Skip to main content
Figure 4 | BioMedical Engineering OnLine

Figure 4

From: Wavelet-based Gaussian-mixture hidden Markov model for the detection of multistage seizure dynamics: A proof-of-concept study

Figure 4

The transition diagram and the posterior probabilities for HMM opt7D using the original feature space. (a) A unidirectional sequence of state progress was found for the HMMopt7D. Even though the validation result indicated that a five state model should be used, S 1 and S 5 can be consolidated to represent the postictal events after expert guided state assignment. (b) The mapping of HMMopt7D output to the electrophysiological data is given. The marginal posterior probabilities illustrate that the HMMopt7D is able to detect the non-ictal, tonic firing, chronic seizures and postictal activities.

Back to article page