Active contour models for segmenting hyphal images

The aim is to characterise the growth and morphology of arbuscular mycorrhizal hyphal distributions (i.e. spatial distributions of hyphae from a type of fungus), in order to enable assessment of the efficiency of hyphal colonisation of a growth medium relative to localised variations in nutrient availability. Patterns of distribution are likely to have a profound influence on nutrient acquisition and on the extent of relocation of nutrients within a medium. Features of interest include length and extension rates of individual hyphae, total length, number of branches, specific growth and branching rates, and fractal dimension. Although software exists to measure the hyphae, once extracted manually from an image, segmenting each hypha for measurement purposes is inefficient, inaccurate, tedious and time consuming. The dataset of images available show the total growth of a hyphal network from a section of root material for days 2 to 11. Segmentation is complicated by background streaking in the image, caused by striations on the membrane upon which the hyphae are grown, faintness of some sections of the hyphae, caused by inconsistent moisture levels on the membrane, and parts of the image which are out of focus at the moment of image capture.