388 nodes, 50,913 hexahedral elements and simulated 16 different organs and tissues, as well as the properties of the electrolyte. The model uses a non-uniform mesh with an average spatial resolution of 0.8 cm in all three dimensions. To validate this model, the voltage across 3 cm2 Ag-AgCl electrodes is measured when currents of 5 mA at 50 kHz are injected into a subject's thorax through the same electrodes. For the same electrode placements and sizes and the same injected current, the FE analysis produced results in good agreement with the experimental data. The optimisation analysis tested seven different electrode placements, five different electrode sizes and six different electrolyte resistivities. The analysis indicates that the anterior-posterior electrode placement, electrode sizes of about 90 cm2 and electrolytes with resistivity of about 800 omega.cm yield the most uniform current distribution through the skin, thus having the best chances to minimise the pain delivered to the patient during successful TCP. The anterior-anterior electrode placement is the second most efficient.

Cardiac Pacing, Artificial↗
PubMed1995

Optimization of cardiac defibrillation by three-dimensional finite element modeling of the human thorax.

The goal of this study was to determine the optimal electrode placement and size to minimize myocardial damage during defibrillation while rendering refractory a critical mass of cardiac tissue of 100%. For this purpose, we developed a 3-D finite element model with 55,388 nodes, 50,913 hexahedral elements, and simulated 16 different organs and tissues, as well as the properties of the electrolyte. The model used a nonuniform mesh with an average spatial resolution of 0.8 cm in all three dimensions. To validate this model, we measured the voltage across 3-cm2 Ag-AgCl electrodes when currents of 5 mA at 50 kHz were injected into a human subject's thorax through the same electrodes. For the same electrode placements and sizes and the same injected current, the finite element analysis produced results in good agreement with the experimental data. For the optimization of defibrillation, we tested 12 different electrode placements and seven different electrode sizes. The finite element analyses showed that the anterior-posterior electrode placement and an electrode size of about 90 cm2 offered the least chance of potential myocardial damage and required a shock energy of less than 350 J for 5-ms defibrillation pulses to achieve 100% critical mass. For comparison, the average cross-sectional area of the heart is approximately 48 cm2, about half of the optimal area. A second best electrode placement was with the defibrillation electrodes on the midaxillary lines under the armpits. Although this placement had higher chances of producing cardiac damage, it required less shock energy to achieve 100% critical mass.

Computer Simulation↗
PubMed1995

A database of cardiac arrhythmias.

OBJECTIVE: To describe a database of cardiac arrhythmia recordings, useful for the development and testing of ECG rhythm processing or monitoring algorithms and devices. METHODS: The raw data were acquired within the Wisconsin-Dane County emergency medical technician-defibrillation program and contained emergency rhythm recordings of an average length of 30 minutes. The raw data were integrated into a software platform designed for the annotation and visualization of the recordings. RESULTS: Currently the database contains the following arrhythmia episodes: ventricular fibrillation (56), asystole (65), electromechanical dissociation (31), and other arrhythmias (42). The software, resident on personal computers, also can transmit any of the database recordings, through a digital-to-analog converter board, to a device under test. CONCLUSIONS: The database technique described will provide a useful means of objectively assessing electronic devices for their ability to detect arrhythmias. The database is unique in that it contains lengthy episodes of arrhythmias. The database will be extended to include additional cases.

Arrhythmias, Cardiac↗
PubMed1994

Finite-element method in electrical impedance tomography.

In electrical impedance tomography (EIT), current patterns are injected into a subject and boundary voltages are measured to reconstruct a cross-sectional image of resistivity distribution. Static EIT image reconstruction requires a computer model of a subject, an efficient data-collection method and robust and fast reconstruction algorithms. The finite-element method is used as the computer model. The paper describes the finite-element analysis software package developed, including an interactive graphical mesh generator and fast algorithms for solving linear systems of equations using sparse-matrix and vector techniques. Various models of irregularly shaped subjects are developed using mesh-design tools, including automatic mesh generation and optimisation using the Delaunay algorithm. Even though the software package is customised for use in electrical impedance tomography, it can be used for other biomedical research areas, such as impedance cardiography, cardiac defibrillation and impedance pneumography.

