Lung-Noudle-Detection

This the extra-low-dose CT lung nodule detection demo code developed by "Shiwen Shen" [email protected] for CDSC project.

This file introduces the workflow and usage of the lung nodule detection pipeline.

######################################################

Workflow

###################################################### On-line detection task:

Input: CT lung image stacks (Analyze file format) Output: Lung nodule binary mask which could be mapped directly to the original images(ANALYZE 7.5 format) 1 segmentation (see folder segmentation): segment the initial nodule candidates from CT images 2 preselection (see folder preselection): reduce the false positive rate based on pre-defined rules 3 feature extraction (see folder feature extraction): generate 27 features for each nodule candidates 4 classification (see mainNoduleDetection.m file): output the final nodule using trained classifier

PLEASE NOTE Depending on the size of the input image, the pipeline may require a large amount (> 4Gb) of memory to complete its task. As a result, the computer may freeze until the pipeline has completed its execution. We recommend that you run the pipeline on a dedicated development environment.

#################################################

Usage

#################################################

############################### main lung nodule detection task

Step 1: Add the folder of current code and its subfolders to path

Step 2: Open "mainNoduleDetectionAnalyzeFile.m" file, this is the main function for lung nodule detection.

Step 3: Run this main function and 3D nodule mask will be automatically shown and you can view it by scrolling the mouse

Author: Shiwen Shen Date: 09/28/2014 Email: [email protected] Copyright: Medical Imaging Informatics Group, UCLA

Citation

Please cite the following papers if this code is used for any publication purpose

[1] Shen, Shiwen, et al. "An automated lung segmentation approach using bidirectional chain codes to improve nodule detection accuracy." Computers in biology and medicine 57 (2015): 139-149.

[2] Duggan, Nóirín, et al. "A Technique for Lung Nodule Candidate Detection in CT Using Global Minimization Methods." Energy Minimization Methods in Computer Vision and Pattern Recognition. Springer International Publishing, 2015.