Proceedings of the 19th Annual International Conference of the IEEE Engineering in Medicine and Biology Society. 'Magnificent Milestones and Emerging Opportunities in Medical Engineering' (Cat. No.97CH36136)
The relationships between the carotid baroreflex and the nonlinear chaotic dynamics of short-term... more The relationships between the carotid baroreflex and the nonlinear chaotic dynamics of short-term arterial blood pressure control was investigated. Studies in conscious dogs with their baroreceptors either intact (control group: n=7) or surgically isolated (intervention group: n=7), demonstrated deterministic chaotic behavior in mean arterial blood pressure. To assess nonlinear dynamical behavior, we constructed phase-space trajectories from the time series data. The correlation dimension of the reconstructed trajectories was measured using the Grassberger-Procaccia algorithm. A significant increase (PcO.01) in the correlation dimension was measurable between the control (3.46k0.38) and intervention groups (4.59k0.23). Sensitive dependence on initial conditions as determined by the largest Lyaponuv exponents, revealed a significant (Pc0.01) decrease in the intervention group (0.26k0.03) when compared with the control group (0.40k0.04). Surrogate data analysis involving the removal of essential nonlinear characteristics from the original data using phase randomisation, revealed that the correlation dimension did not converge over the range of embedding dimensions possible, and that the nonlinear forecasting ability decreased.
Conference proceedings : ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual Conference, Jul 1, 2017
This study was devoted to developing a new auxiliary-model-based damped recursive least squares (... more This study was devoted to developing a new auxiliary-model-based damped recursive least squares (AMB-DRLS) by which the heart rate dynamics can be identified in a real-time manner. Unlike the current conventional schemes for heart rate dynamics modeling, the proposed scheme can simultaneously identify the HR response dynamics and compensate for the existing HR variability while it can also cope with the blowup phenomenon. The performance of the proposed AMB-DRLS scheme was experimentally verified using fifteen healthy male participants who performed treadmill trials following single-cycle square wave protocol. The obtained results revealed a significant difference in goodness-of-fit for the considered parameter estimation schemes. As a result, we conclude that the proposed AMB-DRLS method is able to identify the heart rate response dynamics in a real-time manner while preventing the blowup phenomenon.
When dealing with large-scale systems, manual selection of a subset of components (sensors/actuat... more When dealing with large-scale systems, manual selection of a subset of components (sensors/actuators), or equivalently identification of a favourable structure for the controller, that guarantees a certain closedloop performance, is not very feasible. This paper is dedicated to the problem of concurrent optimal selection of actuators/sensors which can equivalently be considered as the structure identification for the controller. In the context of a multi-channel ℋ 2 dynamic output feedback controller synthesis, we formulate and analyse a framework in which we incorporate two extra terms for penalising the number of actuators and sensors into the variational formulations of controller synthesis problems in order to induce a favourable controller structure. We then develop an explicit scheme as well as an iterative process for the purpose of dealing with the multiobjective problem of controller structure and control law co-design. It is also stressed that the immediate application of the proposed approach lies within the fault accommodation stage of a fault tolerant control scheme. By two numerical examples, we demonstrate the remarkable performance of the proposed approach.
In a home telemonitoring trial, patient adherence with scheduled vital signs measurements is an i... more In a home telemonitoring trial, patient adherence with scheduled vital signs measurements is an important aspect that has not been thoroughly studied and for which data in the literature are limited. Levels of adherence have been reported as varying from approximately 40% to 90%, and in most cases, the adherence rate usually dropped off steadily over time. This drop is more evident in the first few weeks or months after the start. Higher adherence rates have been reported for simple types of monitoring and for shorter periods of intervention. If patients do not follow the intended procedure, poorer results than expected may be achieved. Hence, analyzing factors that can influence patient adherence is of great importance. The goal of the research was to present findings on patient adherence with scheduled vital signs measurements in the recently completed Commonwealth Scientific and Industrial Research Organisation (CSIRO) national trial of home telemonitoring of patients (mean age 7...
