Differential atlas for cancer assessment

Patent metadata

Field	Value
Publication number	US-9851421-B2
Application number	US-201514960539-A
Country	US
Kind code	B2
Filing date	Dec 7, 2015
Priority date	Jan 5, 2015
Publication date	Dec 26, 2017
Grant date	Dec 26, 2017

Abstract

Methods and apparatus associated with producing a quantification of differences associated with biochemical recurrence (BcR) in a region of tissue demonstrating prostate cancer (PCa) are described. One example apparatus includes a set of logics, and a data store that stores a set of magnetic resonance (MR) images acquired from a population of subjects. The set of logics includes an image acquisition logic that acquires a diagnostic image of a region of tissue in a patient demonstrating PCa, a morphology logic that extracts a shape feature, a volume feature, or an intensity feature from the diagnostic image or from a member of the set of MR images, a differential atlas construction logic that constructs a statistical shape differential atlas from the set of MR images, and a quantification logic that produces a quantification of differences based on the shape feature, the volume feature, or the intensity feature, and the differential atlas.

First claim

What is claimed is:

1. A non-transitory computer-readable storage medium storing computer executable instructions that when executed by a computer control the computer to perform a method for predicting biochemical recurrence (BcR) in a region of tissue demonstrating prostate cancer (PCa) pathology, the method comprising:
accessing a set of magnetic resonance (MR) images acquired from a population of patients, where the population includes a set of subpopulations, where a member of the set of MR images includes a prostate capsule shape attribute, a prostate capsule volume attribute, a central gland (CG) shape attribute, or a CG volume attribute;
constructing a statistical shape atlas from the set of MR images;
accessing an MR image of a region of prostate tissue in a patient demonstrating cancerous pathology, where the MR image of the region of prostate tissue has a prostate capsule shape attribute, a prostate capsule volume attribute, a CG shape attribute, or a CG volume attribute;
producing a quantification of differences between the MR image of the region of prostate and the statistical shape atlas based, at least in part, on a comparison of the prostate capsule shape attribute of the MR image of the region of prostate tissue demonstrating cancerous pathology, the prostate capsule volume attribute of the MR image of the region of prostate tissue demonstrating cancerous pathology, the CG shape attribute of the MR image of the region of prostate tissue demonstrating cancerous pathology, or the CG volume attribute of the MR image of the region of prostate tissue demonstrating cancerous pathology, with the statistical shape atlas;
computing a BcR probability score based, at least in part, on the quantification of differences, and
controlling a computer aided diagnosis (CADx) system to generate a classification of the region of tissue in the image based, at least in part, on the BcR probability score or the quantification of differences.

2. The non-transitory computer-readable storage medium of claim 1 , where the set of subpopulations includes a PCa positive (PCa+) subpopulation, a Pea negative (PCa−) subpopulation, or a normal subpopulation.

3. The non-transitory computer-readable storage medium of claim 2 , where the PCa− subpopulation demonstrates increased prostate specific antigen (PSA).

4. The non-transitory computer-readable storage medium of claim 1 , where the subpopulations are acquired across a plurality of institutions.

5. The non-transitory computer-readable storage medium of claim 1 , the method comprising automatically annotating the prostate capsule or the CG in a member of the set of MR images or in the MR image of the region of prostate tissue demonstrating cancerous pathology.

6. The non-transitory computer-readable storage medium of claim 1 , where the CG includes a prostate central zone and a prostate transitional zone.

7. The non-transitory computer-readable storage medium of claim 1 , where a member of the set of MR images or the MR image of the region of prostate tissue demonstrating cancerous pathology is a 1.5 Tesla (T) T2 weighted (T2w) MR image or a 3T T2w MR image.

8. The non-transitory computer-readable storage medium of claim 1 , where a first member of the set of MR images is acquired using a first set of MR acquisition parameters having a first set of values, and a second member of the set of MR images is acquired using a second, different set of MR acquisition parameters having a second, different set of values.

