M. F. Pedersen
Voice Clinic, Ostergade 18,3, DK 1100 Copenhagen, Denmark
It is not only acoustic input into the brain that should be analyzed with neural network but also the resulting voice function. A possibility to get more information is to analyze the voice function in the damaged brain in a co-work between phonetician, phoniatrician, and neuropsychologists. An attempt was made by combining measurement of phonetograms with airflow and stroboscopy. In 22 clients no correlation was found between the damaged frequency and intensity ranges compared with the mean air-flow rates (r=0.15 by multivariate analysis). This is possible because frequency and dynamics are regulated in the acoustic brain and airflow is related to respiration. A discussion should be focused upon multifunctional voice analysis and neural networks in the brain, eventually based on measurements with the multifocal equipment by Key-elemetrics and brain flow measurements [M. F. Pedersen, in Die Krankheit der Stimme, die Stimme der Krankheit, edited by H. Gundermann (Fischer Verlag, New York, 1991), pp. 162--171].