One step forward, several more to go: classification of psychogenic non-epileptic seizures based on automatic clustering
- 1University of Sheffield, Royal Hallamshire Hospital, Sheffield, UK
- 2West of Scotland Regional Epilepsy Service, Southern General Hospital, Glasgow, UK
- Correspondence to Dr Markus Reuber, University of Sheffield, Royal Hallamshire Hospital, Glossop Road, Sheffield S10 2JF, UK; markus.reuber{at}sth.nhs.uk
- Received 4 April 2011
- Accepted 25 April 2011
- Published Online First 25 May 2011
In ordinary neurological practice, the diagnosis of psychogenic non-epileptic seizures (PNES) is often made by exclusion, or on the grounds that seizures lack features of epilepsy. With the exception of linguistic and interactional markers in the conversation with the patient about their symptoms,1 few ‘positive’ signs of PNES have been described. Therefore, this effort to improve our knowledge of the clinical semiology of PNES is welcome. Hubsch et al2 used multiple correspondence analysis and cluster analysis of video-EEG recordings of PNES to identify five clinical subtypes of attack. The authors conclude that their classification of PNES may provide useful criteria for clinical diagnosis, presumably by giving clinicians a series of ‘models’ they can hold in their heads for evaluation of clinical …
