Compositional analysis system

• US 6,415,048 B1
• Filed: 04/07/1995
• Issued: 07/02/2002
• Est. Priority Date: 10/12/1993
• Status: Expired due to Fees

• Current Assignee: Schneider Advanced Technologies, Inc.
• Original Assignee: Schneider Advanced Technologies, Inc.
• Inventors: Schneider, Michael Bret
• Primary Examiner(s): Bella, Matthew C.
• Application Number:
  US08/418,986
• Time in Patent Office: 2,643 Days
• Field of Search: 382/128, 382/130, 382/151, 382/155, 382/203, 382/206, 382/218, 382/131, 382/132, 382/154, 378/19, 378/20, 378/4, 378/901
• US Class Current: 382/131
• CPC Class Codes: G06T 7/0012 Biomedical image inspection

Abstract

An improved method of data and image analysis has been described with reference to specific embodiments. The present invention identifies specific portions of a data file and provides the capability of mathematically comparing those portions with other portions. In one embodiment of the present invention, the user is given the option of reformatting the data file. The user is further given the option of having the processor interpret the result of the analysis step.

Claims

1. In a computer system comprising a processor, a method of performing data analysis on an image of a 3-dimensional object, the image comprising a series of 2-dimensional image slices, the method comprising:

inputting the image into the computer system;

using the processor to analyze each image slice of the image and to divide each image slice into at least one 2-dimesional region that corresponds to a 3-dimensional region within the object to investigate a feature of said 3-dimensional region, wherein the analysis step comprises the steps of;

inputting into the processor criteria corresponding to characteristics of a known region; and

comparing said criteria with each image slice in order to identify said at least one 2-dimensional region;

using the processor to derive information describing said at least one 2-dimensional region in each image slice;

using the processor to combine said derived information for said at least one 2-dimensional region from the image slices to determine said feature of said 3-dimensional region;

using the processor to compare said feature of said 3-dimensional region to a feature of a reference region in order to determine a difference between said 3-dimensional region and said reference region;

using the processor to interpret said difference to form a conclusion as to a likely cause of said difference; and

using the processor to communicate said conclusion as to said likely cause to a user.

2. The method of claim 1, wherein said criteria comprises universal classification criteria applicable to each image slice.

3. The method of claim 1, wherein said criteria comprises slice-specific classification criteria applicable only to one or more but not all of said image slices.

4. The method of claim 1, wherein said comparing step includes an artificial intelligence technique selected from the group consisting of describe-and-match techniques, decision tree techniques, neural net techniques, rule-chaining techniques, ratio pairing techniques, and multisurface pattern separation techniques.

5. The method of claim 1, wherein said communication step further comprises the step of generating a colored image of said difference.

6. The method of claim 1, wherein said communication step comprises the step of generating a textual description of a conclusion.

7. The method of claim 1, wherein said conclusion is a medical diagnosis.

8. The method of claim 1, wherein said combining step comprises the step of adding said derived information together.

9. The method of claim 1, wherein said at least one 2-dimensional region is a structure or an activity level.

10. The method of claim 1, wherein said feature is a size, volume, shape, or intensity.

11. A computer implemented method of performing data analysis, comprising:

inputting at least one input data file describing an object, said object having a plurality of regions;

determining an identity of a known region in said at least one input data file;

comparing features of said known region to features of a reference object in order to determine differences between said known region and said reference object, wherein said known region and said reference object are identified as representing a same type of structure;

generating a conclusion of a cause of said differences between said known region and said reference object; and

communicating said differences and said conclusion to a user.

12. The method of claim 11, wherein said features include means pixel intensities of said known region and said reference object.

13. The method of claim 11, wherein said features include volumes of said known region and said reference object.

14. A computer implanted method of performing data analysis, comprising:

inputting at least one input data file describing an object, said object having a plurality of

determining an identity of a known region in said at least one input data file;

comparing means pixel intensity of said known region to mean pixel intensity of a reference object in order to determine a difference in mean pixel intensity between said known region and said reference object, wherein said known region and said reference object are identified as representing a same type of structure;

generating a medical conclusion of a cause of said mean pixel intensity difference between said known region and said reference object; and

communicating said mean pixel intensity difference and said medical conclusion to a user.

15. A computer implemented method of performing data analysis;

comprising;

inputting at least one input data file describing an object, said object having a plurality of regions;

determining an identity of a known region in said at least one input data file;

comparing a volume of said known region to a volume of a reference object in order to determine a difference in volume between said known region and said reference object, wherein said known region and said reference object are identified as representing a same type of structure;

generating a medical conclusion of a caus of said difference in volume between said known region and said reference object, and

communicating said difference in volume and said medical conclusion to a user.