Spike Review summarizes detections
While the Spike Detection and Density Trend shows the presence of spikes throughout the recording over time, reviewing spike detections is still quite burdensome if thousands of spikes need to be inspected and mentally classified.
Persyst Spike Review alleviates this burden through automatically sorting spikes according to their focus and then then providing an overview to the user. The Overview window depicts averages from the various spike foci detected by Persyst’s Spike Detector. To create these overview averages the spike detections are sorted by detection foci (electrode) and then all detections at a particular focus are mathematically averaged. For example, in the figure below, the first column of EEG represents an average of 1064 events that had their maximum point of detection at the F7 electrode. Notice that columns of EEG are separated from other columns by a thin band of white. Each EEG column represents a distinct group average. The primary electrode focal point of each average, and the number of detection events incorporated into each average, are shown above the columns of EEG. Clicking on any of the electrode location tabs at the top of the EEG window will display the raw (non-averaged) spike detections that arose from that particular electrode location. This feature allows users to select which individuals to keep within a spike group for further analysis.
Dynamic adjustment of detection sensitivity
The sensitivity of the Spike Detector output can be dynamically adjusted during the review process. When Spike Review is initially opened, the detection sensitivity is set to Medium, which is the default. In this position the Spike Detector neural network algorithms identify sharp transients that have a high probability of being epileptiform abnormalities.
However, some spikes and sharp waves that are less well-formed may not be evident at Medium sensitivity. The detector’s sensitivity can be quickly adjusted on the fly. When set to High, Spike Review shows the results of the detector at this more sensitive setting, which is more likely to identify less well-formed or lower amplitude transients. New groups may then appear in the Overview display of spike averages. In concert with the increase in true spike detections, there is also an increase in false positive detections.
