Gordon Lightbody | University College Cork (original) (raw)

Uploads

Papers by Gordon Lightbody

Research paper thumbnail of Seizure detection in neonates: Improved classification through supervised adaptation

Annual International Conference of the IEEE Engineering in Medicine and Biology Society, 2008

The goal of neonatal seizure detection is the development of a patient independent system to aler... more The goal of neonatal seizure detection is the development of a patient independent system to alert staff in the neonatal intensive care unit of ongoing seizures. This study demonstrates the potential in adapting a patient independent classifier using patient specific data. Supervised adaptation is investigated using the basic gradient descent algorithm and least mean squares procedures. An increase in mean

Research paper thumbnail of Nonlinear PLS using Radial Basis Functions

Transactions of The Institute of Measurement and Control, 1997

ABSTRACT

Research paper thumbnail of Nonlinear PLS modelling using radial basis functions

American Control Conference, 1997

An approach to nonlinear partial least squares (PLS) modelling using radial basis function (RBF) ... more An approach to nonlinear partial least squares (PLS) modelling using radial basis function (RBF) neural networks to provide a nonlinear inner relationship is described, along with a technique (the hybrid BFGS algorithm) for training the networks. Results are given to show the performance with a number of different simulation examples, including a model of an industrial overheads condenser and reflux

Research paper thumbnail of Development of a Network-Wide Harmonic Control Scheme Using an Active Filter

IEEE Transactions on Power Delivery, 2007

This paper charts the development of a network-wide harmonic control scheme using an active filte... more This paper charts the development of a network-wide harmonic control scheme using an active filter. The individual components required are discussed and developed. By examining the operation of the scheme as a whole, fresh insights are gained into the applicability of previously proposed techniques. The final scheme was shown to systematically reduce harmonic voltage levels across a network to within

Research paper thumbnail of Validation of an Automated Seizure Detection System on Healthy Babies - Histogram-based Energy Normalization for Montage Mismatch Compensation

Seizures in newborn babies are commonly caused by problems such as lack of oxygen, haemorrhage, m... more Seizures in newborn babies are commonly caused by problems such as lack of oxygen, haemorrhage, meningitis, infection and strokes. The aim of an automated neonatal seizure detection system is to assist clinical staff in a neonatal intensive care unit to interpret the EEG. In this work, the automated neonatal seizure detection system is validated on a set of healthy patients

Research paper thumbnail of Classifier models and architectures for EEG-based neonatal seizure detection

Physiological Measurement, 2008

Neonatal seizures are the most common neurological emergency in the neonatal period and are assoc... more Neonatal seizures are the most common neurological emergency in the neonatal period and are associated with a poor long-term outcome. Early detection and treatment may improve prognosis. This paper aims to develop an optimal set of parameters and a comprehensive scheme for patient-independent multi-channel EEG-based neonatal seizure detection. We employed a dataset containing 411 neonatal seizures. The dataset consists of

Research paper thumbnail of Heart rate variability during sleep in healthy term newborns in the early postnatal period

Physiological Measurement, 2009

Normative time- and frequency-domain heart rate variability (HRV) measures were extracted during ... more Normative time- and frequency-domain heart rate variability (HRV) measures were extracted during quiet sleep (QS) and active sleep (AS) periods in 30 healthy babies. All newborn infants studied were less than 12 h old and the sleep state was classified using multi-channel video EEG. Three bands were extracted from the heart rate (HR) spectrum: very low frequency (VLF), 0.01-0.04 Hz;

Research paper thumbnail of A comparison of quantitative EEG features for neonatal seizure detection

Clinical Neurophysiology, 2008

ObjectiveThis study was undertaken to identify the best performing quantitative EEG features for ... more ObjectiveThis study was undertaken to identify the best performing quantitative EEG features for neonatal seizures detection from a test set of 21.

Research paper thumbnail of EEG 'Diarization' for the Description of Neonatal Brain Injures

Research paper thumbnail of All Things Are Known By Comparison: Standardized Performance Assessment and Metrics

Research paper thumbnail of A Data-Driven Energy Based Estimator of EEG Channel Importance for Improved Patient-Adaptive Neonatal Seizure Detector

Proceedings of the 18th IFAC World Congress, 2011

Research paper thumbnail of Validation of an automated seizure detection algorithm for term neonates