Algorithms↗
PubMed1994

Parameter evaluation of the inverse power-law spectrum of heart rate. A quantitative approach for ECG arrhythmia analysis.

A preliminary study was performed to ensure reliable R-wave detection and fiducial mark location. A generalized phase-sampling model for spectral estimation was then used, based on an interest in beat-to-beat variations. Heart rate spectrum characteristics of 32 presumed-normal subjects and 44 records from the MIT-BIH electrocardiographic database (Massachusetts Institute of Technology, Cambridge, MA) were analyzed and three parameters were calculated: regression line slope, intercept, and cross correlation between the spectral data and the regression lines. Evaluation of these inverse power-law spectrum parameters provides a potential quantitative approach for characterizing erratic fluctuations of heart rate and is possibly used to distinguish between healthy and abnormal subjects. From the 44 recordings in the MIT-BIH database, a V-shaped curve was found in a plot of the regression line slope versus cross correlation. The results for both unmedicated and medicated patients with normal sinus rhythm cluster in the top left region of the graph. Also, 29 of the presumed-normal subjects cluster in the same region. Patients with premature ventricular contraction beats cluster in the top right region of the V-shaped curve. The rest of the recordings from the variety of arrhythmia cases in the database have low slopes and cross correlations, so they cluster near the apex of the V-shaped curve. Three volunteers who each had more than 32 atrial premature contraction and premature ventricular contraction beats also fall in this region of the graph. The results from 10 young volunteers from the United States and 15 volunteers from seven other countries cluster into different regions.(ABSTRACT TRUNCATED AT 250 WORDS)

Adult↗
PubMed1994

Design of an inductive plethysmograph for ventilation measurement.

We have designed an inductive plethysmograph to obtain a non-invasive measure of ventilation. Two elastic bands containing insulated wires encircle the chest and abdomen--the inductance of each band depends on the enclosed cross sectional area. Each inductive band forms an element in a tank circuit, which determines the resonant frequency of a Colpitts oscillatory. By measuring the oscillatory frequency, we indirectly measure the changes in cross sectional area that occur during breathing. Independent measures of chest and abdominal cross sectional area provide a way to detect both normal breathing and airway obstruction. Magnetic coupling due to the mutual inductance between chest and abdominal bands modulates the desired oscillation frequencies. When modulation is excessive, frequency locking occurs and we cannot make independent measures of chest and abdominal area. We have performed simulations that show that, as the chest and abdominal band oscillator frequencies are sufficiently separated, we decrease modulation and avoid frequency locking. We have compared simulataneous recordings of ventilation using our inductive plethysmography and a commercial impedance pneumograph and spirometer. Recordings of normal ventilation by all methods appear similar; however, our inductive device is less prone than the impedance pneumography to artifacts caused by applied pressure and body movements. In addition, during simulated airway obstruction, signals from the chest and abdominal bands are out of phase--suggesting that the inductive technique may be useful for detecting airway obstruction.

Abdomen↗
PubMed1994

A comparison of electrodes for potential use in paediatric/infant apnoea monitoring.

We measured the signal-to-motion-artifact ratio for ten paediatric/infant electrodes for infant apnoea monitoring. Electrodes that have good stability, strong adhesion, low face-to-face impedance, low transthoracic-plus-electrode impedance, large effective area, and large total area are optimal for impedance pneumography. The gel make-up may also affect optimality.

Adhesiveness↗
PubMed1993

Using compound electrodes in electrical impedance tomography.

In electrical impedance tomography, we inject currents and measure voltages to estimate an object's resistivity distribution. The electrode configuration affects measured voltage data because the electrode-skin contact impedance is high and varies with electrode location. We developed a compound electrode which is composed of two electrodes: a large outer electrode to inject current and a small inner electrode to sense voltage. We used these compound electrodes to measure voltages from a physical phantom. We showed that the measured voltages from the compound electrodes are smaller in amplitude than those from conventional electrodes. This demonstrates that the compound electrode can minimize contact impedance voltage drop from the measured data. We used a finite element model for the compound electrode and incorporated the model into the regularized Newton-Raphson reconstruction algorithm. We performed a sensitivity study and showed that the reconstructed resistivity distributions are less dependent on the unknown contact resistance values for a compound electrode than a conventional electrode and that the use of a compound electrode results in improved images for the reconstruction algorithm.

