Lung segmentation on standard and mobile chest radiographs using oriented Gaussian derivatives filter

Wan Siti Halimatul Munirah Wan Ahmad, Wan Mimi Diyana Wan Zaki, Mohammad Faizal Ahmad Fauzi

Research output: Contribution to journalArticle

25 Citations (Scopus)

Abstract

Background: Unsupervised lung segmentation method is one of the mandatory processes in order to develop a Content Based Medical Image Retrieval System (CBMIRS) of CXR. The purpose of the study is to present a robust solution for lung segmentation of standard and mobile chest radiographs using fully automated unsupervised method. Methods: The novel method is based on oriented Gaussian derivatives filter with seven orientations, combined with Fuzzy C-Means (FCM) clustering and thresholding to refine the lung region. In addition, a new algorithm to automatically generate a threshold value for each Gaussian response is also proposed. The algorithms are applied to both PA and AP chest radiographs from both public JSRT dataset and our private datasets from collaborative hospital. Two pre-processing blocks are introduced to standardize the images from different machines. Comparisons with the previous works found in the literature on JSRT dataset shows that our method gives a reasonably good result. We also compare our algorithm with other unsupervised methods to provide fairly comparative measures on the performances for all datasets. Results: Performance measures (accuracy, F-score, precision, sensitivity and specificity) for the segmentation of lung in public JSRT dataset are above 0.90 except for the overlap measure is 0.87. The standard deviations for all measures are very low, from 0.01 to 0.06. The overlap measure for the private image database is 0.81 (images from standard machine) and 0.69 (images from two mobile machines). The algorithm is fully automated and fast, with the average execution time of 12.5 s for 512 by 512 pixels resolution. Conclusions: Our proposed method is fully automated, unsupervised, with no training or learning stage is necessary to segment the lungs taken using both a standard machine and two different mobile machines. The proposed pre-processing blocks are significantly useful to standardize the radiographs from mobile machines. The algorithm gives good performance measures, robust, and fast for the application of the CBMIRS.

Original languageEnglish
Article number20
JournalBioMedical Engineering Online
Volume14
Issue number1
DOIs
Publication statusPublished - 4 Mar 2015

Fingerprint

Thorax
Derivatives
Lung
Image retrieval
Processing
Pixels
Cluster Analysis
Datasets
Learning
Databases
Sensitivity and Specificity

Keywords

  • Chest radiograph
  • Fuzzy c-means
  • Gaussian derivatives
  • Medical image processing
  • Segmentation algorithm
  • Thresholding
  • Unsupervised lung segmentation

ASJC Scopus subject areas

  • Radiological and Ultrasound Technology
  • Radiology Nuclear Medicine and imaging
  • Biomaterials
  • Biomedical Engineering

Cite this

Lung segmentation on standard and mobile chest radiographs using oriented Gaussian derivatives filter. / Wan Ahmad, Wan Siti Halimatul Munirah; Wan Zaki, Wan Mimi Diyana; Ahmad Fauzi, Mohammad Faizal.

In: BioMedical Engineering Online, Vol. 14, No. 1, 20, 04.03.2015.

Research output: Contribution to journalArticle

@article{322f55633e1f4271bef2153f82d4b5c3,
title = "Lung segmentation on standard and mobile chest radiographs using oriented Gaussian derivatives filter",
abstract = "Background: Unsupervised lung segmentation method is one of the mandatory processes in order to develop a Content Based Medical Image Retrieval System (CBMIRS) of CXR. The purpose of the study is to present a robust solution for lung segmentation of standard and mobile chest radiographs using fully automated unsupervised method. Methods: The novel method is based on oriented Gaussian derivatives filter with seven orientations, combined with Fuzzy C-Means (FCM) clustering and thresholding to refine the lung region. In addition, a new algorithm to automatically generate a threshold value for each Gaussian response is also proposed. The algorithms are applied to both PA and AP chest radiographs from both public JSRT dataset and our private datasets from collaborative hospital. Two pre-processing blocks are introduced to standardize the images from different machines. Comparisons with the previous works found in the literature on JSRT dataset shows that our method gives a reasonably good result. We also compare our algorithm with other unsupervised methods to provide fairly comparative measures on the performances for all datasets. Results: Performance measures (accuracy, F-score, precision, sensitivity and specificity) for the segmentation of lung in public JSRT dataset are above 0.90 except for the overlap measure is 0.87. The standard deviations for all measures are very low, from 0.01 to 0.06. The overlap measure for the private image database is 0.81 (images from standard machine) and 0.69 (images from two mobile machines). The algorithm is fully automated and fast, with the average execution time of 12.5 s for 512 by 512 pixels resolution. Conclusions: Our proposed method is fully automated, unsupervised, with no training or learning stage is necessary to segment the lungs taken using both a standard machine and two different mobile machines. The proposed pre-processing blocks are significantly useful to standardize the radiographs from mobile machines. The algorithm gives good performance measures, robust, and fast for the application of the CBMIRS.",
keywords = "Chest radiograph, Fuzzy c-means, Gaussian derivatives, Medical image processing, Segmentation algorithm, Thresholding, Unsupervised lung segmentation",
author = "{Wan Ahmad}, {Wan Siti Halimatul Munirah} and {Wan Zaki}, {Wan Mimi Diyana} and {Ahmad Fauzi}, {Mohammad Faizal}",
year = "2015",
month = "3",
day = "4",
doi = "10.1186/s12938-015-0014-8",
language = "English",
volume = "14",
journal = "BioMedical Engineering Online",
issn = "1475-925X",
publisher = "BioMed Central",
number = "1",

