Automated Seizure Detection Using Ekg

Abstract

Changes in heart rate, most often increases, are associated with the onset of epileptic seizures and may be used in lieu of cortical activity for automated seizure detection. The feasibility of this aim was tested on 241 clinical seizures from 81 subjects admitted to several Epilepsy Centers for invasive monitoring for evaluation for epilepsy surgery. The performance of the EKG-based seizure detection algorithm was compared to that of a validated algorithm applied to electrocorticogram (ECoG). With the most sensitive detection settings [threshold T: 1.15; duration D: 0 s], 5/241 seizures (2%) were undetected (false negatives) and with the highest [T: 1.3; D: 5 s] settings, the number of false negative detections rose to 34 (14%). The rate of potential false positive (PFP) detections was 9.5/h with the lowest and 1.1/h with the highest T, D settings. Visual review of 336 ECoG segments associated with PFPs revealed that 120 (36%) were associated with seizures, 127 (38%) with bursts of epileptiform discharges and only 87 (26%) were true false positives. Electrocardiographic (EKG)-based seizure onset detection preceded clinical onset by 0.8 s with the lowest and followed it by 13.8 s with the highest T, D settings. Automated EKG-based seizure detection is feasible and has potential clinical utility given its ease of acquisition, processing, high signal/noise and ergonomic advantages viz-a-viz EEG (electroencephalogram) or ECoG. Its use as an "electronic" seizure diary will remedy in part, the inaccuracies of those generated by patients/care-givers in a cost-effective manner.

Probability of detection of clinical seizures using heart rate changes

“…Automated detection of epileptic seizures originating from or spreading to central autonomic network (CAN) structures [1][2][3][4][5][6] using heart rate changes [7][8][9][10][11][12][13][14][15], is gaining acceptance as either a valuable complement or as an alternative to cortical electrical signals (e.g., ECoG) due to their: (a) Higher signal-to-noise ratio (mV for EKG vs. mV ECoG; (b) Greater ease and economy of recording (EKG requires 2 electrodes for indirect but all encompassing monitoring of ictal activity in CAN structures, while the same number of scalp or intracranial electrodes would yield inadequate coverage [16]; (c) Lower processing and computational analysis cost [13] due to EKG's lower signal complexity than ECoG.…”

“…The magnitude of relative changes in heart rate at seizure onset, is the single most important biological factor governing the probability of detection using this variable [13], which is subject not only to fluctuations in type and level of physical activity, but also to various other factors. The degree to which age, gender, seizure class (e.g., simple vs. complex partial), etiology (e.g., lesional vs. cryptogenic) or hemispheric location (left vs. right) of the epileptogenic zone may impact the probability of clinical seizure detection is unknown.…”

Computer-Aided Diagnosis of Depression Using EEG Signals

“…Even though a number of papers have been published using the nonlinear methods, there are other nonlinear methods [48,49,50,51,52,53,54,55,56,57,58,59,60,87,88,89,90,91,92,93,94,95,96,97,98] that are worth exploring for the EEG-based diagnosis of depression. As an example, figures 3a and b show sample bispectrum magnitude plots of EEG signals from the left brain hemisphere for normal and depression subjects shown in figure 2, respectively.…”

Adaptive heart rate-based epileptic seizure detection using real-time user feedback

“…Most seizure detection algorithms using heart rate (and in general using all modalities) from literature use so-called patient-independent classifiers (De Cooman et al 2017, Osorio 2014, Milosevic et al 2016, Poh et al 2012). They do not require patient-specific data, making them directly usable in practice.…”