top of page

Seizure Detection: Human vs Computer

Seizure detection can be done by both computers and people, and both methods have their advantages and limitations.

Computer Seizure Detection

A computerized seizure detection system can be more consistent and reliable than human observation, as it is not subject to fatigue, distraction, or bias. Computerized systems can also analyze large amounts of EEG (electroencephalogram) data quickly and accurately, which can be helpful in detecting seizures that may be difficult for a person to identify.

EEG Persyst is a software system used for EEG analysis in clinical settings. It is designed to help healthcare professionals quickly and accurately interpret EEG data, particularly in cases where the patient has experienced seizures or other abnormal brain activity.

The system uses advanced algorithms and machine learning techniques to automatically detect patterns and abnormalities in EEG data, such as spikes or sharp waves that could indicate seizure activity. It also includes a range of tools for visualizing and analyzing EEG data, including trend analysis, spectral analysis, and statistical comparisons.

EEG spike detection
EEG sample showing showing Persyst XLSpike and XLEvent detection.

Overall, EEG Persyst is designed to help healthcare professionals make faster and more accurate diagnoses based on EEG data, which can be critical for patients with epilepsy or other neurological disorders.

EEG Persyst does have it's downfall as it may overcall rhythmic artifact patterns such as patting, scratching, or shaking a seizure. If someone were to rely primarily on the usage of EEG Persyst they should review the video on all rhythmic events to rule out artifact.

Human Seizure Detection

On the other hand, human observation is often more nuanced and can take into account additional factors such as the person's behavior and response to stimuli. A person may be better able to recognize changes in a person's facial expression, body language, or breathing patterns that could be indicative of a seizure. In addition, human observers can provide emotional support and reassurance to the person experiencing the seizure, which a computerized system cannot do.

Right Frontotemporal EEG sample
Seizure that was detected by the technologist observing patient and EEG change. Persyst did not detect seizure.

Another limitation of computerized seizure detection systems is their tendency to generate false positives or false negatives. False positives occur when the system incorrectly identifies a non-seizure event as a seizure, while false negatives occur when the system fails to identify a seizure. These errors can occur due to various factors such as movement artifacts, electrical interference, or limitations in the algorithm used for detection.

Patient event seizure
The Patient Event button was pressed by on-site staff to alert of patient seizure.

The Patient Event button may be pressed in the patients room or if settings allowed, pressed by staff watching EEG from another location. The Patient Event button is often setup differently at each facility. At my prior Level 4 EMU we had it set that if the there was a "push button" in any of our EMU wired rooms it would send out an alarm to the ascom of every staff member working on our Neuroscience unit. This was very beneficial in patient safety and creating a swift response to our patients. The alert was a huge help if the assigned nurse was attending to other patients and the patient was a safety risk.

In Conclusion

Overall, both computerized seizure detection systems and human observers can play an important role in identifying seizures and providing appropriate care to individuals with epilepsy. However, it is important to consider the advantages and limitations of each method and to use them in conjunction with each other whenever possible.

As a change in EEG can occur and may be detected by the EEG Persyst software, Unfortunately, it can not alert staff and the epileptologist of concerning changes which will lead to faster treatment. Tools such as EEG Persyst are beneficial but highly unlikely to remove the human aspect of EEG reading. It is one of the ways to assist with the national shortage of trained neurodiagnostic technologists. As the field of neurophysiology continues to grow and the work force in unable to keep up these advances in EEG are helpful. Do to the many types of seizures and patterns I do not believe that computer equipment will ever be able to fully replace a human reading EEG.

189 views0 comments

Recent Posts

See All

What to expect when your Neurologist orders an EEG?

I’m an EEG Tech; of course, I know what you’re thinking. Like, “Will this stuff come out of my hair?” or “What is that awful smell?”. As EEG Technologists, we understand there is a lot of anxiety that


bottom of page