Clinical Neurophysiology, 2015

The objective of this study was to validate the performance of a seizure detection algorithm (SDA... more The objective of this study was to validate the performance of a seizure detection algorithm (SDA) developed by our group, on previously unseen, prolonged, unedited EEG recordings from 70 babies from 2 centres. EEGs of 70 babies (35 seizure, 35 non-seizure) were annotated for seizures by experts as the gold standard. The SDA was tested on the EEGs at a range of sensitivity settings. Annotations from the expert and SDA were compared using event and epoch based metrics. The effect of seizure duration on SDA performance was also analysed. Between sensitivity settings of 0.5 and 0.3, the algorithm achieved seizure detection rates of 52.6-75.0%, with false detection (FD) rates of 0.04-0.36FD/h for event based analysis, which was deemed to be acceptable in a clinical environment. Time based comparison of expert and SDA annotations using Cohen's Kappa Index revealed a best performing SDA threshold of 0.4 (Kappa 0.630). The SDA showed improved detection performance with longer seizures. The SDA achieved promising performance and warrants further testing in a live clinical evaluation. The SDA has the potential to improve seizure detection and provide a robust tool for comparing treatment regimens.

Research paper thumbnail of Grading Hypoxic-Ischemic Encephalopathy Severity in Neonatal EEG Using GMM Supervectors and the Support Vector Machine

Clinical Neurophysiology, 2015

This work presents a novel automated system to classify the severity of hypoxic-ischemic encephal... more This work presents a novel automated system to classify the severity of hypoxic-ischemic encephalopathy (HIE) in neonates using EEG. A cross disciplinary method is applied that uses the sequences of short-term features of EEG to grade an hour long recording. Novel post-processing techniques are proposed based on majority voting and probabilistic methods. The proposed system is validated with one-hour-long EEG recordings from 54 full term neonates. An overall accuracy of 87% is achieved. The developed grading system has improved both the accuracy and the confidence/quality of the produced decision. With a new label 'unknown' assigned to the recordings with lower confidence levels an accuracy of 96% is attained. The statistical long-term model based features extracted from the sequences of short-term features has improved the overall accuracy of grading the HIE injury in neonatal EEG. The proposed automated HIE grading system can provide significant assistance to healthcare professionals in assessing the severity of HIE. This represents a practical and user friendly implementation which acts as a decision support system in the clinical environment. Its integration with other EEG analysis algorithms may improve neonatal neurocritical care.

Research paper thumbnail of Power system harmonic analysis using the Kalman filter

2003 IEEE Power Engineering Society General Meeting (IEEE Cat. No.03CH37491), 2003

This paper explores the practical application of the Kalman filter to the analysis of harmonic le... more This paper explores the practical application of the Kalman filter to the analysis of harmonic levels in power systems. The merits and limitations of different possible implementations are investigated and the effect of fundamental frequency variation is examined. The tuning of the Kalman filter for desired dynamic response is discussed and an adaptive tuning algorithm derived for the improved convergence

Research paper thumbnail of Nonlinear dynamic matrix control using local models

Proceedings of the 1998 American Control Conference. ACC (IEEE Cat. No.98CH36207), 1998

This paper proposes the concept of using a local model network (LMN) to identify a highly nonline... more This paper proposes the concept of using a local model network (LMN) to identify a highly nonlinear chemical process, and to implement a dynamic matrix controller (DMC) that uses the local model network as its internal model. The LMN is constructed of local linear autoregressive with external input (ARX) models, and is trained using a hybrid learning approach developed by

Research paper thumbnail of Improving distributed model predictive control performance via weight optimization using PSO

2nd IFAC International Conference on Intelligent Control Systems and Signal Processing (2009), 2009

Research paper thumbnail of Advances in Automated Neonatal Seizure Detection

Studies in Computational Intelligence, 2011

This chapter highlights the current approaches in automated neonatal seizure detection and in par... more This chapter highlights the current approaches in automated neonatal seizure detection and in particular focuses on classifier based methods. Automated detection of neonatal seizures has the potential to greatly improve the outcome of patients in the neonatal intensive care ...

Research paper thumbnail of A Gaussian mixture model based statistical classification system for neonatal seizure detection

2009 IEEE International Workshop on Machine Learning for Signal Processing, 2009

Research paper thumbnail of A comparison of generative and discriminative approaches in automated neonatal seizure detection

2009 IEEE International Symposium on Intelligent Signal Processing, 2009

... was comprised of multichannel video-EEG data from 17 fullterm neonates (ges-tational age rang... more ... was comprised of multichannel video-EEG data from 17 fullterm neonates (ges-tational age range 39-42 weeks) recorded in Cork University Maternity Hospital, Cork, Ireland ... First, the per channel probabilities of seizure are filtered using a 15-epoch central moving average filter ...

