Novel Features to Capture Temporal Variations of Rhythmic Limb Movement to Distinguish Convulsive Epileptic and Psychogenic Non -epileptic Seizures

Objective: Diagnosis of convulsive psychogenic nonepileptic seizures (PNES) pose a challenge due to their clin ical similarities with generalized tonic-clonic seizures. Thedelay in correct diagnosis of PNES increases the vulnerability to si deeffects of anti epileptic drugs and imposes physical, socia l, and economic burden on patients with PNES. Definite diagnosis of PNES require long term video-electroencephalography moni toring (VEM). However, VEM requires significant system resources and thus, arises the need for alternative methods for PNES diagnosis. In this study, we aim to investigate the potentia l of Poincaré derived temporal variations in rhythmic limb movement recorded using a wrist-worn accelerometer based device for differentiating convulsive ES and PNES. Methods: The temporal variations in the accelerometer traces corresponding to 39 generalized tonic-clonic seizures (GTCS) and 44 convulsive PNES events are obtained using Poincar é maps. Poincaŕe maps are geometrical representation of a time series signal, when plotted in Cartesian plane. The two prop osed indexes: tonic index (TI) and dispersion decay index ( DDI ) are calculated to quantify the temporal variations derived usi ng Poincaré maps. Key Findings: The TI captures the presence of a tonic phase in an event. Whereas, theDDI captures the subsiding behavior of an event.TI and DDI of GTCS events was higher in comparison to convulsive PNES events ( p< 0.001). A maximum AUC (area under ROC curve) of 0.96 was obtained for GTCS and PNES differentiation using TI alone. A linear discriminant classifier build using a combination of TI and DDI of all Poincaré derived descriptors could correctly differentiate 42 (sensitivity: 95.45%) of 44PNES events and36 (specificity: 92.30%) of 39GTCS events. A blinded review of the Poincaŕe derived temporal variations in rhythmic limb movement during seizures could correctly differentiate 26 (sensitivity: 70.27%) of 44 PNES events and 33 (specificity: 86.84%) of 39 GTCS events. In comparison to the proposed method, the coefficient of variation (COV) of li mb movement frequency in time windows of2.56s resulted in an AUC value of 0.78. Significance: Temporal variations in rhythmic limb movement S. Kusmakar and M. Palaniswami are with the Department of Ele ctrical and Electronic Engineering, The University of Melbourne, Vic 3052, Australia. Email: skusmakar@student.unimelb.edu.au; palani@unimelb.edu .au C. K. Karmakar is with the Department of Electrical and Elect ronic Engineering, The University of Melbourne, Vic 3052, Austr alia and also with Deakin University Geelong, Vic 3125, Australia, Email: karmakar@unimelb.edu.au R. Muthuganapathy is with the Department of Engineering Des ign, Indian Institute of Technology Madras, India. Email: mraman@iitm.ac.in B. Yan, Patrick kwan, and T. O’Brien are with the Melbourne Br ain Centre, Royal Melbourne Hospital, Dept. of Medicine, The Un iversity of Melbourne, VIC 3052, Australia. Email: Bernard.Yan@mh.org.au, obrientj@unimelb.edu.au can be used to distinguish generalized tonic-clonic seizur es (GTCS) from convulsive psychogenic non-epileptic seizure s (PNES). Thus, enabling real-time monitoring and differential diagnosis of convulsive PNES, using a wrist-worn accelerom eter based biosensor.

[1]  Chandan Karmakar,et al.  Novel features for capturing temporal variations of rhythmic limb movement to distinguish convulsive epileptic and psychogenic nonepileptic seizures , 2018, Epilepsia.

[2]  T. Loddenkemper,et al.  Automated seizure detection systems and their effectiveness for each type of seizure , 2016, Seizure.

[3]  Roland D. Thijs,et al.  Non-EEG based ambulatory seizure detection designed for home use: What is available and how will it influence epilepsy care? , 2016, Epilepsy & Behavior.

[4]  Tobias Loddenkemper,et al.  Seizure detection, seizure prediction, and closed-loop warning systems in epilepsy , 2014, Epilepsy & Behavior.

[5]  Poul Jennum,et al.  Quantitative analysis of surface electromyography during epileptic and nonepileptic convulsive seizures , 2014, Epilepsia.