9. The non-transitory computer-readable storage medium of claim 8 , where the first set of MR acquisition parameters and the second set of MR acquisition parameters includes pixel dimensions, resolution, or slice spacing.

10. The non-transitory computer-readable storage medium of claim 1 , where the member of the set of MR images or the MR image is acquired using a surface coil or using an endorectal coil.

11. The non-transitory computer-readable storage medium of claim 2 , where constructing the statistical shape atlas comprises:
annotating a prostate capsule represented in a member of the set of MR images;
annotating a central gland (CG) represented in the member of the set of MR images;
upon determining that the member of the set of MR images was acquired using an endorectal coil, bias field correcting the member of the set of MR images that was acquired using an endorectal coil;
applying template-based anatomically constrained registration to a subset of the set of MR images associated with a subpopulation, where the registration is based, at least in part, on a prostate capsule shape attribute, a prostate capsule volume attribute, a CG shape attribute, a CG volume attribute, or on a T2w intensity associated with the member of the set of MR images;
computing a spatial median of the prostate capsule shape attribute or the CG shape attribute for the subset of the set of MR images associated with a subpopulation, or a spatial median of the prostate capsule volume attribute or the CG volume attribute for the subset;
progressively updating a template based, at least in part, on the spatial median of the prostate capsule shape attribute, the spatial median of the CG shape attribute, the spatial median of the prostate capsule volume attribute, or the spatial median of the CG volume attribute;
progressively increasing the complexity of an optimized three dimensional (3D) transformation, where the 3D transformation is a translation;
correcting an atlas shift between an atlas associated with a first subpopulation and an atlas associated with a second subpopulation; and
applying a per-voxel comparison between members of a subset of the set of MR images associated with the first subpopulation and members of a subset of a set of MR images associated with the second, different subpopulation.

12. The non-transitory computer-readable storage medium of claim 11 , where progressively increasing the complexity of the optimized three dimensional (3D) transformation includes:
converting the optimized 3D transformation from a translation to an affine translation, or converting the optimized 3D transformation from an affine translation to an elastic deformation.

13. The non-transitory computer-readable storage medium of claim 12 , where correcting an atlas shift between the atlas associated with the first subpopulation and the atlas associated with the second subpopulation includes optimizing the affine translation relative to the first subpopulation and the second subpopulation, where the affine translation optimization is based, at least in part, on the prostate capsule shape attribute or the CG shape attribute.

14. The non-transitory computer-readable storage medium of claim 13 where the per-voxel comparison includes a non-parametric Wilcoxon test.

15. The non-transitory computer-readable storage medium of claim 14 , where the per-voxel comparison further includes applying multiple comparison correction using a Bonferonni correction.

16. A method for producing a quantification of differences between a diagnostic image of a region of prostate tissue and a statistical shape atlas, the method comprising:
accessing a statistical shape atlas, where the statistical shape atlas includes a set of registered medical images of prostate tissue acquired from a population of subjects, the population including a subpopulation of prostate cancer positive ( PCa+) subjects, a subpopulation of, prostate cancer negative (PCa−) subjects, and a subpopulation of normal subjects, where a prostate represented in an image in the set of medical images includes a shape and a volume;
accessi

Assignees

• Univ Case Western Reserve

Inventors

• Madabhushi Anant
• Rusu Mirabela

Classifications

• G01R33/34084 Primary

  implantable coils or coils being geometrically adaptable to the sample, e.g. flexible coils or coils comprising mutually movable parts · CPC title

• G06V10/772 Primary

  Determining representative reference patterns, e.g. averaging or distorting patterns; Generating dictionaries · CPC title

• G06F18/2433

  Single-class perspective, e.g. one-against-all classification; Novelty detection; Outlier detection · CPC title

• G06F18/28

  Determining representative reference patterns, e.g. by averaging or distorting; Generating dictionaries · CPC title

• G06K9/6284

  Physics · mapped topic