Three-dimensional Reconstruction of Peripheral Nerve Internal Fascicular Groups.

Yingchun Zhong,Jian Qi,Xiaolin Liu,Yi Zhang,Peng Luo,Cory J Xian,Jianghui Dong,Liping Wang

doi:10.1038/srep17168

Abstract

Peripheral nerves are important pathways for receiving afferent sensory impulses and sending out efferent motor instructions, as carried out by sensory nerve fibers and motor nerve fibers. It has remained a great challenge to functionally reconnect nerve internal fiber bundles (or fascicles) in nerve repair. One possible solution may be to establish a 3D nerve fascicle visualization system. This study described the key technology of 3D peripheral nerve fascicle reconstruction. Firstly, fixed nerve segments were embedded with position lines, cryostat-sectioned continuously, stained and imaged histologically. Position line cross-sections were identified using a trained support vector machine method, and the coordinates of their central pixels were obtained. Then, nerve section images were registered using the bilinear method, and edges of fascicles were extracted using an improved gradient vector flow snake method. Subsequently, fascicle types were identified automatically using the multi-directional gradient and second-order gradient method. Finally, a 3D virtual model of internal fascicles was obtained after section images were processed. This technique was successfully applied for 3D reconstruction for the median nerve of the hand-wrist and cubital fossa regions and the gastrocnemius nerve. This nerve internal fascicle 3D reconstruction technology would be helpful for aiding peripheral nerve repair and virtual surgery.

Highlights

Peripheral nerves are important pathways for receiving afferent sensory impulses and sending out efferent motor instructions, as carried out by sensory nerve fibers and motor nerve fibers
The main objectives of this study were four folds: (1) to design templates for position line cross sections, to recognize the position line images and to register images of the nerve serial cross sections; (2) to extract images of the fascicle edges and merge them onto the nerve cross section images to obtain continuous and smooth edges of nerve fascicles; (3) to use multi-directional gradient and second-order gradient method combined with rough-means clustering method to achieve automatic functional type classification of nerve fascicles; and (4) to process the serial cross section images and achieve a 3D reconstruction of Peripheral nerves (PNs) fascicles
The entire procedure consists of six stages, namely, (1) serial section preparation and imaging, (2) position line recognition of section images, (3) image registration, (4) edge acquisition of nerve fascicles, (5) functional type recognition of nerve fascicles, and (6) 3D reconstruction of internal fascicles

Summary

Introduction

Peripheral nerves are important pathways for receiving afferent sensory impulses and sending out efferent motor instructions, as carried out by sensory nerve fibers and motor nerve fibers. A 3D virtual model of internal fascicles was obtained after section images were processed This technique was successfully applied for 3D reconstruction for the median nerve of the hand-wrist and cubital fossa regions and the gastrocnemius nerve. The main objectives of this study were four folds: (1) to design templates for position line cross sections, to recognize the position line images and to register images of the nerve serial cross sections (or transform them into one coordinate system); (2) to extract images of the fascicle edges and merge them onto the nerve cross section images to obtain continuous and smooth edges of nerve fascicles; (3) to use multi-directional gradient and second-order gradient method combined with rough-means clustering method to achieve automatic functional type classification of nerve fascicles; and (4) to process the serial cross section images and achieve a 3D reconstruction of PN fascicles. Establishment of a 3D visualization system of nerve fascicle types will be of significance for peripheral nerve repair

Objectives

Methods

Results

Conclusion