An interactive framework for reconstructing 3D neuronal structures.

Kanuj Kumar, Vijay Natarajan, S. K. Sikdar and Kalyan V. Srinivas.
ICVGIP 2014: Proc. Indian Conference on Computer Vision, Graphics and Image Processing, 2014, 19:1-19:8.

Abstract

Neuron reconstruction is a complex and tedious task and neurobiologists today have to rely on time-consuming manual methods. Methods that entail manual tracing may introduce systematic inaccuracies. This necessitates the use of automated methods for reconstruction of neuronal structures. Despite recent advancements, the automation of reconstruction process either does not exhibit robustness against noise of microscopy images or fails to capture precise dendritic structures. This motivates the development of semi-automated methods that require crucial but minimal expert user input. In this paper, we present a fast and interactive framework for reconstruction of neuronal structures that employs automatic methods while allowing a user to provide expert input if the results are unsatisfactory. The framework is designed as a multi-stage pipeline, where in the user provides numerical input parameters to guide the reconstruction process and validates the output of automated methods visually. The user is also assisted by the framework in calculating the optimal values of required parameters for reconstruction. The framework is also able to handle discontinuities and produce a connected geometric model of the neuron structure.

Download

Supplementary Material:

  1. Poster (Download)
  2. Demonstartion video (Download)