Automatic morphometry of synaptic boutons of cultured cells using granulometric analysis of digital images

Numbers, linear density, and surface area of synaptic boutons can be important parameters in studies on synaptic plasticity in cultured neurons. We present a method for automatic identification and morphometry of boutons based on filtering of digital images using granulometric analysis. Cultures of cortical neurons (DIV8 and DIV21) were fixed and marked with fluorescently labeled antibodies for synapsin I (a marker for synaptic boutons) and MAP-2 (a marker for dendrites). Images were acquired on a confocal microscope and automatically processed. Granulometry, a morphological operator sensitive to the geometry and size of objects, was used to construct a filter passing fuzzy fluorescent grains of a certain size. Next, the filter was overlaid with the original image (masking) and the positive pixels were identified by an integral intensity threshold (thresholding). Disjoint grains, representing individual boutons, were reconstructed from the connected pixels above the threshold, numbered and their area was measured. In total, 1498 boutons with a mean diameter of 1.63 ± 0.49 μm (S.D.) were measured. Comparisons with manual counts showed that the proposed method was capable of identifying boutons in a systematic manner at the light microscopic level and was a viable alternative to manual bouton counting.