Spike detection: a review and comparison of algorithms

Wearable Electroencephalography

“…in the same EEG recording, they will often mark noticeably different sections, with agreement ranging from 0 to 90% of the time [53]. This can affect the reported sensitivity of the monitoring scheme as any two experts will disagree over its performance.…”

Cluster-based spike detection algorithm adapts to interpatient and intrapatient variation in spike morphology

“…In order to do so, the average of all pairwise expert vs. expert sensitivity and selectivity values, and the average of all pairwise algorithm vs. expert sensitivity and selectivity values were compared, as proposed by Wilson and Emerson (2002). The sensitivity of one reader (A) versus another (B) is defined as:…”

“…Many automated spike detection algorithms have been published (as shown in Frost (1985), Wilson and Emerson (2002) and Halford (2009) for review). Algorithms are typically based on methods such as template matching (Ji et al, 2011), mimetic analysis (Exarchos et al, 2006), power spectral analysis (Sugi et al, 2002;Adjouadi et al, 2004;Exarchos et al, 2006), wavelet analysis (Adeli et al, 2003;Indiradevi et al, 2008) and artificial neural networks (Guler and Ubeyli, 2005;Tzallas et al, 2006).…”

“…One of the major problems of automatic spike detection is that both spike morphology and background vary widely between patients (interpatient variation) (Wilson and Emerson, 2002) as well as within a single recording (intrapatient variation) (Wahlberg and Lantz, 2000). This is why spike detection algorithms usually give better performance when the EEG expert is asked to select samples of spikes in order to tailor the algorithm to the patient (Sankar and Natour, 1992;Aarabi et al, 2009).…”

A Multistage System for Automatic Detection of Epileptic Spikes

“…Much effort has been spent on spike detection over the last 40 years. The reader is referred to excellent reviews on the topic that can be consulted in [2][3][4].…”