Research paper thumbnail of Modulation frequency analysis of seizures in neonatal EEG

2014 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2014

Research paper thumbnail of Seizure detection in neonates: Improved classification through supervised adaptation

Annual International Conference of the IEEE Engineering in Medicine and Biology Society, 2008

The goal of neonatal seizure detection is the development of a patient independent system to aler... more The goal of neonatal seizure detection is the development of a patient independent system to alert staff in the neonatal intensive care unit of ongoing seizures. This study demonstrates the potential in adapting a patient independent classifier using patient specific data. Supervised adaptation is investigated using the basic gradient descent algorithm and least mean squares procedures. An increase in mean

Research paper thumbnail of Nonlinear PLS using Radial Basis Functions

Transactions of The Institute of Measurement and Control, 1997

ABSTRACT

Research paper thumbnail of Nonlinear PLS modelling using radial basis functions

American Control Conference, 1997

An approach to nonlinear partial least squares (PLS) modelling using radial basis function (RBF) ... more An approach to nonlinear partial least squares (PLS) modelling using radial basis function (RBF) neural networks to provide a nonlinear inner relationship is described, along with a technique (the hybrid BFGS algorithm) for training the networks. Results are given to show the performance with a number of different simulation examples, including a model of an industrial overheads condenser and reflux

Research paper thumbnail of Development of a Network-Wide Harmonic Control Scheme Using an Active Filter

IEEE Transactions on Power Delivery, 2007

This paper charts the development of a network-wide harmonic control scheme using an active filte... more This paper charts the development of a network-wide harmonic control scheme using an active filter. The individual components required are discussed and developed. By examining the operation of the scheme as a whole, fresh insights are gained into the applicability of previously proposed techniques. The final scheme was shown to systematically reduce harmonic voltage levels across a network to within

Research paper thumbnail of Validation of an Automated Seizure Detection System on Healthy Babies - Histogram-based Energy Normalization for Montage Mismatch Compensation

Seizures in newborn babies are commonly caused by problems such as lack of oxygen, haemorrhage, m... more Seizures in newborn babies are commonly caused by problems such as lack of oxygen, haemorrhage, meningitis, infection and strokes. The aim of an automated neonatal seizure detection system is to assist clinical staff in a neonatal intensive care unit to interpret the EEG. In this work, the automated neonatal seizure detection system is validated on a set of healthy patients

Research paper thumbnail of Classifier models and architectures for EEG-based neonatal seizure detection

Physiological Measurement, 2008

Neonatal seizures are the most common neurological emergency in the neonatal period and are assoc... more Neonatal seizures are the most common neurological emergency in the neonatal period and are associated with a poor long-term outcome. Early detection and treatment may improve prognosis. This paper aims to develop an optimal set of parameters and a comprehensive scheme for patient-independent multi-channel EEG-based neonatal seizure detection. We employed a dataset containing 411 neonatal seizures. The dataset consists of

Research paper thumbnail of Heart rate variability during sleep in healthy term newborns in the early postnatal period

Physiological Measurement, 2009

Normative time- and frequency-domain heart rate variability (HRV) measures were extracted during ... more Normative time- and frequency-domain heart rate variability (HRV) measures were extracted during quiet sleep (QS) and active sleep (AS) periods in 30 healthy babies. All newborn infants studied were less than 12 h old and the sleep state was classified using multi-channel video EEG. Three bands were extracted from the heart rate (HR) spectrum: very low frequency (VLF), 0.01-0.04 Hz;

Research paper thumbnail of A comparison of quantitative EEG features for neonatal seizure detection

Clinical Neurophysiology, 2008

ObjectiveThis study was undertaken to identify the best performing quantitative EEG features for ... more ObjectiveThis study was undertaken to identify the best performing quantitative EEG features for neonatal seizures detection from a test set of 21.

Research paper thumbnail of EEG 'Diarization' for the Description of Neonatal Brain Injures

Research paper thumbnail of All Things Are Known By Comparison: Standardized Performance Assessment and Metrics

Research paper thumbnail of A Data-Driven Energy Based Estimator of EEG Channel Importance for Improved Patient-Adaptive Neonatal Seizure Detector

Proceedings of the 18th IFAC World Congress, 2011

Research paper thumbnail of Validation of an automated seizure detection algorithm for term neonates

