Method and system for determining a phenotype of a neoplasm in a human or animal body

• US 9,811,904 B2
• Filed: 04/17/2014
• Issued: 11/07/2017
• Est. Priority Date: 04/19/2013
• Status: Active Grant

• Current Assignee: Stichting Maastricht Radiation Oncology "Maastro-Clinic"
• Original Assignee: Stichting Maastricht Radiation Oncology "Maastro-Clinic"
• Inventors: Lambin, Philippe, Aerts, Hugo Johannes Wilhelmus Louis
• Primary Examiner(s): Akhavannik, Hadi
• Application Number:
  US14/783,974
• Publication Number:
  US 20160078613A1
• Time in Patent Office: 1,300 Days
• Field of Search: None
• CPC Class Codes: G06T 2207/10081 Computed x-ray tomography [CT]
• CPC Class Codes: G06T 2207/30064 Lung nodule
• CPC Class Codes: G06T 2207/30096 Tumor; Lesion
• CPC Class Codes: G06T 7/0012 Biomedical image inspection
• CPC Class Codes: G06V 20/695 Preprocessing, e.g. image segmentation

Abstract

The present invention relates to a decision support system and an image analysis method for providing information for enabling determination of a phenotype of a neoplasm in a human or animal body for enabling prognostication, comprising the steps of: receiving, by a processing unit, image data of the neoplasm; and deriving, by the processing unit, a plurality of image feature parameter values from the image data, said image parameter values relating to image features associated with the neoplasm; and deriving, by said processing unit using a signature model, one or more neoplasm signature model values associated with the neoplasm from said image feature parameter values, wherein said signature model includes a functional relation between or characteristic values of said image feature parameter values for deriving said neoplasm signature model values.

Claims

1. An image analysis method for providing information for enabling determination of a phenotype of a neoplasm in a human or animal body for enabling prognostication, comprising the steps of:

receiving, by a processing unit, image data of the neoplasm; and

deriving, by the processing unit, a plurality of image feature parameter values from the image data, said image parameter values relating to image features associated with the neoplasm; and

deriving, by said processing unit using a signature model, one or more neoplasm signature model values associated with the neoplasm from said image feature parameter values, wherein said signature model includes a functional relation between or characteristic values of said image feature parameter values for deriving said neoplasm signature model values therefrom;

wherein the image feature parameter values are indicative of image feature parameters, wherein the signature model includes at least all of the image feature parameters from a group comprising;

gray-level non-uniformity, and wavelet high-low-high gray-level run-length gray-level non-uniformity.

2. The image analysis method according to claim 1, wherein the signature model further includes at least all of the image feature parameters from a group comprising:

statistics energy, and shape compactness.

3. The image analysis method according to claim 1, wherein the method further comprises the steps of:

obtaining, by said processing unit from a memory, said signature model comprising one or more signature selector values associated with the image features, wherein the signature selector values indicate whether the associated image features are comprised by the signature model; and

multiplying for the at least one signature model, by the processing unit, the image feature parameter values with the associated signature selector values for obtaining said one or more neoplasm signature model values associated with the neoplasm.

4. The image analysis method according to claim 1, further comprising a step of comparing, by the processing unit, the neoplasm signature model values to at least one signature model reference value for the at least one signature model, for associating the neoplasm with the phenotype.

5. The image analysis method according to claim 1, wherein the signature selector values further comprise one or more weighting values associated with the image features comprised by the signature model, for weighing the image features in the signature model.

6. The image analysis method according to claim 1, further comprising a step of calculating, by the processing unit, a neoplasm signature model score as a function of the one or more neoplasm signature model values, wherein the step of comparing includes comparing the neoplasm signature model score with a reference score.

7. The image analysis method according to claim 1, wherein the at least one signature model comprises a plurality of distinct signature models.

8. The image analysis method according to claim 1, wherein the image feature parameters further include at least one element of a group comprising:

first-order gray level statistics obtained from image pixels or areas of the image from the image data;

second-order gray level statistics obtained from co-occurrence matrices of the image data;

run-length gray level statistics, short run emphasis, long run emphasis, run percentage; and

shape and size based features.

9. The image analysis method according to claim 2, wherein in accordance with the signature model, the image feature parameter statistics energy has an associated absolute weighting value within a range of 1.0e−

20 through 1.0e−

5;

the image feature parameter shape compactness has an associated absolute weighting value within a range of 1.0e−

7through 1.0e−

1;

the image feature parameter gray-Level non-uniformity has an associated absolute weighting value within a range of 1.0e−

9 through 1.0e−

1; and

the image feature parameter wavelet high-low-high gray-level run-length gray-level non-uniformity has an associated absolute weighting value within a range of 1.0e−

9 through 1.0e−

1.