This paper reports the development and application of virtual classroom player (VCPlayer), a mult... more This paper reports the development and application of virtual classroom player (VCPlayer), a multimedia educational technology tool for the recording, playback and review of educational sessions in both local and tele-teaching environments. The VCPlayer allows the combination of two media streams synchronously, one displaying the lecture/tutor and one showing the electronic content. VCPlayer-based lectures have been trialled within postgraduate courses at the School of Electrical Engineering. Evaluation has revealed that students show a strong preference for the high-quality, more realistic delivery mode of the VCPlayer over static PowerPoint presentations. Lecturers producing content for the VCPlayer observed that it was as easy to prepare as traditional methods (such as handwritten notes or PowerPoint slides), but found that the combined voice and handwritten annotation gave them substantially better opportunity to communicate complex concepts. Educational benefits revealed during the VCPlayer trial will be discussed in this paper, along with its implementation for various platforms.
We investigate and compare a number of computationally efficient classifiers for categorising the... more We investigate and compare a number of computationally efficient classifiers for categorising the Frank lead ECG as normal or one of six disease conditions using a neural network expert system. These include a power spectral density estimate, and two discrete wavelets, the Daubechies wavelet of order 10 (dblO) and the Symlet wavelet of order 8 (sym8) applied to a single beat of the X, Y and Z Frank leads. Simple statistical parameters derived from these transforms and from reconstructed filtered signals were used as inputs to a neural network with no hidden units and a softmax output stage. We used multiple runs of 10 fold cross validation to obtain estimates of classifier performance. Best results were obtained for the dblO parameters when age and sex were also added. Overall accuracy was 68.8+0.6%. These results are comparable to those derived from neural nets trained with over 229 scalar parameters (70.9&0.6%) and were derived at much lower computational cost. The methods derived can be easily implemented in real time using a DSP processor.
A clinical trial of a home telecare system [1-5] for the management of chronic disease at home on... more A clinical trial of a home telecare system [1-5] for the management of chronic disease at home on 22 patients with a primary diagnosis of Congestive Heart Failure (CHF) and / or Chronic Obstructive Pulmonary Disease (COPD) has demonstrated a high level of patient and doctor satisfaction with respect to usability, functionality and effectiveness. Measurement and medications compliance was high and all patients were able to use the system daily. Although a detailed evaluation of health care outcomes and cost benefits of the clinical trial was beyond the scope of this project, individual case studies suggest that this home telecare system could have a profound impact on the active management of chronic and complex disease at home.
In our laboratory we have developed a rule based expert system for the automated interpretation o... more In our laboratory we have developed a rule based expert system for the automated interpretation of the clinical ECG. Robust algorithms, validated against the CSE ECG databases, have been designed for the reliable determination of QRS complexes, T waves and P waves. Individual wave morphometries are then classified and analysed to produce a matrix of over thirty parameters, which is then passed to a rule based expert system for final interpretation and diagnosis. The performance of this system is now beginning to approach that of the best commercial systems available. Difficulties exist however in the cross validation of rules obtained from a variety of clinical sources and journal publications, even when highly experienced cardiologists are available as consultants. We are thus investigating alternative methods based on Probabilistic Neural Networks. We measure over 180 parameters from the ECG derived from a variety of methods including power spectral density, discrete wavelet decomposition and more classical scalar or vectorcardiographic lead analysis. Receiver operator characteristic (ROC) plots are used to optimise the selection of suitable independent parameters by providing an index of the discriminating ability between alternate categories of a parameter over the complete spectrum of the parameter values. Once an optimal set of independent discriminating parameters are determined for each disease condition, they are used as the input nodes to a neural network. Preliminary results from a database of 500 fully diagnosed ECG records indicates that diagnostic accuracies of 75-80% are achievable. Artificial neural networks, however are relatively new technologies which pose particular challenges regarding the verification and validation of core processing architectures and the neural network topology. Some of these issues are now been considered by the FDA and the TGA in Australia and could result in the imposition of significant constraints on the use of Neural Networks in medical diagnostic devices.
