Ion mobility spectrometry (IMS) technique that carries out ionization and ion-species drift measurements at atmospheric pressure is used for the detection of the lung cancer at early stages. The species required for diagnostic support are only present in very low quantities so robust data pre-processing is needed to the evaluation stage. A discrete wavelet transformation is applied which clearly emphasizes the true signals within the data set and an additional noise reduction step using log normal description provides much clearer data. With the reduction in noise, the next data processing requirement is for an improvement of the individual components within the matrices against noise generated from false signals. Discriminant analysis techniques applied to the reduced data have proven to be effective in the diagnosis of lung cancer.
|Number of pages||4|
|Publication status||Published - Oct 2008|