Codes

Three-dimensional neurite tracing under globally varying contrast
by S. Gulyanon, N. Sharifai, S. Bleykhman, E. Kelly, M.D. Kim, A. Chiba, and G. Tsechpenakis, in Int'l Symposium on Biomedical Imaging: from Nano to Macro (ISBI), Brooklyn, NY, 2015.
[Matlab GUI & Code for MAC, PC][Manual]
Description: This software traces single neuron in image stacks with varying contrast background using multiple active contour models or snakes in the bottom-up fashion, where we spilt the stack into sub-volumes and trace them, then we merge traces together.
Instruction:
- Download the MATLAB package "SnakeTracingMAC.mlappinstall" for Mac OS X or "SnakeTracingPC.mlappinstall" for Windows (Tested on MATLAB 2014a).
- Double click the file to install.
- To run, select APPS > Snake Tracing. The GUI manual is here.

CRF formulation of active contour population for efficient three-dimensional neurite tracing
by S.Gulyanon, N. Sharifai, M.D. Kim, A. Chiba, and G. Tsechpenakis, in Int'l Symposium on Biomedical Imaging: from Nano to Macro (ISBI), Prague, Czech Republic, 2016.
[Matlab Code]
Description: This code traces single neuron in image stacks with varying contrast background using the population of active contour models or snakes under the CRF framework.
Instruction:
- Download the code and unzip it.
- Main tracing function is 'trace_main.m'. See the example of how to use in 'script.m'. The detail of the function is described in 'README.md'.

Neurite reconstruction from time-lapse sequences using co-segmentation
by S. Gulyanon, N. Sharifai, M.D. Kim, A. Chiba, and G. Tsechpenakis, in Int'l Symposium on Biomedical Imaging: from Nano to Macro (ISBI), Melbourne, Australia, 2017.
[Matlab Code]
Description: This code segments single neuron over time from time-lapse image stacks using co-segmentation principle postulated as MRF, which is optimized by the graph cut method.
Instruction:
- Download the code and unzip it.
- Preprocessing images by running 'RobustGPMatching3D/generateDrosophiladata.m'.
- The main function is 'mycoseg/main_coseg.m'. Example is in 'mycoseg/script.m'. The detailed instruction is in 'README.md'.

Tracking neuron reconstruction in calcium imaging using markov random field
by S. Gulyanon, W. D. Tracy, L. He, and G. Tsechpenakis
[Matlab Code]
Description: This code tracks the location of dendrites of neurons over time for both 3D (XYT) and 4D (XYZT) calcium image datasets. The method models the neuron as the articulated body and solves the transformation using MRF with fusion moves (QPBO + alpha-expansion).
Instruction:
- Download the code and unzip it.
- Main tracking function is 'mrftrack.m'.
- Optical flow and soma location must be computed first. Examples are shown in trackGFPRFPXYTdata.m' (for 3D XYT) and 'trackGFPRFPXYZTdata.m' (for 4D XYZT). The detailed instruction is in 'README.md'.