Resumo

PAREDES, José L  e  SOSA, Enedina. A baseline correction algorithm for capillary electrophoresis data using local optimization of the legend algorithm in the wavelet domain. INCI [online]. 2009, vol.34, n.8, pp.556-562. ISSN 0378-1844.

A baseline (BL) correction algorithm for capillary electrophoresis (CE) data is developed. The proposed algorithm exploits a low-resolution wavelet representation of the original signal to locally model the BL as a polinomial function and applies the LEGEND optimization algorithm to obtain the model parameters, such that a robust non-symmetric cost function is minimized. The proposed algorithm outputs a suitable BL correction on those subsections of the electropherogram where peaks related to low concentration substances are near those of high concentration substances. The performance of the proposed algorithm was tested by measuring the glutamate mass on 24 electropherograms after the BL had been suppressed using the proposed algorithm. The resulting values are compared to those yielded by a conventional measuring method performed by a CE specialist and by a second CE baseline correction method recently introduced.

Palavras-chave : Algoritmo LEGEND; Corrección de Línea Base; Electroforesis Capilar; Transformada Wavelet .