Analysis on the erythrocyte shape changes using wavelet transforms

Abstract

The mechanics of red blood cell shape changes under normal and deformed conditions are analyzed using wavelet based approach. Images of intact and deformed human red blood cells obtained from normal adults are subject to morphological image processing and the corresponding shape descriptions at two different levels of approximations using different wavelet functions are analyzed. The results demonstrate that using wavelets it is possible to classify normal and deformed red blood cell shapes. Uniform and consistent results are obtained for cells with similar shapes, for all chosen wavelet functions. The variation indices are significant (p<0.005) for all the chosen wavelet functions at both the approximation levels. Further it seems that this approach could be useful for identifying closely identical cell shapes. As cell shape deformations are significant in describing the flow behavior in micro or macro vessels the study seems to be clinically relevant. The methodologies, algorithms and observations based on wavelet based analysis are discussed in detail.