Advanced

Automated interpretation of PET/CT images in patients with lung cancer.

Gutte, Henrik; Jakobsson, David; Olofsson, Fredrik; Ohlsson, Mattias LU ; Valind, Sven LU ; Loft, Annika; Edenbrandt, Lars LU and Kjær, Andreas (2007) In Nuclear Medicine Communications 28(2). p.79-84
Abstract
Purpose: To develop a completely automated method based on image processing techniques and artificial neural networks for the interpretation of combined [18F]fluorodeoxyglucose (FDG) positron emission tomography (PET) and computed tomography (CT) images for the diagnosis and staging of lung cancer.



Methods: A total of 87 patients who underwent PET/CT examinations due to suspected lung cancer comprised the training group. The test group consisted of PET/CT images from 49 patients suspected with lung cancer. The consensus interpretations by two experienced physicians were used as the 'gold standard' image interpretation. The training group was used in the development of the automated method. The image processing techniques... (More)
Purpose: To develop a completely automated method based on image processing techniques and artificial neural networks for the interpretation of combined [18F]fluorodeoxyglucose (FDG) positron emission tomography (PET) and computed tomography (CT) images for the diagnosis and staging of lung cancer.



Methods: A total of 87 patients who underwent PET/CT examinations due to suspected lung cancer comprised the training group. The test group consisted of PET/CT images from 49 patients suspected with lung cancer. The consensus interpretations by two experienced physicians were used as the 'gold standard' image interpretation. The training group was used in the development of the automated method. The image processing techniques included algorithms for segmentation of the lungs based on the CT images and detection of lesions in the PET images. Lung boundaries from the CT images were used for localization of lesions in the PET images in the feature extraction process. Eight features from each examination were used as inputs to artificial neural networks trained to classify the images. Thereafter, the performance of the network was evaluated in the test set.



Results: The performance of the automated method measured as the area under the receiver operating characteristic curve, was 0.97 in the test group, with an accuracy of 92%. The sensitivity was 86% at a specificity of 100%.



Conclusions: A completely automated method using artificial neural networks can be used to detect lung cancer with such a high accuracy that the application as a clinical decision support tool appears to have significant potential. (Less)
Please use this url to cite or link to this publication:
author
organization
publishing date
type
Contribution to journal
publication status
published
subject
in
Nuclear Medicine Communications
volume
28
issue
2
pages
79 - 84
publisher
Lippincott Williams & Wilkins
external identifiers
  • wos:000243759100003
  • scopus:33846010170
ISSN
1473-5628
DOI
10.1097/MNM.0b013e328013eace
language
English
LU publication?
yes
id
ed1f5b88-f7bc-4a36-9a45-b3899d791779 (old id 165186)
alternative location
http://www.nuclearmedicinecomm.com/pt/re/nucmedcomm/abstract.00006231-200702000-00003.htm;jsessionid=HZHdsQzGCthLvRHHfj3qvRq1J1F1jDQhpBdTnGp2lt53wttQG3ng!-1288052477!181195628!8091!-1
date added to LUP
2007-07-23 13:03:56
date last changed
2017-01-01 04:57:19
@article{ed1f5b88-f7bc-4a36-9a45-b3899d791779,
  abstract     = {Purpose: To develop a completely automated method based on image processing techniques and artificial neural networks for the interpretation of combined [18F]fluorodeoxyglucose (FDG) positron emission tomography (PET) and computed tomography (CT) images for the diagnosis and staging of lung cancer.<br/><br>
<br/><br>
Methods: A total of 87 patients who underwent PET/CT examinations due to suspected lung cancer comprised the training group. The test group consisted of PET/CT images from 49 patients suspected with lung cancer. The consensus interpretations by two experienced physicians were used as the 'gold standard' image interpretation. The training group was used in the development of the automated method. The image processing techniques included algorithms for segmentation of the lungs based on the CT images and detection of lesions in the PET images. Lung boundaries from the CT images were used for localization of lesions in the PET images in the feature extraction process. Eight features from each examination were used as inputs to artificial neural networks trained to classify the images. Thereafter, the performance of the network was evaluated in the test set.<br/><br>
<br/><br>
Results: The performance of the automated method measured as the area under the receiver operating characteristic curve, was 0.97 in the test group, with an accuracy of 92%. The sensitivity was 86% at a specificity of 100%.<br/><br>
<br/><br>
Conclusions: A completely automated method using artificial neural networks can be used to detect lung cancer with such a high accuracy that the application as a clinical decision support tool appears to have significant potential.},
  author       = {Gutte, Henrik and Jakobsson, David and Olofsson, Fredrik and Ohlsson, Mattias and Valind, Sven and Loft, Annika and Edenbrandt, Lars and Kjær, Andreas},
  issn         = {1473-5628},
  language     = {eng},
  number       = {2},
  pages        = {79--84},
  publisher    = {Lippincott Williams & Wilkins},
  series       = {Nuclear Medicine Communications},
  title        = {Automated interpretation of PET/CT images in patients with lung cancer.},
  url          = {http://dx.doi.org/10.1097/MNM.0b013e328013eace},
  volume       = {28},
  year         = {2007},
}