The Japanese journal of neuropsychology

Full Text of this Article
in Japanese PDF (597K)

ArticleTitle	A recovery estimate of traumatic brain injury patients of 500 scales based on electroencephalogram
Language	J
AuthorList	Tatsuaki Shirataki, Shin Hibino, Michitaka Matsubara, Hiromi Tatsuda, Hiromi Kishi
Affiliation	Nagoya City Rehabilitation Center
Publication	Japanese Journal of Neuropsychology: 23 (2), 106-112, 2007
Received
Accepted
Abstract	We estimated it about IQ, Miyake-style note signature test, The Rivermead Behavioral Memory Test, PASAT and the index that said from electroencephalograms of 554 traumatic brain injury patients. The method of analysis used a fuzzy neural network (FNN). FNN is technology superior for a prediction and pattern recognition, and even the field of medical care is the technology that begins to be used as clinical invocation means. An estimate and an average error between things of actual survey value were 4.5 points in VIQ, 7.5 points in PIQ, 6.3 points in FIQ and the error ratio was less than 6% both. In addition, survey value were 3.8 points in Miyake-style, 2.6 points in The Rivermead Behavioral Memory Test, 5.6 points in PASAT and the error ratio was around 10% about all. By this method, we were able to estimate the intelligence quotient with enough precision. For a study of a long-term prediction, we must pile up data collection of time to get conclusive evidence of a prediction of the far future and we want to prepare help for estimates.
Keywords	traumatic brain injury, intelligence quotient, electroencephalography, fuzzy neural network, estimate