Veterinary Thermographic Image Analysis



[pic]

Veterinary Thermographic Image Analysis

Data and Temperature Normalization

Project Number 7- 64878

Report Number 4878-4

Scott E Umbaugh, BSE, MSEE, PhD

Patrick Solt, BSEE, MSCS, PhD Candidate

Computer Vision and Image Processing Laboratory

Southern Illinois University Edwardsville

June 30, 2008

Submitted to:

Dr. Catherine Loughin

Dr. Dominic Marino, Chief of Staff

Long Island Veterinary Specialists

TABLE OF CONTENTS

Executive Summary 2

Introduction and Overview 3

Materials and Methods 3

Images 3

Software 3

Data Normalization 6

Color Normalization 8

Experimental Results and Discussion 11

Results from Experiment Set #1 11

Results from Experiment Set #2 13

Conclusions 15

Future Work 16

References 17

Appendix –Result File Names 18

Executive Summary

Research and development was undertaken to investigate the application of image analysis and pattern classification to thermographic images of 34 canines. Specifically, thermographic images of canines of the breed Cavalier King Charles Spaniel were examined to investigate the Chiari malformation, or COMS, pathology. Pattern classification algorithms were developed for: severe, moderate and mild classes of the pathology. For these experiments the classification was performed using pairs of images, instead of single images as in previous experiments. The head images were used, and for this report the pair was top and front of head images. In progress are experiments for the: 1) top and right side, 2) front and left side, 3) front and right side, 4) top and left side. The eventual goal for the research is to be able to differentiate normal and abnormal thermographic patterns in canines as a diagnostic and research tool.

Two sets of experiments were performed using three data normalization methods and the original and four color normalization methods for a total for of 7,864,290 experiments. Each experiment consisted of two views of the head, specific classes, a specific data normalization method, a specific color normalization method and specific features. Experiments were performed with all permutations of the nine selected features on each pair of selected body parts. These nine are the five texture features and four of the histogram features – histogram mean was not used. Based on previous experimentation the MinMax, Softmax and Standard Normal Density data normalization methods provided the best results and were selected for these experiments. The texture distance of 6 provided the best overall results in previous experimentation and was selected for use here. Regarding color normalization, the original and the four peviously defined methods were used.

The first set of experiments used the clinical images that were classified as severe and moderate classes of the pathology. Here, the front of head (A1) and top of head (A1D) images were used. With top and front of head image pair, 100% successful classification was achieved. The Norm-RGB-Lum is the best color normalization method. These results indicate that there is a difference between the moderate and severe classes with the clinical images. These findings support previous results.

The second set of experiments used the nonclinical images that were classified as severe, moderate and mild classes of the pathology. Here, the front and top of head images were used. With this image pair, a classification success rate of 78% was achieved with a texture distance of six. The best color normalization method here was normGray.These results are inconclusive but indicate that there may be a difference between the mild, moderate and severe classes in the nonclinical images.

After the first two sets of experiments were performed, comparisons of success rates were made between results on the clinical and nonclinical canine images. Comparing the results we find that the separate classes in the clinical (moderate and severe) images are more homogeneous than the classes (mild, moderate and severe) in the nonclinical images. This finding supports previous results, and is much more strongly indicated by these experiments.

Introduction and Overview

Thermographic images of canines were analyzed using feature extraction and pattern classification tools. The purpose of this research is to examine the thermographic images of canines of the breed Cavalier King Charles Spaniel to investigate the Chiari malformation, or COMS, pathology.

The primary question here was to explore the use of pairs of images, instead of single images as in previous experiments.

1. Can we successfully differentiate between the classes moderate and severe in the clinical images? Here this pair of images was used:

➢ Use top (A1D) and front (A1) head images

2. Can we successfully differentiate between the classes mild, moderate and severe in the nonclinical images? Here this pair of images was used:

➢ Use top (A1D) and front (A1) head images

Materials and Methods

Images

The thermal images are from Long Island Veterinary Specialists taken with a Meditherm Med2000 IRIS. The images are in TIF file format as RGB images with 319 columns by 238 rows, 8-bits per pixel per color band. Note that a total of 18 colors are used in these images. The images are thermal images of 34 canines of the breed Cavalier King Charles Spaniel. Included in the data set are images of front of head, top of head, left and right side of head.

