Introduction: Bone cancer is a common type of cancer in canines. There are various bone cancers, of which osteosarcoma is the most prevalent in large breed dogs. It usually appears in limbs and grows fast and is highly metastatic. Imaging methods such as x-ray, computed tomography (CT), and magnetic resonance imaging (MRI), in addition to biopsies, are used primarily for diagnosis. These imaging methods are expensive, comprising large dose of radiation (CT/X-ray), and require the patient (canine) to be motionless for long period during the imaging process. Also, these techniques may not diagnose this type of cancer at an early stage, which will have direct consequences as it is highly metastatic.
Objectives: This research study investigates the possibility of using thermographic images for diagnosis of bone cancer in dogs, the advantages include: it is noninvasive, cheaper, faster, uses no radiation, uses real time imaging, is portable and accessible for computer aided image analysis.
Experimental Methods: Experiments were performed at two different stages. The images were divided into two groups based on classification, cancer and no-cancer. Next, the images were grouped according to body parts, elbow/knee, and wrist. Each of the groups was then further divided into two sub-groups corresponding to the camera views, anterior, and lateral. Thermographic patterns of normal and abnormal canines were analyzed using feature extraction and pattern classification. Second-order histogram texture, histogram and Laws texture features were extracted from the thermograms and were used for pattern classification. Artificial neural networks were used for the Pattern classification.
Results: The best classification success rate in canine bone cancer detection was 87.5% with sensitivity of 85% and specificity of 90% produced by lateral view of wrist region with ANN. ANN has the better classification success than CVIP-FEPC in two cases, whereas for one data set FEPC has good accuracy. This shows that the pattern classification algorithms that were used in CVIP-FEPC were sufficient for pattern classification for this research. Features with best classification success with each data normalization method were chosen and used for ANN using Matlab. The feature values that were not normalized were used. Principal component transform was performed before these were input to the artificial neural network. Only the principal components that contain 93 - 96% of the variance were considered. The best classification success rate of 100% with 100% sensitivity and 100% specificity was obtained for lateral view of wrist region with artificial neural network as the pattern classification method.
Conclusion: Our results show that it is possible to use thermographic imaging as a prescreening
tool for detection of canine bone cancer. We have also determined that filtering of features, normalizations by considering the success rate of individual experiments and perform experiments combining filtered features, classifications has significant effect on the overall accuracy.
|Advisor:||Umbaugh, Scott E.|
|Commitee:||Klingensmith, Jon, Leander, Robert|
|School:||Southern Illinois University at Edwardsville|
|Department:||Electrical and Computer Engineering|
|School Location:||United States -- Illinois|
|Source:||MAI 81/12(E), Masters Abstracts International|
|Subjects:||Computer Engineering, Artificial intelligence|
|Keywords:||Artificial neural networks, Bone cancer, Computer vision, CVIP-FEPC, Pattern classification, Thermograms|
Copyright in each Dissertation and Thesis is retained by the author. All Rights Reserved
The supplemental file or files you are about to download were provided to ProQuest by the author as part of a
dissertation or thesis. The supplemental files are provided "AS IS" without warranty. ProQuest is not responsible for the
content, format or impact on the supplemental file(s) on our system. in some cases, the file type may be unknown or
may be a .exe file. We recommend caution as you open such files.
Copyright of the original materials contained in the supplemental file is retained by the author and your access to the
supplemental files is subject to the ProQuest Terms and Conditions of use.
Depending on the size of the file(s) you are downloading, the system may take some time to download them. Please be