Odor to sensor space transformations in biological and artificial noses

作     者：Pearce, TC 

作者机构：Univ Leicester Dept Engn Leicester LE1 7RH Leics England 

出 版 物：《NEUROCOMPUTING》 (神经计算)

年 卷 期：2000年第32卷

页      面：941-952页

核心收录：

学科分类：08[工学] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：Bass Breweries Ltd Marconi Applied Technologies University of Warwick 

主　　题：olfaction artificial nose population coding 

摘      要：A geometric interpretation of population coding is described which gives an intuitive understanding of how individual sensor/receptor tunings define the overall performance of a sensory system based upon such a code. The approach is demonstrated by applying it to biologically realistic, and synthetic olfactory systems. The analytical measures developed as part of this treatment may be widely applied to optimize the detection performance of artificial electronic nose systems as well as for comparing different tuning scenarios in the biological olfactory pathway. The examples show clearly how a population coded sensory system can deliver both extreme selectivity to compounds of interest without comprising the range of stimuli that may be perceived. (C) 2000 Elsevier Science B.V. All rights reserved.