Seizure detection/What is a seizure and when to detect it?

What is a seizure and when to detect it?

Seizures are a symptom associated with abnormal electrical activity in the brain, sometimes described as an electrical storm in the brain or earthquake in the brain (Osorio et al. 2010). Unfortunately, answering these questions remains a challenge for purposes of seizure detection, since there is currently no objective definition of what constitutes a seizure. This significantly hinders the perfection of seizure detection, as it is obviously difficult to develop an algorithm to detect events with perfect precision when the events themselves aren’t objectively identifiable. Strategies to deal with this problem generally fall into two groups: (1) develop an SDA that detects those seizures that have certain characteristics in common, or (2) develop multi-faceted SDAs which attempt to detect all abnormal epileptiform activity present in brain signals, including not just unequivocal seizures but other relevant activity such as brief seizures, bursts, spike trains, and even single spikes, then correlate the occurrence of these detections with clinical events of interest in hopes of providing the most complete information possible about the brain system dynamics underlying seizures for that patient. The first approach has limitations in that seizures outside the target set may go undetected, while the second approach tends to detect both artifacts and other paroxysmal events (both epileptic and non-epileptic) that are not generally considered to be seizures. There is an ongoing debate regarding the dynamical relationship between spikes and seizures (Frei et al. 2010).

Even when the EEG exhibits changes that are considered to be an unequivocal seizure, significant differences (tens of seconds or more) often exist between experts as to when the electrographic onset (EO) and electrographic end (EE) should be marked. Whether or not seizures have clinical/behavioral manifestations and the time when such manifestations first begin, the so-called clinical onset (CO), are determined subjectively as well, since there isn’t always an experienced or trained observer there during the seizure to make the determination, patients are often unaware of their own seizures, and cognitive/functional testing is rarely administered during seizures (Osorio and Frei 2010). Inter-rater variability in EO and EE markings, and difficulty in determining which seizures are “clinical” complicate SDA development and performance assessment, since most applications desire detection of the signal changes that immediately follow EO (which can, e.g., be spike detection in some cases, or a shift in power spectral density in others, depending on what is marked as EO) and sometimes it is desirable to only detect clinical seizures, which may not be differentiable from other “subclinical seizures” (more properly termed “seizures not known to be clinical”) until well after EO. #F3-#F8 provide some illustrative examples.

Figure 1: (A) 60s of single channel ECoG signal containing a seizure and annotations of 3 different patterns that could be used to train or adapt an SDA. The first pattern (“seg 1”) is a quasi-periodic train of spike and slow waves (SSW) with approximately one wave every 2s; the second pattern (“seg 2”) consists of more frequent SSWs occurring approximately 2/s; and the third pattern (“seg 3”) is a higher frequency rhythmic discharge detected by the generic SDA in (Osorio et al. 1998). (B) The “detection ratio” outputs of three SDAs, each adapted to detect one of the three annotated patterns identified in (A). Each output increases above its baseline (with “signal-to-noise ratios” between 100 and 700) temporally correlated with the occurrence of the respective pattern it is trained to detect. Note that the generic algorithm only detects two of the three high frequency bursts present in the signal, as the third doesn’t last long enough (i.e., <1s) for it to be detected without some other SDA adaptation.

Figure 2: When the detector adapted to detect the “seg 1” pattern is applied to a longer recording, the first 30 detections are shown (i.e., 30 ECoG traces, from successive detected events, with detection occurring for each trace at the right margin of the figure). The reader is challenged to consider each trace and see if they can determine differences that might foreshadow whether or not the signal will further evolve to the high frequency rhythmic discharge pattern (“seg 3”) often associated with unequivocal electrographic seizures. (See #F5 for the answer.) A further important challenge, typically requested of a SDA, is to determine by this point in time whether or not the signal change is going to evolve into a seizure with clinical/behavioral manifestations.

Figure 3: The same data as shown in #F4 except the 60s of ECoG following each detection is also shown. Annotations in the right margin indicate the 14 events which evolved to produce generic SDA detections of high frequency rhythmic activity. For this subject, during 85.6 hr of continuous monitoring, 36% of such SSW train detections evolved to generic detections of high frequency rhythmic activity and 18% of those generic SDA detections evolved to become seizures with known clinical/behavioral manifestations. On the other hand, every known clinical seizure was preceded by a generic SDA detection and every such detection was preceded even earlier by a SSW train. This observation calls into question what exactly constitutes “a seizure” and when exactly one begins.

One significant limitation in comparing expert visual analysis (EVA) scores with the results of an SDA is that EVA is typically retrospective – the electroencephalographer identifies that a seizure has occurred and then pages forward and backward through the signal display/printout to set the times at which they believe the event starts and ends. On the other hand, any SDA that operates in real-time is required to issue a decision at a particular time using only signal information available up to that point in time. While some SDAs have been developed that perform retrospectively to first determine a seizure’s presence and then back up to determine when it started (Chan et al. 2008), and these approaches have some utility for offline processing, the associated detection delay limits their usefulness for warning or closed-loop control applications.

Figure 4: (A) 20s of single channel ECoG signal containing a seizure. (B) Magnification of the seizure signal around the electrographic onset. This signal is used to adapt/train an SDA to rapidly detect events with similar power spectral density that is sustained for 2s.

Figure 5: The adapted SDA, trained using the epoch in #F6(B), is applied to a longer epoch of ECoG from the same subject as in #F6 and 15s traces are shown for the first 50 resulting detections. All detections have similar signal characteristics in the two-second window immediately preceding the detections. The reader is challenged to guess which of the detections will evolve to electrographic seizures that last at least 10s. (See #F8 for the answer.)

Figure 6: The same data as shown in #F7 except the 15s of ECoG following each detection is also shown. Only 8 of the 50 detections correspond to electrographic seizure events that last at least 10s.