Software

The CVIPtools [1] (Computer Vision and Image Processing Tools) software was used to investigate and analyze the images. The primary tools used include Analysis->Features and Analysis->Pattern Classification. The features considered include (see Figure 1) spectral (Fourier), histogram, and texture features. The histogram features include mean, standard deviation, skew, energy and entropy; and the texture features include energy, inertia, correlation, inverse difference, and entropy. The pattern classification tools utilized include various normalization methods, distance and similarity measures and three classification methods (see Figure 2).

[pic]

Figure 1 – CVIPtools Feature Extraction window

[pic]

Figure 2 – CVIPtools Pattern Classification window

In the course of the image analysis it was determined that the previously developed special purpose software would be further enhanced for this phase of the research. The software is designed to ease the analysis process and enable us to perform a significantly greater number of experiments. We added the capability to run groups of experiments simultaneously, with the same parameters, which allowed for the analysis of comparative results. One of the primary factors under investigation was color normalization, and the software was employed to experiment with a number of normalization techniques and methods.

Data Normalization

Data normalization is used to compensate for potential bias due to the varying range on different components of the feature vectors. For example, one component may only range from 1 to 5 and another may range from 1 to 5,000, so a difference of 5 would be a maximum for the first feature, while this same difference might be insignificant for the second feature. Here we explored five data normalization methods:

• Standard normal density (SND): A statistical-based method to normalize feature vectors by taking each vector component and subtracting the mean and dividing by the standard deviation. The probability distributions are estimated by using the existing data.

This is done as follows, given a set of k feature vectors, Fj is a member of the set {F1, F2, …, Fk}, with n features in each vector:

[pic]

Now, for each feature component, we subtract the mean and divide by the standard deviation:

[pic]

• Min-max normalization: Maps the data to a specified range, but still retains the relationship between the values, as follows:

[pic]

• Softmax scaling: A nonlinear method for skewed data distributions, that is, not evenly distributed about the mean. It compresses the data into the range of 0 to 1. The process compresses the data exponentially as it gets farther away from the mean, as follows:

[pic]

With the mean and standard deviation defined as in the standard normal density distribution above.

Color Normalization

For this phase of the research, several different methods for normalizing the colors in the images were compared. Each of the images uses the same color palette consisting of 18 colors. Within an image, each color represents a specific temperature. The color to temperature mapping for an image is based on camera settings that were in effect at the time the image was captured. Since the camera may be recalibrated between each image capture, to obtain best results for each image, within a set of images one color may map to several different temperatures. This shifting of temperature scale between images could introduce noise in the pattern classification process. To compensate for this, a remapping was applied to the color values of each image so that all of the images are mapped to a common temperature scale. Four different color normalization methods were used:

• Luminance: For each pixel in the original image, the color components are weighted as follows to generate a gray level value:

[pic]

• Norm-Gray: Temperatures from 15° C to 45° C are mapped to gray levels from 0 to 255. Each pixel in the normalized image is colored with the gray level that is mapped to the temperature represented by the color of the corresponding pixel in the original image. Figure 3 (below) shows the color remapping that occurs when this normalization method is applied to an image for which the temperature range is 21° C to 34° C. The final step of this normalization method is a histogram stretch that occurs when the image is saved as a TIFF file.

[pic]

Figure 3 – Norm-Gray remapping for image with temperature range of 21° C to 34° C

• Norm-RGB: Similar to Norm-Gray except that instead of temperatures being mapped to gray levels from 0 to 255, they are mapped to a continuous version of the original color palette. An example of the remapping that occurs when this method is applied to an image for which the temperature range is 21° C to 34° C is shown in Figure 4.

[pic]

Figure 4 – Norm-RGB remapping for image with temperature range of 21° C to 34° C

• Norm-RGB-Lum: This method produces a gray-scale image by applying the Norm-RGB normalization method to an image and then applying the luminance normalization method to the result.

Figures 5a through 5e show an original image, along with the images resulting from each of the normalization methods used:

[pic]

Figure 5a – Original Image

[pic]

Figure 5c – Norm-Gray Normalized Image

[pic]

Figure 5e – Norm-RGB-Lum Normalized Image

[pic]

Figure 5b – Luminance Normalized Image

[pic]

Figure 5d – Norm-RGB Normalized Image

Experimental Results and Discussion

Results from Experiment Set #1

Images: Clinical group, the front and top images of the head.

Classes: moderate and severe.