}

TY - JOUR

T1 - Lung segmentation on standard and mobile chest radiographs using oriented Gaussian derivatives filter

AU - Wan Ahmad, Wan Siti Halimatul Munirah

AU - Wan Zaki, Wan Mimi Diyana

AU - Ahmad Fauzi, Mohammad Faizal

PY - 2015/3/4

Y1 - 2015/3/4

N2 - Background: Unsupervised lung segmentation method is one of the mandatory processes in order to develop a Content Based Medical Image Retrieval System (CBMIRS) of CXR. The purpose of the study is to present a robust solution for lung segmentation of standard and mobile chest radiographs using fully automated unsupervised method. Methods: The novel method is based on oriented Gaussian derivatives filter with seven orientations, combined with Fuzzy C-Means (FCM) clustering and thresholding to refine the lung region. In addition, a new algorithm to automatically generate a threshold value for each Gaussian response is also proposed. The algorithms are applied to both PA and AP chest radiographs from both public JSRT dataset and our private datasets from collaborative hospital. Two pre-processing blocks are introduced to standardize the images from different machines. Comparisons with the previous works found in the literature on JSRT dataset shows that our method gives a reasonably good result. We also compare our algorithm with other unsupervised methods to provide fairly comparative measures on the performances for all datasets. Results: Performance measures (accuracy, F-score, precision, sensitivity and specificity) for the segmentation of lung in public JSRT dataset are above 0.90 except for the overlap measure is 0.87. The standard deviations for all measures are very low, from 0.01 to 0.06. The overlap measure for the private image database is 0.81 (images from standard machine) and 0.69 (images from two mobile machines). The algorithm is fully automated and fast, with the average execution time of 12.5 s for 512 by 512 pixels resolution. Conclusions: Our proposed method is fully automated, unsupervised, with no training or learning stage is necessary to segment the lungs taken using both a standard machine and two different mobile machines. The proposed pre-processing blocks are significantly useful to standardize the radiographs from mobile machines. The algorithm gives good performance measures, robust, and fast for the application of the CBMIRS.

AB - Background: Unsupervised lung segmentation method is one of the mandatory processes in order to develop a Content Based Medical Image Retrieval System (CBMIRS) of CXR. The purpose of the study is to present a robust solution for lung segmentation of standard and mobile chest radiographs using fully automated unsupervised method. Methods: The novel method is based on oriented Gaussian derivatives filter with seven orientations, combined with Fuzzy C-Means (FCM) clustering and thresholding to refine the lung region. In addition, a new algorithm to automatically generate a threshold value for each Gaussian response is also proposed. The algorithms are applied to both PA and AP chest radiographs from both public JSRT dataset and our private datasets from collaborative hospital. Two pre-processing blocks are introduced to standardize the images from different machines. Comparisons with the previous works found in the literature on JSRT dataset shows that our method gives a reasonably good result. We also compare our algorithm with other unsupervised methods to provide fairly comparative measures on the performances for all datasets. Results: Performance measures (accuracy, F-score, precision, sensitivity and specificity) for the segmentation of lung in public JSRT dataset are above 0.90 except for the overlap measure is 0.87. The standard deviations for all measures are very low, from 0.01 to 0.06. The overlap measure for the private image database is 0.81 (images from standard machine) and 0.69 (images from two mobile machines). The algorithm is fully automated and fast, with the average execution time of 12.5 s for 512 by 512 pixels resolution. Conclusions: Our proposed method is fully automated, unsupervised, with no training or learning stage is necessary to segment the lungs taken using both a standard machine and two different mobile machines. The proposed pre-processing blocks are significantly useful to standardize the radiographs from mobile machines. The algorithm gives good performance measures, robust, and fast for the application of the CBMIRS.

KW - Chest radiograph

KW - Fuzzy c-means

KW - Gaussian derivatives

KW - Medical image processing

KW - Segmentation algorithm

KW - Thresholding

KW - Unsupervised lung segmentation

UR - http://www.scopus.com/inward/record.url?scp=84925010179&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84925010179&partnerID=8YFLogxK

U2 - 10.1186/s12938-015-0014-8

DO - 10.1186/s12938-015-0014-8

M3 - Article

C2 - 25889188

AN - SCOPUS:84925010179

VL - 14

JO - BioMedical Engineering Online

JF - BioMedical Engineering Online

SN - 1475-925X

IS - 1

M1 - 20

ER -