This study investigates the automatic classification of the Frank lead electrocardiogram (ECG) in... more This study investigates the automatic classification of the Frank lead electrocardiogram (ECG) into different pathophysiological disease categories. Coefficients from the discrete wavelet transform are used to represent the ECG diagnostic information and a comparison of the performance of classifiers processing feature sets generated using different mother wavelets is made. Fifteen feature sets are calculated from three Daubechies wavelets, with the decomposition level varied between 3 and 7. The classification performance of each feature set was optimised using automatic feature selection and by combining classifications of multi-beat ECG information. Throughout the study a database of 500 ECG records with examples from seven disease categories was used. The classification of each record is known with 100% confidence and is based on ECG independent information. Using multiple runs of 10-fold cross-validation to obtain all results, it was shown that the overall classification performance of the different feature sets was 71.6-74.2%. In addition, the wavelet order and level had little influence on the overall performance. Analysis of the automatically chosen features reveal that time-frequency bands in the vicinity of the QRS onset and the T-wave are consistently selected.
This case involves a B2B2C model (service provider to community to the elderly) of innovative tel... more This case involves a B2B2C model (service provider to community to the elderly) of innovative tele-care services. The service provider under study is Chu-Shan Show-Chwan Hospital, a district hospital in Nantou County, Taiwan. The community comprises social welfare institutions such as temples, community care centers, and nursing homes. Finally, the elderly individuals who are the ultimate targets of the service are either community residents or residents of nursing homes. The community, as a service promoter, contributes to the increased acceptance of the newly introduced tele-care service compared with the traditional B2C model. Numerous communities under study and some social welfare organizations are extremely interested in providing healthcare services to the elderly and will support future operations provided satisfactory results are achieved. This project can help district hospitals extend their primary care function to communities, build trust, and improve relationships between physicians and patients.
Information and communications technology (ICT) is increasingly being used in management of chron... more Information and communications technology (ICT) is increasingly being used in management of chronic illness to facilitate shared services (virtual health networks and electronic health records), knowledge management (care rules and protocols, scheduling, information directories), as well as consumer-based health education and evidence-based clinical protocols. Common applications of ICT include home monitoring of vital signs for patients with chronic disease, as well as replacing home visits by nurses in person with telemedicine videophone consultations. A patient-managed Home Telecare System with integrated clinical signs monitoring, automated scheduling and medication reminders, as well as access to health education and daily logs, is presented as an example of ICT use for chronic disease self-management. A clinical case study demonstrates how early identification of adverse trends in clinical signs recorded in the home can either avoid hospital readmission or reduce the length of...
International journal of telemedicine and applications, 2013
Despite recent advances in the area of home telemonitoring, the challenge of automatically detect... more Despite recent advances in the area of home telemonitoring, the challenge of automatically detecting the sound signatures of activities of daily living of an elderly patient using nonintrusive and reliable methods remains. This paper investigates the classification of eight typical sounds of daily life from arbitrarily positioned two-microphone sensors under realistic noisy conditions. In particular, the role of several source separation and sound activity detection methods is considered. Evaluations on a new four-microphone database collected under four realistic noise conditions reveal that effective sound activity detection can produce significant gains in classification accuracy and that further gains can be made using source separation methods based on independent component analysis. Encouragingly, the results show that recognition accuracies in the range 70%-100% can be consistently obtained using different microphone-pair positions, under all but the most severe noise conditi...