For complete results see the text files:

Exp1-A1-A1D-MinMax-results.txt

    Exp1-A1-A1D-SoftMax-results.txt

    Exp1-A1-A1D-StdNorm-results.txt

|Classification results for Experiment #1 (Clinical), |

|Top and Front Head images, SoftMax Data Normalization |

|Experiments with best success rate of 100% (9/9 correct) |

|124,272 out of 1,310,715 experiments had 100% success |

|Color |Texture distance |Number of Experiments with |Percentage |Color Normalization Ranking |

|Normalization | |100% | | |

|Original |6 |15,824 |20% |5 |

|Luminance |6 |16,998 |22% |4 |

|normGray |6 |30,058 |38% |2 |

|normRGB |6 |19,441 |25% |3 |

|normRGBLum |6 |41,951 |53% |1 |

|Classification results for Experiment #1 (Clinical), |

|Top and Front Head images, MinMax Data Normalization |

|Experiments with best success rate of 100% (9/9 correct) |

|111,343 out of 1,310,715 experiments had 100% success |

|Color |Texture distance |Number of Experiments with |Percentage |Color Normalization Ranking |

|Normalization | |100% | | |

|Original |6 |6,957 |9% |5 |

|Luminance |6 |16,226 |22% |3 |

|normGray |6 |34,634 |46% |2 |

|normRGB |6 |12,659 |17% |4 |

|normRGBLum |6 |40,867 |54% |1 |

|Classification results for Experiment #1 (Clinical), |

|Top and Front Head images, StdNorm Data Normalization |

|Experiments with best success rate of 100% (9/9 correct) |

|106,389 out of 1,310,715 experiments had 100% success |

|Color |Texture distance |Number of Experiments with |Percentage |Color Normalization Ranking |

|Normalization | |100% | | |

|Original |6 |9,225 |13% |5 |

|Luminance |6 |12,276 |17% |4 |

|normGray |6 |22,849 |32% |3 |

|normRGB |6 |23,083 |33% |3 |

|normRGBLum |6 |38,956 |55% |1 |

➢ We believe that the 100% classification success indicates that the top and front of head pair is the useful for differentiation of the two classes.

➢ These results indicate that there is a difference between the moderate and severe classes in the clinical images.

➢ The best previous rate for front of head alone images was also 100%, but only 12 of the experiments from this group had a 100% classification rate. Here we had over 100,000 experiments with the 100% success rate for each data normalization method.

➢ These results indicate that all three data normalization methods, softmax, minmax and stdnorm, are useful

➢ Norm-RGB-Lum provided the best color normalization results with all three data normalization methods

➢ The results are limited by the small size of the data sets

Results from Experiment Set #2

Images: Nonclinical group, the front and top images of the head.

Classes: mild, moderate and severe.

For complete results see the text files:

Exp2-A1-A1D-MinMax-results.txt

    Exp2-A1-A1D-SoftMax-results.txt

    Exp2-A1-A1D-StdNorm-results.txt

|Classification results for Experiment #2 (Non-clinical), |

|Top and Front Head images, SoftMax Data Normalization |

|Experiments with best success rate of 78% (18/23 correct) |

|73 out of 1,310,715 experiments had 78% success |

|Color |Texture distance |Number of Experiments with |Percentage (x/73) |Color Normalization Ranking |

|Normalization | |78% | | |

|Original |6 |5 |7% |3 |

|Luminance |6 |6 |8% |2 |

|normGray |6 |67 |92% |1 |

|normRGB |6 |0 |0% |X |

|normRGBLum |6 |0 |0% |X |

|Classification results for Experiment #2 (Non-clinical), |

|Top and Front Head images, MinMax Data Normalization |

|Experiments with best success rate of 78% (18/23 correct) |

|15 out of 1,310,715 experiments had 78% success |

|Color |Texture distance |Number of Experiments with |Percentage (x/15) |Color Normalization Ranking |

|Normalization | |78% | | |

|Original |6 |6 |40% |2 |

|Luminance |6 |0 |0% |X |

|normGray |6 |9 |60% |1 |

|normRGB |6 |0 |0% |X |

|normRGBLum |6 |0 |0% |X |

|Classification results for Experiment #2 (Non-clinical), |

|Top and Front Head images, StdNorm Data Normalization |

|Experiments with best success rate of 78% (18/23 correct) |

|30 out of 1,310,715 experiments had 78% success |

|Color |Texture distance |Number of Experiments with |Percentage (x/30) |Color Normalization Ranking |

|Normalization | |78% | | |

|Original |6 |7 |23% |2 |

|Luminance |6 |6 |20% |3 |

|normGray |6 |17 |57% |1 |

|normRGB |6 |0 |0% |X |

|normRGBLum |6 |0 |0% |X |

➢ These results indicate that normGray is the best color normalization method of the non-clinical images, with all three classes

➢ These results are inconclusive but indicate that there may be a difference between the mild, moderate and severe classes in the nonclinical images.