[6]  Markus Reuber,et al.  Minimum requirements for the diagnosis of psychogenic nonepileptic seizures: A staged approach , 2013, Epilepsia.

[7]  Bernard Yan,et al.  Time‐frequency mapping of the rhythmic limb movements distinguishes convulsive epileptic from psychogenic nonepileptic seizures , 2013, Epilepsia.

[8]  Phillip E M Smith.  Epilepsy: mimics, borderland and chameleons , 2012, Practical Neurology.

[9]  Jefferson Luiz Brum Marques,et al.  Comparison of heart rate variability parameters during complex partial seizures and psychogenic nonepileptic seizures , 2012, Epilepsia.

[10]  Lojini Logesparan,et al.  Optimal features for online seizure detection , 2012, Medical & Biological Engineering & Computing.

[11]  Elske Ammenwerth,et al.  Measurement and quantification of generalized tonic–clonic seizures in epilepsy patients by means of accelerometry—An explorative study , 2011, Epilepsy Research.

[12]  Orrin Devinsky,et al.  Differentiating between nonepileptic and epileptic seizures , 2011, Nature Reviews Neurology.

[13]  Chih-Jen Lin,et al.  LIBSVM: A library for support vector machines , 2011, TIST.

[14]  Marimuthu Palaniswami,et al.  Analyzing temporal variability of standard descriptors of Poincaré plots. , 2010, Journal of electrocardiology.

[15]  A. Schulze-Bonhage,et al.  Views of patients with epilepsy on seizure prediction devices , 2010, Epilepsy & Behavior.

[16]  Dennis Velakoulis,et al.  Clinical Characteristics and Outcome in Patients With Psychogenic Nonepileptic Seizures , 2010, Psychosomatic medicine.

[17]  Sándor Beniczky,et al.  Diagnostic usefulness and duration of the inpatient long-term video-EEG monitoring: Findings in patients extensively investigated before the monitoring , 2009, Seizure.

[18]  Markus Reuber,et al.  Psychogenic nonepileptic seizures: Answers and questions , 2008, Epilepsy & Behavior.

[19]  Simon Vogrin,et al.  “Convulsive” Nonepileptic Seizures Have a Characteristic Pattern of Rhythmic Artifact Distinguishing Them from Convulsive Epileptic Seizures , 2004, Epilepsia.

[20]  Terence J O'Brien,et al.  Evaluating the Utility of Inpatient Video‐EEG Monitoring , 2004, Epilepsia.

[21]  G. Baker,et al.  Failure to recognize psychogenic nonepileptic seizures may cause death , 2004, Neurology.

[22]  A. House,et al.  Somatization, dissociation and general psychopathology in patients with psychogenic non-epileptic seizures , 2003, Epilepsy Research.

[23]  C. Elger,et al.  Outcome in psychogenic nonepileptic seizures: 1 to 10‐year follow‐up in 164 patients , 2003, Annals of neurology.

[24]  W. LaFrance,et al.  Treatment of nonepileptic seizures , 2002, Epilepsy & Behavior.

[25]  M. Reuber,et al.  Diagnostic delay in psychogenic nonepileptic seizures , 2002, Neurology.

[26]  Marimuthu Palaniswami,et al.  Do existing measures of Poincare plot geometry reflect nonlinear features of heart rate variability? , 2001, IEEE Transactions on Biomedical Engineering.

[27]  W. Hauser,et al.  An estimate of the prevalence of psychogenic non-epileptic seizures , 2000, Seizure.

[28]  Sukhan Lee,et al.  Micromachined inertial sensors , 1999, Proceedings 1999 IEEE/RSJ International Conference on Intelligent Robots and Systems. Human and Environment Friendly Robots with High Intelligence and Emotional Quotients (Cat. No.99CH36289).

[29]  W H Theodore,et al.  The secondarily generalized tonic‐clonic seizure , 1994, Neurology.

[30]  J. Hanley,et al.  A method of comparing the areas under receiver operating characteristic curves derived from the same cases. , 1983, Radiology.

[31]  Marimuthu Palaniswami,et al.  Poincaré Plot Methods for Heart Rate Variability Analysis , 2013, Springer US.

[32]  Christopher G. Wilson,et al.  Periodicity, mixed-mode oscillations, and quasiperiodicity in a rhythm-generating neural network. , 2002, Biophysical journal.