Title :
Feature extraction and pattern classification of hormonal time series
Author :
Wang, T.P. ; Vagnucci, A.H. ; Pratt, V. ; Li, C.C.
Author_Institution :
Pittsburgh Univ., PA, USA
Abstract :
A pattern recognition system for computer classification of cortisol time series into classes of Cushing´s syndrome with different etiology or as normal is described. Discriminatory features are selected from the Fourier analysis and Karhunen-Loeve expansion coefficients of cortisol time series. The performance of the cortisol pattern recognition system as tested on 86 sample patterns, 41 normals and 45 patients is summarized. The recognition accuracy is 100% for the normal, pituitary, and adrenal classes, but only 60% for the ectopic class
Keywords :
computerised pattern recognition; medical diagnostic computing; Cushing´s syndrome; Fourier analysis; Karhunen-Loeve expansion coefficients; adrenal class; computer classification; cortisol time series; discriminatory features; ectopic class; feature extraction; hormonal time series; normals; patients; pattern classification; pituitary class; recognition accuracy; Artificial intelligence; Feature extraction; Hospitals; Medical diagnostic imaging; Neoplasms; Pattern classification; Pattern recognition; Plasmas; Testing; Time series analysis;
Conference_Titel :
Engineering in Medicine and Biology Society, 1989. Images of the Twenty-First Century., Proceedings of the Annual International Conference of the IEEE Engineering in
Conference_Location :
Seattle, WA
DOI :
10.1109/IEMBS.1989.95965