Proceedings of the 22nd Annual International Conference of the IEEE Engineering in Medicine and Biology Society (Cat. No.00CH37143), 2000
ABSTRACT Previous modelling studied by other investigators indicates that electrotonic coupling b... more ABSTRACT Previous modelling studied by other investigators indicates that electrotonic coupling between cells underlies mutual entrainment and synchronization of the intact sinus node. In this study, interaction between a pair of resistively coupled pacemaker cells was investigated using a complex, physiologically accurate model of the single sinoatrial node cell. Variation in intrinsic cycle length was achieved by multiplying the background sodium conductance gb,Na by a constant modulating parameter (0<α⩽1.0). For each value of the modulating parameter a single cell was integrated until a steady state was achieved, at which point the model parameters were saved for use as initial conditions in the simulations of coupled cell pairs. The effect of cycle length and magnitude of coupling resistance on entrainment were investigated. Simulation results are presented demonstrating both simple and complex entrainment phenomena
Proceedings of the 19th Annual International Conference of the IEEE Engineering in Medicine and Biology Society. 'Magnificent Milestones and Emerging Opportunities in Medical Engineering' (Cat. No.97CH36136)
The relationships between the carotid baroreflex and the nonlinear chaotic dynamics of short-term... more The relationships between the carotid baroreflex and the nonlinear chaotic dynamics of short-term arterial blood pressure control was investigated. Studies in conscious dogs with their baroreceptors either intact (control group: n=7) or surgically isolated (intervention group: n=7), demonstrated deterministic chaotic behavior in mean arterial blood pressure. To assess nonlinear dynamical behavior, we constructed phase-space trajectories from the time series data. The correlation dimension of the reconstructed trajectories was measured using the Grassberger-Procaccia algorithm. A significant increase (PcO.01) in the correlation dimension was measurable between the control (3.46k0.38) and intervention groups (4.59k0.23). Sensitive dependence on initial conditions as determined by the largest Lyaponuv exponents, revealed a significant (Pc0.01) decrease in the intervention group (0.26k0.03) when compared with the control group (0.40k0.04). Surrogate data analysis involving the removal of essential nonlinear characteristics from the original data using phase randomisation, revealed that the correlation dimension did not converge over the range of embedding dimensions possible, and that the nonlinear forecasting ability decreased.
Conference proceedings : ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual Conference, Jul 1, 2017
This study was devoted to developing a new auxiliary-model-based damped recursive least squares (... more This study was devoted to developing a new auxiliary-model-based damped recursive least squares (AMB-DRLS) by which the heart rate dynamics can be identified in a real-time manner. Unlike the current conventional schemes for heart rate dynamics modeling, the proposed scheme can simultaneously identify the HR response dynamics and compensate for the existing HR variability while it can also cope with the blowup phenomenon. The performance of the proposed AMB-DRLS scheme was experimentally verified using fifteen healthy male participants who performed treadmill trials following single-cycle square wave protocol. The obtained results revealed a significant difference in goodness-of-fit for the considered parameter estimation schemes. As a result, we conclude that the proposed AMB-DRLS method is able to identify the heart rate response dynamics in a real-time manner while preventing the blowup phenomenon.
When dealing with large-scale systems, manual selection of a subset of components (sensors/actuat... more When dealing with large-scale systems, manual selection of a subset of components (sensors/actuators), or equivalently identification of a favourable structure for the controller, that guarantees a certain closedloop performance, is not very feasible. This paper is dedicated to the problem of concurrent optimal selection of actuators/sensors which can equivalently be considered as the structure identification for the controller. In the context of a multi-channel ℋ 2 dynamic output feedback controller synthesis, we formulate and analyse a framework in which we incorporate two extra terms for penalising the number of actuators and sensors into the variational formulations of controller synthesis problems in order to induce a favourable controller structure. We then develop an explicit scheme as well as an iterative process for the purpose of dealing with the multiobjective problem of controller structure and control law co-design. It is also stressed that the immediate application of the proposed approach lies within the fault accommodation stage of a fault tolerant control scheme. By two numerical examples, we demonstrate the remarkable performance of the proposed approach.
In a home telemonitoring trial, patient adherence with scheduled vital signs measurements is an i... more In a home telemonitoring trial, patient adherence with scheduled vital signs measurements is an important aspect that has not been thoroughly studied and for which data in the literature are limited. Levels of adherence have been reported as varying from approximately 40% to 90%, and in most cases, the adherence rate usually dropped off steadily over time. This drop is more evident in the first few weeks or months after the start. Higher adherence rates have been reported for simple types of monitoring and for shorter periods of intervention. If patients do not follow the intended procedure, poorer results than expected may be achieved. Hence, analyzing factors that can influence patient adherence is of great importance. The goal of the research was to present findings on patient adherence with scheduled vital signs measurements in the recently completed Commonwealth Scientific and Industrial Research Organisation (CSIRO) national trial of home telemonitoring of patients (mean age 7...
