Automatic 3D Neuron Tracing Based on Terminations Detection

Abstract

The digital reconstruction of neuron morphology in volumetric images is important for understanding the connectivity of neurons system. In this paper, we developed an automatic neuron reconstruction (tracing) framework based on neuron termination points. The neuron terminations are detected by an adaptive ray-shooting model which analyzes the intensity distribution of the neighborhood around the termination point candidates. Based on the detected neuron terminations, a graph-augmented deformable model based on the shortest path rule is employed to reconstruct the 3D neuron structure. The experimental results on multiple 3D neuron images validated the effectiveness of the proposed method to reconstruct the neuron structures.