A background elimination method based on wavelet transform in wound infection detection by electronic nose

When mice are used as experimental subjects in the detection of wound infection based on electronic nose (enose), the background, i.e., the smell of the mice themselves, is very strong, and most useful information is buried in it. A new method is proposed to eliminate the background and discriminate wound infection based on a gas sensor array composed of 15 gas sensors. It employs thirteen-scale and the first order Daubechies (db1) wavelet analysis to decompose each signal of the sensor array. Direct multiplication of wavelet transform coefficients at corresponding scales between response signals of the wounded and healthy mice are used to eliminate background smell. The approximation coefficients of which the background had been eliminated are used as the inputs of RBF (Radical Basis Function) network for discrimination. The result shows that this method is effective and practical for background elimination in the detection of wound infection. Besides, this method is also useful in dimensionality reduction.