Knowledge-based brachytherapy planning based on primitive features computed from anatomical contour data

Patent metadata

Field	Value
Publication number	US-10967199-B2
Application number	US-201716077237-A
Country	US
Kind code	B2
Filing date	Feb 9, 2017
Priority date	Feb 11, 2016
Publication date	Apr 6, 2021
Grant date	Apr 6, 2021

Abstract

Described here are systems and methods for knowledge-based brachytherapy planning. These systems and methods are capable of automatically generating treatment plans for prostate brachytherapy in clinically relevant times. Primitive features computed from anatomical contours of the target organ are used to retrieve a template plan from a database. The template plan is then adjusted in a stochastic search algorithm.

First claim

The invention claimed is:

1. A computer-implemented method for generating a brachytherapy plan, the steps of the method comprising:
(a) providing contour data for a patient to a computer system, the contour data indicating an anatomical contour for a target organ;
(b) computing with the computer system, a primitive feature from the contour data;
(c) selecting a template brachytherapy plan by querying a database using the primitive feature, the database containing stored contour data and stored brachytherapy plan configurations for different previously treated patients; and
(d) generating a brachytherapy plan for the patient by performing with the computer system, a stochastic search algorithm to adjust the selected template brachytherapy plan based on at least one clinical rule.

2. The method as recited in claim 1 , wherein the primitive feature is a symmetry value that indicates a symmetry of the target organ in an anatomical plane.

3. The method as recited in claim 2 , wherein the target organ is a prostate and the anatomical plane is a sagittal plane.

4. The method as recited in claim 3 , wherein the symmetry value is computed by generating a binary mask of the prostate from the contour data, dividing the binary mask into a left lobe area and a right lobe area, and computing the symmetry value as a ratio of an intersection of the left lobe area and the right lobe area to a union of the left lobe area and the right lobe area.

5. The method as recited in claim 4 , wherein dividing the binary mask include calculating a centroid of the binary mask and dividing the binary mask based on the centroid.

6. The method as recited in claim 1 , wherein step (c) includes selecting the template brachytherapy plan based on a comparison of the primitive feature to primitive features extracted from the stored contour data.

7. The method as recited in claim 6 , wherein the database is queried using a multi-dimensional range-query algorithm.

8. The method as recited in claim 7 , wherein the range-query algorithm implements a hierarchy of primitive features to select the template brachytherapy plan from the database.

9. The method as recited in claim 8 , wherein the range-query algorithm is performed in a cascading format for the hierarchy of primitive features, wherein the cascading format implements a cascading pathway based on a decision tree.

10. The method as recited in claim 1 , wherein step (d) includes adjusting the template brachytherapy plan based on the primitive feature.

11. The method as recited in claim 10 , wherein the primitive feature is a symmetry value that indicates a symmetry of the target organ in an anatomical plane, and adjusting the template brachytherapy plan based on the primitive feature includes maximizing the symmetry value.

12. The method as recited in claim 1 , wherein step (d) includes initializing the stochastic search algorithm based on the selected brachytherapy plan template to define a restricted search space for the stochastic search algorithm.

13. The method as recited in claim 1 , wherein step (d) includes identifying dosimetric cold spots in the template brachytherapy plan and updating the template brachytherapy plan using the at least one clinical rule to determine where additional seeds should be added in the template brachytherapy plan.

14. The method as recited in claim 1 , wherein the contour data provided in step (a) also indicates an anatomical contour for at least one organ-at-risk (OAR).

15. The method as recited in claim 1 , wherein the contour data provided in step (a) comprises a series of two-dimensional contours that collectively indicate a three-dimensional contour.

16. The method as recited in claim 15 , wherein each contour in the series of two-dimensional contours is a contour in an axial plane through the target organ.

17. A computer-implemented method for generating a symmetry metric from anatomical contour data computed from medical images that depict a subject, the steps of the method comprising:
(a) providing the anatomical contour data to a computer system, the anatomical contour data indicating an anatomical contour for a target organ in the subject;
(b) computing with the computer system, a binary mask of the target organ using the anatomical contour data;
(c) computing with the computer system, an axis of reflection for the target organ, the axis of reflection defining an axis about which symmetry of the target organ will be measured;
(d) computing with the computer system based on the binary mask, a symmetry metric that provides a quantitative value indicating symmetry of the target organ about the axis of reflection; and
(e) generating a report based on the computed symmetry metric.

18. The method as recited in claim 17 , wherein step (d) includes dividing the binary mask into a first area and a second area using the axis of reflection, and computing the symmetry metric as a ratio of an intersection of the first area and the second area to a union of the first area and the second area.

19. The method as recited in claim 18 , wherein the target organ is a prostate, the first area indicates an area of a left lobe of the prostate in a sagittal plane, and the second area indicates an area of a right lobe of the prostate in the sagittal plane.

Assignees

• Sunnybrook Res Inst

Inventors

• Ravi Ananth
• Nicolae Alexandru

Classifications

• G16H50/20

  for computer-aided diagnosis, e.g. based on medical expert systems · CPC title

• A61N5/103 Primary

  Treatment planning systems · CPC title

• A61B34/10

  Computer-aided planning, simulation or modelling of surgical operations · CPC title

• A61N2005/1041

  using a library of previously administered radiation treatment applied to other patients · CPC title

• G16H20/40

  relating to mechanical, radiation or invasive therapies, e.g. surgery, laser therapy, dialysis or acupuncture · CPC title