This paper reports the development and application of virtual classroom player (VCPlayer), a mult... more This paper reports the development and application of virtual classroom player (VCPlayer), a multimedia educational technology tool for the recording, playback and review of educational sessions in both local and tele-teaching environments. The VCPlayer allows the combination of two media streams synchronously, one displaying the lecture/tutor and one showing the electronic content. VCPlayer-based lectures have been trialled within postgraduate courses at the School of Electrical Engineering. Evaluation has revealed that students show a strong preference for the high-quality, more realistic delivery mode of the VCPlayer over static PowerPoint presentations. Lecturers producing content for the VCPlayer observed that it was as easy to prepare as traditional methods (such as handwritten notes or PowerPoint slides), but found that the combined voice and handwritten annotation gave them substantially better opportunity to communicate complex concepts. Educational benefits revealed during the VCPlayer trial will be discussed in this paper, along with its implementation for various platforms.
We investigate and compare a number of computationally efficient classifiers for categorising the... more We investigate and compare a number of computationally efficient classifiers for categorising the Frank lead ECG as normal or one of six disease conditions using a neural network expert system. These include a power spectral density estimate, and two discrete wavelets, the Daubechies wavelet of order 10 (dblO) and the Symlet wavelet of order 8 (sym8) applied to a single beat of the X, Y and Z Frank leads. Simple statistical parameters derived from these transforms and from reconstructed filtered signals were used as inputs to a neural network with no hidden units and a softmax output stage. We used multiple runs of 10 fold cross validation to obtain estimates of classifier performance. Best results were obtained for the dblO parameters when age and sex were also added. Overall accuracy was 68.8+0.6%. These results are comparable to those derived from neural nets trained with over 229 scalar parameters (70.9&0.6%) and were derived at much lower computational cost. The methods derived can be easily implemented in real time using a DSP processor.
A clinical trial of a home telecare system [1-5] for the management of chronic disease at home on... more A clinical trial of a home telecare system [1-5] for the management of chronic disease at home on 22 patients with a primary diagnosis of Congestive Heart Failure (CHF) and / or Chronic Obstructive Pulmonary Disease (COPD) has demonstrated a high level of patient and doctor satisfaction with respect to usability, functionality and effectiveness. Measurement and medications compliance was high and all patients were able to use the system daily. Although a detailed evaluation of health care outcomes and cost benefits of the clinical trial was beyond the scope of this project, individual case studies suggest that this home telecare system could have a profound impact on the active management of chronic and complex disease at home.
In our laboratory we have developed a rule based expert system for the automated interpretation o... more In our laboratory we have developed a rule based expert system for the automated interpretation of the clinical ECG. Robust algorithms, validated against the CSE ECG databases, have been designed for the reliable determination of QRS complexes, T waves and P waves. Individual wave morphometries are then classified and analysed to produce a matrix of over thirty parameters, which is then passed to a rule based expert system for final interpretation and diagnosis. The performance of this system is now beginning to approach that of the best commercial systems available. Difficulties exist however in the cross validation of rules obtained from a variety of clinical sources and journal publications, even when highly experienced cardiologists are available as consultants. We are thus investigating alternative methods based on Probabilistic Neural Networks. We measure over 180 parameters from the ECG derived from a variety of methods including power spectral density, discrete wavelet decomposition and more classical scalar or vectorcardiographic lead analysis. Receiver operator characteristic (ROC) plots are used to optimise the selection of suitable independent parameters by providing an index of the discriminating ability between alternate categories of a parameter over the complete spectrum of the parameter values. Once an optimal set of independent discriminating parameters are determined for each disease condition, they are used as the input nodes to a neural network. Preliminary results from a database of 500 fully diagnosed ECG records indicates that diagnostic accuracies of 75-80% are achievable. Artificial neural networks, however are relatively new technologies which pose particular challenges regarding the verification and validation of core processing architectures and the neural network topology. Some of these issues are now been considered by the FDA and the TGA in Australia and could result in the imposition of significant constraints on the use of Neural Networks in medical diagnostic devices.
