Comparison and validation of visual assessment and image processing algorithms to quantify morphology dynamics of Euglena gracilis.

Image processing algorithms were developed and compared with visual assessment from 12 volunteers to quantify the temporal morphological structure of a single Euglena gracilis organism. Representative images of E. gracilis, showing different morphological characteristics from ovate to cylindrical and elongate, were captured with a bright-field video microscopy system. These images were ranked by the volunteers in order from ovate to elongate. The images were analyzed in the spatial and spatial frequency domain, and the order of the images from each analysis was ranked against the visual assessment. The assessment methods agreeing with the volunteer's preferred sequence were an eccentricity measurement (major axis over the sum of the minor axis at three points), the cross correlation of the image without high pass filtering or edge detection, and cross correlation of the power spectral density.