Spectral analysis of muscle fiber images as a means of assessing sarcomere heterogeneity.

A new image-analysis-based method is described for assessing sarcomere heterogeneity in skinned rabbit psoas muscle fiber segments. This method consists of off-line, two-dimensional Fourier spectral analysis of video-taped muscle images. Local sarcomere length is assessed by partitioning the muscle images into half and quarter images spanning the original image and analyzing the associated spectra. The spectra are analyzed in two different ways, yielding two measures of sarcomere length. The first measure is obtained by calculating and inverting the centroid frequency of the first-order peak associated with each two-dimensional Fourier spectrum. The second measure is obtained in a similar manner, the only difference being that the two-dimensional spectra are first collapsed into one-dimensional line spectra by summing the pixels perpendicular to the fiber axis. Comparison of the two measures provides a measure of striation skewness that cannot be obtained by other image analysis based methods that determine sarcomere length by analyzing selected line luminance profiles.