System for assessing radiation treatment plan segmentations

• US 8,081,813 B2
• Filed: 05/29/2009
• Issued: 12/20/2011
• Est. Priority Date: 05/30/2008
• Status: Active Grant

• Current Assignee: Standard Imaging Incorporated
• Original Assignee: Standard Imaging Incorporated
• Inventors: Nelms, Benjamin E., Miller, Neal R., Manion, Jeffrey H.
• Primary Examiner(s): Nguyen, Sang
• Application Number:
  US12/475,139
• Publication Number:
  US 20090297014A1
• Time in Patent Office: 935 Days
• Field of Search: 382128-132, 382/173, 382/199, 378/65, 378/62, 600/425, 250/252.1
• US Class Current: 382/132
• CPC Class Codes: G06F 18/22 Matching criteria, e.g. proximity measures
• CPC Class Codes: G06T 2207/10072 Tomographic images
• CPC Class Codes: G06T 2207/30004 Biomedical image processing
• CPC Class Codes: G06T 7/11 Region-based segmentation
• CPC Class Codes: G06V 10/751 Comparing pixel values or logical combinations thereof, or feature values having positional relevance, e.g. template matching
• CPC Class Codes: G06V 10/761 Proximity, similarity or dissimilarity measures
• CPC Class Codes: G06V 2201/03 Recognition of patterns in medical or anatomical images
• CPC Class Codes: G16H 20/40 relating to mechanical, radiation or invasive therapies, e.g. surgery, laser therapy, dialysis or acupuncture
• CPC Class Codes: G16H 30/20 for handling medical images, e.g. DICOM, HL7 or PACS

Abstract

Segmentations used to describe structures to be treated by radiotherapy are evaluated by converting the segmentations into volume models and examining volume elements that are extra or missing in the volume model of the second segmentation with respect to the volume model of the first segmentation. This characterization of volume elements may be displayed graphically to show differences in segmentations for training or evaluation purposes and may be quantified by a metric method tallying volume elements as optionally weighted by distance from volume elements shared by the segmentation.

Claims

1. An apparatus for automatically assessing radiation therapy segmentations comprising an electronic computer executing a stored program to:

(a) receive a first and second electronic file each providing data points describing different three-dimensional surfaces circumscribing a structure in a human or animal patient intended for radiation therapy;

(b) generate a first and second volume model defined by the different three-dimensional surfaces of the first and second electronic file respectively;

(c) compare the first and second volume model to identify;

(i) common volume elements in common to both of the first and second volume models;

(ii) missing volume elements of the first volume model that are not in the second volume model; and

(iii) extra volume elements of the second volume model that are not in the first volume model; and

(d) output an indication of a conformance between the three-dimensional surfaces circumscribing the structure defined by the first and second electronic files based on a metric method of measuring volumes of the missing volume elements and extra volume elements;

whereby segmentation represented by the second electronic file is evaluated.

2. The apparatus of claim 1 wherein the first and second volume models are constructed by identifying a set of voxels within the three-dimensional surfaces, and wherein the step of comparing the first and second volume models evaluates each voxel of a union of the set of voxels of the three-dimensional surfaces on a voxel by voxel basis to measure the missing and extra volume elements.

3. The apparatus of claim 1 wherein the electronic computer further includes a graphic display screen and wherein the stored program further displays a cross-sectional image through the first and second volume models along a user-defined cross-sectional plane separately identifying the common volume elements, missing volume elements, and extra volume elements by different colors.

4. The apparatus of claim 3 wherein the common volume elements are colored green, the missing volume elements colored blue and the extra volume elements colored red.

5. The apparatus of claim 3 wherein the stored program further receives a third electronic file providing a cross-sectional image of patient tissue at the user defined cross-sectional plane, and wherein cross-sectional images through the first and second volume models are displayed superimposed on the cross-sectional image of patient tissue obtained from a third electronic file.

6. The apparatus of claim 1 wherein the metric method provides a weighted sum of missing volume elements and extra volume elements normalized by a number of common volume elements so that the metric score decreases with increasing missing volume elements and extra volume elements.

7. The apparatus of claim 1 wherein the step of comparing the first and second volume models also identifies a distance measure for each missing volume element from a closest common volume element and a distance measure for each extra volume element from a closest common volume element wherein the distance measure provides a weighting of the volume of the missing volume elements and extra volume elements in the metric method.

8. The apparatus of claim 1 wherein the electronic computer further includes a user input device for accepting a representation of the metric method to allow a user to set the metric method.

9. The apparatus of claim 8 wherein the metric method is a combination of:

(i) a constant value,

(ii) a linear function of a number of error volume elements with distance, and

(iii) an exponential function of the number of error volume elements with distance;

wherein the error volume elements are missing volume elements or extra volume elements.

10. The apparatus of claim 1 wherein the electronic computer further includes a graphic display screen and wherein execution of the stored program produces a histogram displaying cumulative missing volume elements as a function of distance ranges and cumulative extra volume elements as the function of distance ranges.

11. The apparatus of claim 1 wherein the first and second electronic files provide text strings identifying the structure and wherein the stored program provides for a selection of the metric methods used for the measurement from a set of metric methods using a table mapping the text strings identifying the structure of the first and second electronic files to one of the set of metric methods to be used as the metric method.

12. The apparatus of claim 1 wherein the electronic computer further includes a graphic display screen and a user input device for accepting function parameters;

and

wherein the stored program executes to display a function of the metric method as a graph together with function parameters entered by the user and to change the graph as the function parameters are changed by the user.

13. A method for automatically assessing radiation therapy segmentations comprising using an electronic computer executing stored program to perform the steps of:

(a) receiving a first and second electronic file each providing data points describing different three-dimensional surfaces circumscribing a structure in a human or animal patient intended for radiation therapy;

(b) generating a first and second volume model defined by the different three-dimensional surfaces of the first and second electronic file respectively;

(c) comparing the first and second volume model to identify;

(i) common volume elements in common to both of the first and second volume models;

(ii) missing volume elements of the first volume model that are not in the second volume model; and

(iii) extra volume elements of the second volume model that are not in the first volume model; and

(d) outputting an indication of a conformance between the three-dimensional surfaces circumscribing the structure defined by the first and second electronic files based on a metric method of measuring volumes of the missing volume elements and extra volume elements;

whereby segmentation represented by the second electronic file is evaluated.

14. The method of claim 13 wherein the first and second volume model are constructed by identifying a set of voxels within the three-dimensional surfaces, and wherein the step of comparing the first and second volume models evaluates each voxel of a union of the set of voxels of the three-dimensional surfaces on a voxel by voxel basis to identify the missing and extra volume elements.

15. The method of claim 14 further including the step of displaying a cross-sectional image through the first and second volume models along a user-defined cross-sectional plane separately identifying the common volume elements, missing volume elements, and extra volume elements by different colors.

16. The method of claim 13 wherein the common volume elements are displayed as green, the missing volume elements are displayed as blue and the extra volume elements are displayed as red.

17. The method of claim 16 wherein the common volume elements, missing volume elements and extra volume elements are displayed superimposed on the cross-sectional image of patient tissue showing the structure.

18. The method of claim 14 wherein the step of comparing the first and second volume models also identifies a distance of each missing volume element from a closest common volume element and a distance of each extra volume element from a closest common volume element and wherein the measure is based on a metric method of the missing volume elements and extra volume elements weighted by the respective distances.