Clinical Neurophysiology, 2015

The objective of this study was to validate the performance of a seizure detection algorithm (SDA... more The objective of this study was to validate the performance of a seizure detection algorithm (SDA) developed by our group, on previously unseen, prolonged, unedited EEG recordings from 70 babies from 2 centres. EEGs of 70 babies (35 seizure, 35 non-seizure) were annotated for seizures by experts as the gold standard. The SDA was tested on the EEGs at a range of sensitivity settings. Annotations from the expert and SDA were compared using event and epoch based metrics. The effect of seizure duration on SDA performance was also analysed. Between sensitivity settings of 0.5 and 0.3, the algorithm achieved seizure detection rates of 52.6-75.0%, with false detection (FD) rates of 0.04-0.36FD/h for event based analysis, which was deemed to be acceptable in a clinical environment. Time based comparison of expert and SDA annotations using Cohen's Kappa Index revealed a best performing SDA threshold of 0.4 (Kappa 0.630). The SDA showed improved detection performance with longer seizures. The SDA achieved promising performance and warrants further testing in a live clinical evaluation. The SDA has the potential to improve seizure detection and provide a robust tool for comparing treatment regimens.

Research paper thumbnail of Grading Hypoxic-Ischemic Encephalopathy Severity in Neonatal EEG Using GMM Supervectors and the Support Vector Machine

Clinical Neurophysiology, 2015

This work presents a novel automated system to classify the severity of hypoxic-ischemic encephal... more This work presents a novel automated system to classify the severity of hypoxic-ischemic encephalopathy (HIE) in neonates using EEG. A cross disciplinary method is applied that uses the sequences of short-term features of EEG to grade an hour long recording. Novel post-processing techniques are proposed based on majority voting and probabilistic methods. The proposed system is validated with one-hour-long EEG recordings from 54 full term neonates. An overall accuracy of 87% is achieved. The developed grading system has improved both the accuracy and the confidence/quality of the produced decision. With a new label 'unknown' assigned to the recordings with lower confidence levels an accuracy of 96% is attained. The statistical long-term model based features extracted from the sequences of short-term features has improved the overall accuracy of grading the HIE injury in neonatal EEG. The proposed automated HIE grading system can provide significant assistance to healthcare professionals in assessing the severity of HIE. This represents a practical and user friendly implementation which acts as a decision support system in the clinical environment. Its integration with other EEG analysis algorithms may improve neonatal neurocritical care.

Research paper thumbnail of Power system harmonic analysis using the Kalman filter

2003 IEEE Power Engineering Society General Meeting (IEEE Cat. No.03CH37491), 2003

This paper explores the practical application of the Kalman filter to the analysis of harmonic le... more This paper explores the practical application of the Kalman filter to the analysis of harmonic levels in power systems. The merits and limitations of different possible implementations are investigated and the effect of fundamental frequency variation is examined. The tuning of the Kalman filter for desired dynamic response is discussed and an adaptive tuning algorithm derived for the improved convergence

Research paper thumbnail of Nonlinear dynamic matrix control using local models

Proceedings of the 1998 American Control Conference. ACC (IEEE Cat. No.98CH36207), 1998

This paper proposes the concept of using a local model network (LMN) to identify a highly nonline... more This paper proposes the concept of using a local model network (LMN) to identify a highly nonlinear chemical process, and to implement a dynamic matrix controller (DMC) that uses the local model network as its internal model. The LMN is constructed of local linear autoregressive with external input (ARX) models, and is trained using a hybrid learning approach developed by

Research paper thumbnail of Improving distributed model predictive control performance via weight optimization using PSO

2nd IFAC International Conference on Intelligent Control Systems and Signal Processing (2009), 2009

Research paper thumbnail of Advances in Automated Neonatal Seizure Detection

Studies in Computational Intelligence, 2011

This chapter highlights the current approaches in automated neonatal seizure detection and in par... more This chapter highlights the current approaches in automated neonatal seizure detection and in particular focuses on classifier based methods. Automated detection of neonatal seizures has the potential to greatly improve the outcome of patients in the neonatal intensive care ...

Research paper thumbnail of A Gaussian mixture model based statistical classification system for neonatal seizure detection

2009 IEEE International Workshop on Machine Learning for Signal Processing, 2009

Research paper thumbnail of A comparison of generative and discriminative approaches in automated neonatal seizure detection

2009 IEEE International Symposium on Intelligent Signal Processing, 2009

... was comprised of multichannel video-EEG data from 17 fullterm neonates (ges-tational age rang... more ... was comprised of multichannel video-EEG data from 17 fullterm neonates (ges-tational age range 39-42 weeks) recorded in Cork University Maternity Hospital, Cork, Ireland ... First, the per channel probabilities of seizure are filtered using a 15-epoch central moving average filter ...

Research paper thumbnail of Modulation frequency analysis of seizures in neonatal EEG

2014 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2014