This study investigates the automatic classification of the Frank lead electrocardiogram (ECG) in... more This study investigates the automatic classification of the Frank lead electrocardiogram (ECG) into different pathophysiological disease categories. Coefficients from the discrete wavelet transform are used to represent the ECG diagnostic information and a comparison of the performance of classifiers processing feature sets generated using different mother wavelets is made. Fifteen feature sets are calculated from three Daubechies wavelets, with the decomposition level varied between 3 and 7. The classification performance of each feature set was optimised using automatic feature selection and by combining classifications of multi-beat ECG information. Throughout the study a database of 500 ECG records with examples from seven disease categories was used. The classification of each record is known with 100% confidence and is based on ECG independent information. Using multiple runs of 10-fold cross-validation to obtain all results, it was shown that the overall classification performance of the different feature sets was 71.6-74.2%. In addition, the wavelet order and level had little influence on the overall performance. Analysis of the automatically chosen features reveal that time-frequency bands in the vicinity of the QRS onset and the T-wave are consistently selected.
This case involves a B2B2C model (service provider to community to the elderly) of innovative tel... more This case involves a B2B2C model (service provider to community to the elderly) of innovative tele-care services. The service provider under study is Chu-Shan Show-Chwan Hospital, a district hospital in Nantou County, Taiwan. The community comprises social welfare institutions such as temples, community care centers, and nursing homes. Finally, the elderly individuals who are the ultimate targets of the service are either community residents or residents of nursing homes. The community, as a service promoter, contributes to the increased acceptance of the newly introduced tele-care service compared with the traditional B2C model. Numerous communities under study and some social welfare organizations are extremely interested in providing healthcare services to the elderly and will support future operations provided satisfactory results are achieved. This project can help district hospitals extend their primary care function to communities, build trust, and improve relationships between physicians and patients.
Information and communications technology (ICT) is increasingly being used in management of chron... more Information and communications technology (ICT) is increasingly being used in management of chronic illness to facilitate shared services (virtual health networks and electronic health records), knowledge management (care rules and protocols, scheduling, information directories), as well as consumer-based health education and evidence-based clinical protocols. Common applications of ICT include home monitoring of vital signs for patients with chronic disease, as well as replacing home visits by nurses in person with telemedicine videophone consultations. A patient-managed Home Telecare System with integrated clinical signs monitoring, automated scheduling and medication reminders, as well as access to health education and daily logs, is presented as an example of ICT use for chronic disease self-management. A clinical case study demonstrates how early identification of adverse trends in clinical signs recorded in the home can either avoid hospital readmission or reduce the length of...
International journal of telemedicine and applications, 2013
Despite recent advances in the area of home telemonitoring, the challenge of automatically detect... more Despite recent advances in the area of home telemonitoring, the challenge of automatically detecting the sound signatures of activities of daily living of an elderly patient using nonintrusive and reliable methods remains. This paper investigates the classification of eight typical sounds of daily life from arbitrarily positioned two-microphone sensors under realistic noisy conditions. In particular, the role of several source separation and sound activity detection methods is considered. Evaluations on a new four-microphone database collected under four realistic noise conditions reveal that effective sound activity detection can produce significant gains in classification accuracy and that further gains can be made using source separation methods based on independent component analysis. Encouragingly, the results show that recognition accuracies in the range 70%-100% can be consistently obtained using different microphone-pair positions, under all but the most severe noise conditi...
Proceedings of the 22nd Annual International Conference of the IEEE Engineering in Medicine and Biology Society (Cat. No.00CH37143), 2000
ABSTRACT Previous modelling studied by other investigators indicates that electrotonic coupling b... more ABSTRACT Previous modelling studied by other investigators indicates that electrotonic coupling between cells underlies mutual entrainment and synchronization of the intact sinus node. In this study, interaction between a pair of resistively coupled pacemaker cells was investigated using a complex, physiologically accurate model of the single sinoatrial node cell. Variation in intrinsic cycle length was achieved by multiplying the background sodium conductance gb,Na by a constant modulating parameter (0<α⩽1.0). For each value of the modulating parameter a single cell was integrated until a steady state was achieved, at which point the model parameters were saved for use as initial conditions in the simulations of coupled cell pairs. The effect of cycle length and magnitude of coupling resistance on entrainment were investigated. Simulation results are presented demonstrating both simple and complex entrainment phenomena
Uploads
Papers by B. Celler