Algorithms↗
PubMed1993

Finite element modeling of electrode-skin contact impedance in electrical impedance tomography.

In electrical impedance tomography (EIT), we inject currents through and measure voltages from an array of surface electrodes. The measured voltages are sensitive to electrode-skin contact impedance because the contact impedance and the current density through this contact impedance are both high. We used large electrodes to provide a more uniform current distribution and reduce the contact impedance. A large electrode differs from a point electrode in that it has shunting and edge effects which cannot be modeled by a single resistor. We used the finite element method (FEM) to study the electric field distributions underneath an electrode, and developed three models: a FEM model, a simplified FEM model and a weighted load model. We showed that the FEM models considered both shunting and edge effects and matched closely the experimental measurements. FEM models for electrodes can be used to improve the performance of an electrical impedance tomography reconstruction algorithm.

Algorithms↗
PubMed1993

Applications of artificial neural networks for ECG signal detection and classification.

The authors have investigated potential applications of artificial neural networks for electrocardiographic QRS detection and beat classification. For the task of QRS detection, the authors used an adaptive multilayer perceptron structure to model the nonlinear background noise so as to enhance the QRS complex. This provided more reliable detection of QRS complexes even in a noisy environment. For electrocardiographic QRS complex pattern classification, an artificial neural network adaptive multilayer perceptron was used as a pattern classifier to distinguish between normal and abnormal beat patterns, as well as to classify 12 different abnormal beat morphologies. Preliminary results using the MIT/BIH (Massachusetts Institute of Technology/Beth Israel Hospital, Cambridge, MA) arrhythmia database are encouraging.

Arrhythmias, Cardiac↗
PubMed1992

Skin impedance measurements using simple and compound electrodes.

We have studied the effect of the electrode configuration on the measurement of body impedance and found that the electrode configuration greatly affects the impedance measurement using the four-electrode method. We studied the characteristics of the compound electrode and found that the compound electrode provides the four-electrode method in a compact form. A new method of measuring the skin impedance using simple electrodes at low frequencies was developed. At high frequencies where the effect of internal tissue impedance is not negligible, we used the compensation method using compound electrodes, because they measure the voltage right under the skin. At 50 kHz, we measured the real part of the skin impedance of less than 80 omega on the thorax. We propose a simple instrument which can measure accurate skin impedance at various frequencies.

Electric Conductivity↗
PubMed1992

Neural-network-based adaptive matched filtering for QRS detection.

We have developed an adaptive matched filtering algorithm based upon an artificial neural network (ANN) for QRS detection. We use an ANN adaptive whitening filter to model the lower frequencies of the ECG which are inherently nonlinear and nonstationary. The residual signal which contains mostly higher frequency QRS complex energy is then passed through a linear matched filter to detect the location of the QRS complex. We developed an algorithm to adaptively update the matched filter template from the detected QRS complex in the ECG signal itself so that the template can be customized to an individual subject. This ANN whitening filter is very effective at removing the time-varying, nonlinear noise characteristic of ECG signals. Using this novel approach, the detection rate for a very noisy patient record in the MIT/BIH arrhythmia database is 99.5%, which compares favorably to the 97.5% obtained using a linear adaptive whitening filter and the 96.5% achieved with a bandpass filtering method.

Algorithms↗
PubMed1992

Measuring lung resistivity using electrical impedance tomography.

We propose the use of electrical impedance tomography (EIT) imaging techniques in the measurement of lung resistivity for detection and monitoring of apnea and edema. In EIT, we inject currents into a subject using multiple electrodes and measure boundary voltages to reconstruct a cross-sectional image of internal resistivity distribution. We found that a simplified, therefore fast, version of the impedance imaging method can be used for detection and monitoring of apnea and edema. We have showed the feasibility of this method through computer simulations and human experiments. We speculate that the EIT imaging technique will be more reliable than the current impedance apnea monitoring method, since we are monitoring the change of internal lung resistivity. However, more study is required to verify that this method performs better in the presence of motion artifact than the conventional two-electrode impedance apnea monitoring method. Future work should include experiments which carefully simulate different kinds of motion artifacts.

Adult↗