Interactive segmentation

Patent metadata

Field	Value
Publication number	US-10147185-B2
Application number	US-201515509925-A
Country	US
Kind code	B2
Filing date	Sep 8, 2015
Priority date	Sep 11, 2014
Publication date	Dec 4, 2018
Grant date	Dec 4, 2018

Abstract

A method for three-dimensional interactive segmentation, including: receiving a three-dimensional medical image of an interior volume of a patient's body; automatically performing three dimensional segmentation on the three dimensional medical image to detect and define a region of interest, wherein the performing of the three dimensional segmentation comprises automatically determining a boundary defining the region of interest; receiving from a user spatial information indicating one or more regions of disagreement in the three-dimensional medical image with respect to the determined boundary; and updating the three dimensional segmentation of the three dimensional medical image based on the spatial information received from the user, wherein the updating comprises updating the determined boundary based on the spatial information to redefine the area of interest.

First claim

What is claimed is:

1. A method comprising using at least one hardware processor for:
receiving a three-dimensional medical image of an interior volume of a patient's body;
automatically performing three dimensional segmentation on the three dimensional medical image to detect and define a region of interest in the three dimensional medical image, wherein the performing of the three dimensional segmentation comprises automatically determining a boundary defining the region of interest;
receiving an input from a user with respect to the three dimensional segmentation of the three dimensional medical image, wherein the input comprises spatial information indicating one or more regions of disagreement in the three-dimensional medical image with respect to the determined boundary; and
updating the three dimensional segmentation of the three dimensional medical image based on the input received from the user, wherein the updating comprises updating the determined boundary based on the spatial information to redefine the area of interest, and wherein the updating is performed according to weights determining the relative contribution of each one of the three dimensional segmentation and the input from the user to the updating of the three dimensional segmentation, thereby determining the extent of influence of the input from the user on the determined boundary.

2. The method according to claim 1 , wherein the three dimensional medical image is of an imaging modality type selected from the group consisting of: Computerized Tomography (CT), Magnetic Resonance Imaging (MM) and ultrasound.

3. The method according to claim 1 , further comprising using said at least one hardware processor for performing three dimensional visualization of the segmented region of interest.

4. The method of claim 1 , further comprising using said at least one hardware processor for estimating at least one measurable feature of the region of interest, wherein the at least one measurable feature is selected from the group consisting of: volume, length, width, height, pose, location and texture.

5. The method of claim 1 , wherein the receiving of the input from the user and the updating of the three dimensional segmentation are performed in an iterative manner.

6. The method of claim 1 , wherein the receiving of the input from the user is performed via a Graphical User Interface (GUI), and wherein the GUI presents to the user one or more views of the 3D medical image selected from the group consisting of: slice views in one or more body planes and a 3D view.

7. The method of claim 6 , wherein the spatial information comprises anchors indicated by the user via a pointing device on the one or more views of the 3D medical image presented by the GUI.

8. The method of claim 1 , wherein the automatically performing of the three dimensional segmentation, and the updating of the three dimensional segmentation, comprise using a probabilistic model.

9. The method of claim 8 , wherein the probabilistic model is based on Gaussian Mixture Model (GMM), and wherein the GMM is used to model a distribution of the intensities of the 3D medical image within and outside of the region of interest.

10. The method of claim 8 , wherein the performing of the three dimensional segmentation further comprises defining a cost functional, the cost functional comprising:
i) an image likelihood term representing the three dimensional segmentation, and
ii) a user interaction term representing the input from the user,
assigning the weight of the weights determining the relative contribution of the three dimensional segmentation to the image likelihood term,
assigning the weight of the weights determining the relative contribution of the input from the user to the user interaction term, and
setting the weight of the weights assigned to the user interaction term to zero.

11. The method of claim 10 , wherein the performing of the three dimensional segmentation and the updating the three dimensional segmentation further comprise finding a segmentation that optimizes the cost functional by using a gradient descent process.

12. A medical system comprising:
a medical imaging device;
at least one hardware processor configured to:
i) automatically perform three dimensional segmentation on the three dimensional medical image to detect and define a region of interest in the three dimensional medical image, wherein the performing of the three dimensional segmentation comprises automatically determining a boundary defining the region of interest;
ii) receive an input from a user with respect to the three dimensional segmentation of the three dimensional medical image, wherein the input comprises spatial information indicating one or more regions of disagreement in the three-dimensional medical image with respect to the determined boundary; and
iii) update the three dimensional segmentation of the three dimensional medical image based on the input received from the user, wherein the updating comprises updating the determined boundary based on the spatial information to redefine the area of interest, and wherein the updating is performed according to weights determining the relative contribution of each one of the three dimensional segmentation and the input from the user to the updating of the three dimensional segmentation, thereby determining the extent of influence of the input from the user on the determined boundary.

13. The medical system according to claim 12 , wherein said medical imaging device is selected from the group consisting of: a Computerized Tomography (CT) device, a Magnetic Resonance Imaging (MRI) device and an ultrasound imaging device.

14. The medical system according to claim 12 , wherein said at least one hardware processor is further configured to perform three dimensional visualization of the segmented region of interest.

15. The medical system of claim 12 , wherein said at least one hardware processor is further configured to estimate at least one measurable feature of the region of interest, wherein the at least one measurable feature is selected from the group consisting of: volume, length, width, height, pose, location and texture.

16. The medical system of claim 12 , wherein the receiving of the input from the user and the updating of the three dimensional segmentation are performed in an iterative manner.

17. The medical system of claim 12 , wherein the receiving of the input from the user is performed via a GUI, and wherein the GUI presents to the user one or more views of the 3D medical image selected from the group consisting of: slice views in one or more body planes and a 3D view.

18. The medical system of claim 17 , wherein the spatial information comprises anchors indicated by the user via a pointing device on the one or more views of the 3D medical image presented by the GUI.

19. The medical system of claim 12 , wherein the automatically performing of the three dimensional segmentation, and the updating of the three dimensional segmentation, comprise using a probabilistic model, wherein the probabilistic model is based on Gaussian Mixture Model (GMM), and wherein the GMM is used to model a distribution of the intensities of the 3D medical image within and outside of the region of interest.

20. The medical system of claim 19 , wherein the performing of the three dimensional segmentation further comprises defining a cost functional, the cost functional comprising:
i) an image likelihood term representing the three dimensional segmentation, and
ii) a user

Assignees

• B G Negev Technologies And Applications Ltd At Ben Gurion Univ
• Mor Research Applic Ltd

Inventors

• Riklin Raviv Tamar
• Shelef Ilan
• Hershkovich Itzhak
• Chitrit Ohad

Classifications

• G06F19/321

  Physics · mapped topic

• G06T7/12

  Edge-based segmentation · CPC title

• G16H50/50

  for simulation or modelling of medical disorders · CPC title

• G06T2207/10081

  Computed x-ray tomography [CT] · CPC title

• G06T7/0012 Primary

  Biomedical image inspection · CPC title