Dendritic spines detection based on directional morphological filter and shortest path

Dendritic spines are membranous protrusions from neuron´s dendrites. They play a very important role in the nervous system. They are very small and have various shapes; hence it is very challenging to detect them in neuron images. This paper presents a novel method for detecting dendritc spines in 2D images. A new dendrite backbone or centerline extraction method is introduced herein which is based on an iterative process between smoothing and extraction. The proposed method iteratively refines the extraction result using both directional morphological filtering and improved Hessian filtering until a satisfactory extraction result is obtained. A shortest path method is applied along a backbone to extract the boundary of the dendrites. Spines are then segmented from the dendrites outside the extracted boundary. The proposed algorithm has been tested with many images and good results are achieved.