Persyst Artifact Reduction — detect and reduce artifact with the touch of a button

Persyst Artifact Reduction (AR) works in near-real-time to reduce EMG, Electrode Artifact and Eye Movement. Using Blind Source Separation, Persyst AR selectively reduces these artifacts without attenuating or otherwise altering the underlying Cerebral signal. This results in an artifact-reduced EEG with 40x less cerebral signal distortion than a standard 15Hz high-cut filter.

When reviewing EEG, one of the big issues is the presence of artifacts. The signal from EMG and Eye Movements can often be much larger than the cerebral signal. Additionally, in real-life recordings, electrodes frequently come loose or experience other issues further obscuring or confounding the signal. Persyst AR recognizes artifct from loose electrodes and excludes the relevant channels to allow the reader to clearly see the remaining channels.

Persyst Artifact Reduction reduces artifact in the EEG at the touch of a button without the use of distorting filters. The original EEG is always shown in the background, so the reviewer can see what portions of the signal have been removed. The artifact reduced signal provides a better recording, so the Trending and Spike and Seizure detection algorithms produce more clinically relevant information.

Artifact Reduction is even more important when computing Trends.

The impact of artifacts on quantitative EEG is even more significant. When computing an FFT (CSA) or other Spectral analysis in the conventional manner all of the artifacts are simply included in the calculation. Since in many patients these artifact signals can be very large it can be very difficult to distinguish between changes generated from artifacts and those that are actually cerebral in nature. This dramatically reduces the value of qEEG since true changes can be obscured, and artifacts can create a large number of false positives that have to be eliminated by examining the EEG.

Persyst Trending largely removes this issue with the use of Artifact Reduction. A very significant portion of artifact is removed from the EEG prior to calculation. This means that changes in the Trends generally reflect changes in the cerebral signal.

Artifact Reduction improves detection performance

Artifact Reduction is also a key part of Persyst’s industry leading detection software. By operating on waveforms after artifacts have been removed Persyst is able to make the detectors far more sensitive to changes in the cerebral signal without introducing a high rate of false positives. This is one of the reasons that the Persyst Spike Detector was shown to be non-inferior to skilled human readers (2017, 2018).