10. The image analysis method according to claim 1, wherein the image data is received using an imaging method selected from a group comprising magnetic resonance imaging, computed tomography, positron emission tomography, single-photon emission computed tomography, ultrasonography, thermography, and photo-acoustic imaging.

11. The image analysis method in accordance with claim 1, any of the previous claims, wherein the step of receiving image data comprises the steps of receiving first image data of the neoplasm at a first moment in time and receiving second image data of the neoplasm at a second moment in time, and wherein the steps of deriving the image feature parameter values, obtaining the signature model, and multiplying the image feature parameter values with the associated signature selector values is performed for said first and second image data, further comprising a step of determining a difference between the neoplasm signature model values of the first and second image data.

16. The image analysis method according to claim 8, wherein the first order gray level statistics are selected from minimum intensity, maximum intensity, mean intensity, intensity range, intensity variance, intensity standard deviation, skewness, kurtosity and entropy.

17. The image analysis method according to claim 8, wherein the second-order gray level statistics are selected from contrast, correlation between neighboring image axels or pixels, energy, homogeneity, inverse difference moment, sum average, sum variance, and sum entropy.

18. The image analysis method according to claim 8, wherein the shape and size based features are selected from perimeter, cross-sectional area, major axis length, maximum diameter, and volume.

12. A decision support system arranged for performing an image analysis method for providing information for enabling determination of a phenotype of a neoplasm in a human or animal body for enabling prognostication, said system comprising an input connected to a processing unit for receiving by the processing unit image data of the neoplasm;

wherein the processing unit is further arranged for deriving a plurality of image feature parameter values from the received image data, said image parameter values relating to image features associated with the neoplasm, wherein the processing unit is connected to a memory for obtaining therefrom at least one signature model comprising one or more signature selector values associated with the image features, wherein the signature selector values indicate whether the associated image features are comprised by the signature model, and wherein the processing unit is arranged for multiplying for the at least one signature model the image feature parameter values with the associated signature selector values for obtaining one or more neoplasm signature model values associated with the neoplasm;

wherein the image feature parameter values are indicative of image feature parameters, wherein the signature model includes at least all of the image feature parameters from a group comprising;

statistics energy, shape compactness, gray-level non-uniformity, wavelet high-low-high gray-level run-length gray-level non-uniformity.

13. The decision support system according to claim 12, wherein the processing unit is further arranged for comparing the neoplasm signature model values to at least one signature model reference value for the at least one signature model, for associating the neoplasm with the phenotype, the system further comprising an output for providing an indicator value indicative of a result of said step of associating.

14. The decision support system according to claim 12, wherein the signature selector values further comprise one or more weighting values associated with the image features comprised by the signature model, for weighing the image features in the signature model, and wherein in accordance with the signature model, the image feature parameter statistics energy has an associated absolute weighting value within a range of 1.0e−

20through 1.0e−

5;

the image feature parameter shape compactness has an associated absolute weighting value within a range of 1.0e−

7 through 1.0e−

1;

the image feature parameter gray-level non-uniformity has an associated absolute weighting value within a range of 1.0e−

9 through 1.0e−

1; and

the image feature parameter wavelet high-low-high gray-level run-length gray-level non-uniformity has an associated absolute weighting value within a range of 1.0e−

9 through 1.0e−

1.

15. A non-transistory computer-readable medium comprising computer-executable instructions which, when run on a computer, are arranged for performing an image analysis method for providing information for enabling determination of a phenotype of a neoplasm in a human or animal body for enabling prognostication, the method comprising the steps of:

receiving, by a processing unit, the image data of the neoplasm; and

deriving, by the processing unit, a plurality of image feature parameter values from the image data, said image parameter values relating to image features associated with the neoplasm; and

deriving, by said processing unit using a signature model, one or more neoplasm signature model values associated with the neoplasm from said image feature parameter values, wherein said signature model includes a functional relation between or characteristic values of said image feature parameter values deriving said neoplasm signature model values therefrom,

wherein the image feature parameter values are indicative of image feature parameters, and

wherein the signature model includes at least all of the image feature parameters from a group comprising;

gray-level non-uniformity, and

wavelet high-low-high gray-level run-length gray-level non-uniformity.