COMPUTER AIDED DIAGNOSIS APPARATUS AND METHOD BASED ON SIZE MODEL OF REGION OF INTEREST

• US 20180047164A1
• Filed: 10/27/2017
• Published: 02/15/2018
• Est. Priority Date: 09/16/2014
• Status: Active Grant

• Current Assignee: Samsung Electronics Co. Ltd.
• Original Assignee: Samsung Electronics Co. Ltd.
• Inventors: RYU, Seung Woo, KIM, Won Sik
• Granted Patent: US 10,664,968 B2
• Time in Patent Office: Days
• CPC Class Codes: A61B 8/13 Tomography A61B8/10, A61B8/12 take precedence; tomography for radiation diagnosis A61B6/02
• CPC Class Codes: A61B 8/56 Details of data transmission or power supply
• CPC Class Codes: G06T 2207/10016 Video; Image sequence
• CPC Class Codes: G06T 2207/10132 Ultrasound image
• CPC Class Codes: G06T 2207/30096 Tumor; Lesion
• CPC Class Codes: G06T 7/0012 Biomedical image inspection
• CPC Class Codes: G06T 7/12 Edge-based segmentation
• CPC Class Codes: G06T 7/143 involving probabilistic approaches, e.g. Markov random field [MRF] modelling

Abstract

A Computer Aided Diagnosis (CAD) apparatus and CAD method for detection of ROIs based on an ROI size transition model. The CAD apparatus includes: an image receiver configured to sequentially receive images; a region of interest (ROI) acquirer configured to acquire an ROI from a current image based on an ROI size transition model; and an ROI output configured to output visual information indicating the acquired ROI.

Claims

1. A computer program product comprising a non-transitory computer-readable storage medium configured to store a computer readable program comprising instructions configured to, when executed by a computing device, cause the computing device to:

detect at least one region of interest (ROI) from a current image among sequentially received images,

determine whether at least one ROI, that is erroneously detected, of the detected at least one ROI exists based on an ROI size transition information, and

output visual information indicating the at least one ROI except for the erroneously detected at least one ROI based on a result of the determination.

2. The computer program product of claim 1, wherein the sequentially received images are real-time ultrasound images acquired through a probe in units of frames.

3. The computer program product of claim 1, wherein the computing device is further configured to:

determine a size state of the detected at least one ROI by matching a size of the detected at least one ROI with the ROI size transition information; and

determine that the detected at least one ROI is erroneously detected, in response to determining that a size state of a previous image on the ROI size transition information is not capable of being transformed to the determined size state of the current image.

4. The computer program product of claim 3, wherein the computing device is further configured to not output visual information on the at least one ROI that is erroneously detected.

5. The computer program product of claim 3, wherein the computing device is further configured to:

in response to the at least one ROI being erroneously detected, determine the biggest size state that is possible to transition from the size state of the previous image using the ROI size transition information, and

output visual information to a size corresponding to the biggest size state.

6. The computer program product of claim 1, wherein the computing device is further configured to predict an ROI in the current image using the ROI size transition information, in response to the ROI not being detected from the current image.

7. The computer program product of claim 6, wherein the computing device is further configured to:

determine a size state transitionable from the size state of the ROI on an image before the current image using the ROI size transition information, and

predict the ROI in the current image using the determined size state.

8. The computer program product of claim 1, wherein the ROI size transition information is created using the Markov model learning based on a change in sizes of ROIs in a sequence of images for each interested item, and the ROI size transition information comprises size states, a transition possibility between the size states, transition direction, and transition probability information.

9. The computer program product of claim 1, wherein in order to indicate a location and size of the outputted at least one ROI in the current image, the visual information includes:

first information comprising at least one of square, round, oval, or cross shapes; and

second information comprising at least one of color, types of lines, or thickness of lines of the first information.

10. The computer program product of claim 1, wherein the sequentially received images are configured to a three-dimensional image.

11. A computer aided diagnosis (CAD) method, comprising:

sequentially receiving images;

detecting at least one region of interest (ROI), by a processor, from a current image among the sequentially received images;

determining whether at least one ROI, that is erroneously detected, of the detected at least one ROI exists based on an ROI size transition information; and

outputting visual information that indicates the at least one ROI except for the erroneously detected at least one ROI based on a result of the determination.

12. The method of claim 11, wherein the determining comprises:

determining a size state of the detected at least one ROI by matching a size of the detected at least one ROI with the ROI size transition information; and

determining that the detected at least one ROI is erroneously detected, in response to determining that a size state of a previous image on the ROI size transition information is not capable of being transformed to the determined size state of the current image.

13. The method of claim 12, further comprising:

not outputting visual information on the at least one ROI that is erroneously detected.

14. The method of claim 12, wherein the outputting comprises:

determining the biggest size state that is possible to transition from the size state of the previous image using the ROI size transition information; and

outputting visual information to a size corresponding to the biggest size state.

15. The method of claim 11, further comprising:

predicting an ROI in the current image using the ROI size transition information in response to the ROI not being detected from the current image.

16. The method of claim 15, wherein the predicting of the ROI comprises:

determining a size state transitionable from the size state of the ROI on an image before the current image using the ROI size transition information; and

predicting the ROI in the current image using the determined size state.

17. The method of claim 11, wherein the ROI size transition information is created using the Markov model learning based on a change in sizes of ROIs in a sequence of images for each interested item, and the ROI size transition information comprises size states, a transition possibility between the size states, transition direction and transition probability information.

18. The method of claim 11, wherein the sequentially receiving of the images comprises:

receiving real-time ultrasound images through a image receiver in units of frames.

19. The method of claim 11, wherein the sequentially received images are configured to a three-dimensional image.

20. A computer aided diagnosis (CAD) apparatus, comprising:

a memory configured to store instructions; and

at least one processor, that upon executing the stored instructions, configured to;

detect at least one region of interest (ROI) from a current image among sequentially received images,

determine whether at least one ROI, that is erroneously detected, of the detected at least one ROI exists based on an ROI size transition information, and

output visual information indicating the at least one ROI except for the erroneously detected at least one ROI based on a result of the determination.