➢ Comparing then results from experiment #1, where we had over 100,000, 100% success rate experiments to eperiment #2 results, where we had at best 73, 78% success rate experiments, we feel confident in stating that the Clinical image classes of Moderate and Severe are much better separated than the Non-clinical image classes of Mild, Moderate and Severe.

Conclusions

Two sets of experiments were performed for a total for of 7,864,290 experiments. The top and front of head image pair was used. Previous experimentation showed that the MinMax, Softmax and Standard Normalization data normalization methods provided the best results and were thus selected for further experimentation. The texture distances of 6 provided the best overall results and were selected for these experiments. Regarding color normalization, the original and the four normalization methods were used.

The first set of experiements was to investigate classification of the moderate and severe classes with the clinical images. With top and front of head image pair, 100% successful classification was achieved. The Norm-RGB-Lum is the best color normalization method. These results indicate that there is a difference between the moderate and severe classes with the clinical images. These findings support previous results.

The second set of experiments used the nonclinical images and classified that were classified as severe, moderate and mild classes of the pathology. Here, the front and top of head images were used. With this image pair, a classification success rate of 78% was achieved with a texture distance of six. The best color normalization method here was normGray.These results are inconclusive but indicate that there may be a difference between the mild, moderate and severe classes in the nonclinical images.

Comparing the results we find that the separate classes in the clinical (moderate and severe) images are more homogeneous than the classes (mild, moderate and severe) in the nonclinical images. This finding supports previous results, and is much more strongly indicated by these experiments.

Future Work

• Complete experiments on the other pairs of images: 1) top and right side, 2) front and left side, 2) front and right side, 4) top and left side.

• Rerun experiment #2 with a texture distance of 7 to achieve the best success from the previous report of 83% (19/23)

• Collect more images for differentiation of the mild, moderate and severe classes of both clinical and nonclinical images

• Further analyze the results from the experiments, including analysis of frequency of features relative to classification success and feature set size relative to success

• Analyze Experiment Set2 #2 to determine why specific examples were misclassified

• Investigate use of support vector machines for analysis of veterinary thermographic images

• Research the use of neural networks for the analysis of veterinary thermographic images

• Investigate use of genetic algorithms for the analysis of veterinary thermographic images

• Research methods to determine the “best” feature set(s)

• Enhance the graphical user interface (GUI) for the automatic pattern classification and application-specific software

• Explore finding of feature vector “norms” for various pathologies/abnormalities

• Explore multimodal image analysis

• Investigate development of automatic preprocessing software for thermographic images

• Investigate development of automatic segmentation software for thermographic images

• Observe first hand the image acquisition process to explore standardization of this process and observation of the methods

• Consult the veterinarians to learn more regarding analysis and diagnostic uses of the thermographic images

• Confer with the veterinarians and/or technicians in the development and use of the application-specific software

References

1. Veterinary Thermographic Image Analysis, Project Number 7-64878, Report Number 4878-1, December 22, 2006

2. Veterinary Thermographic Image Analysis, Project Number 7-64878, Report Number 4878-2, August 31, 2007

3. Veterinary Thermographic Image Analysis, Project Number 7-64878, Report Number 4878-3, January 23, 2008

4. Computer Imaging: Digital Image Analysis and Processing , Scott E Umbaugh, The CRC Press, Boca Raton, FL, January 2005

5. Pattern Recognition Engineering, Morton Adler and Eric Smith, John-Wiley & Sons, NY, 1993

6. Pattern Classification, 2nd Edition, Richard Duda, Peter Hart, David Stork, John-Wiley & Sons, NY, 2001

7. Pattern Recognition, Sergios Theodoris and Konstantinos Koutroumbas, Academic Press, NY, 2006

Appendix –Result File Names

The resultant files are not included here, but are in text files with the following file names. The files are available upon request. Each file is approximately 25 megabytes.

Result Data Files

|Experiment |File names |

|Experiment Set #1 | Exp1-A1-A1D-MinMax-results.txt |

| |    Exp1-A1-A1D-SoftMax-results.txt |

| |    Exp1-A1-A1D-StdNorm-results.txt |

|Experiment Set #2 | Exp2-A1-A1D-MinMax-results.txt |

| |    Exp2-A1-A1D-SoftMax-results.txt |

| |    Exp2-A1-A1D-StdNorm-results.txt |

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download