Granted Patent B2

US 12,380,554 · App. 17/471,001 · Granted Aug 5, 2025

Inventors: Minki Chung (Seoul, KR); Beomhee Park (Seoul, KR); Seo Taek Kong (Yongin-si, KR); Younjoon Chung (Seoul, KR)

Assignee: VUNO Inc.

G06T7/0012G16H30/40G16H50/20G16H50/30G06T2207/30096

Abstract

Claims (91)

1. A method performed on one or more processors of a computing device, the method comprising:

receiving a body medical image;

detecting one or more lesions from the received body medical image using a lesion detection model;

extracting a body region having anatomical significance from the received body medical image using an anatomical analysis model;

generating anatomical location information for the one or more lesions by matching the detection result for the one or more lesions with the extracted body region; and

generating a diagnosis result for the received body medical image, based on the anatomical location information for the one or more lesions,

wherein the body region having the anatomical significance is a body organ that is likely to be a point of occurrence of the one or more lesions or a body organ that is a target of the detection of the one or more lesions, and

wherein the generating the diagnosis result for the received body medical image, based on the anatomical location information for the one or more lesions comprising:

generating clinical information indicating clinical significance in a situation in which the one or more lesions are related to the anatomical location information,

wherein the clinical information comprising:

an occurrence frequency of lesion for each of regions divided from the body region having the anatomical significance; and

a degree of risk of lesion for each of the divided regions,

wherein the generating the clinical information indicating the clinical significance in the situation in which the one or more lesions are related to the anatomical location information comprising:

changing a confidence score determined in the step of detecting the one or more lesions, based on the occurrence frequency of lesion for each of the divided regions which is obtained from clinical statistical information; and

generating the clinical information, based on the changed confidence score and based on the degree of risk of lesion for each of the divided regions which is obtained from the clinical statistical information.

2. The method of claim 1 , wherein the detecting the one or more lesions from the received body medical image using the lesion detection model comprising:

determining type information of the one or more lesions included in the received body medical image;

determining a confidence score corresponding to the determined type information of the one or more lesions; and

generating contour information of the detected one or more lesions.

3. The method of claim 1 , wherein the generating the diagnosis result for the received body medical image, based on the anatomical location information for the one or more lesions further comprising:

generating a readout using a readout generation model, based on the generated clinical information.

4. The method of claim 3 , wherein the generating the diagnosis result for the received body medical image, based on the anatomical location information for the one or more lesions further comprising:

modifying the generated readout by reflecting a modification input of a user; and

generating the diagnosis result for the received body medical image based on the modified readout.

5. The method of claim 1 , wherein the method further comprising:

generating a user interface including information on the diagnosis result, and

wherein the user interface comprising:

a first area for displaying the detected one or more lesions on the received body medical image;

a second area for displaying summary information about the detected one or more lesions; and

a third area for displaying a readout corresponding to the diagnosis result.

6. The method of claim 5 , wherein the user interface further comprising:

a notification area for providing additional information according to a degree of risk of lesion, and

wherein whether to display the notification area is determined based on at least one of the anatomical location information for the one or more lesions or a confidence score for the one or more lesions.

7. The method of claim 1 , wherein the extracted body region is divided into a plurality of sub regions, and

wherein the generating the anatomical location information for the one or more lesions comprising:

determining a location in which the one or more lesions exist in the divided sub regions by matching the detection result for the one or more lesions with the plurality of sub regions for the extracted body region; and

generating the anatomical location information indicating that the one or more lesions exist in the determined location.

8. A computer program comprising instructions stored in a non-transitory computer-readable storage medium to cause a computer to perform the following operations, the operations comprising:

receiving a body medical image;

detecting one or more lesions from the received body medical image using a lesion detection model;

extracting a body region having anatomical significance from the received body medical image using an anatomical analysis model;

generating anatomical location information for the one or more lesions by matching the detection result for the one or more lesions with the extracted body region; and

generating a diagnosis result for the received body medical image, based on the anatomical location information for the one or more lesions,

wherein the body region having the anatomical significance is a body organ that is likely to be a point of occurrence of the one or more lesions or a body organ that is a target of the detection of the one or more lesions, and

wherein the generating the diagnosis result for the received body medical image, based on the anatomical location information for the one or more lesions comprising:

generating clinical information indicating clinical significance in a situation in which the one or more lesions are related to the anatomical location information,

wherein the clinical information comprising:

an occurrence frequency of lesion for each of regions divided from the body region having the anatomical significance; and

a degree of risk of lesion for each of the divided regions,

wherein the generating the clinical information indicating the clinical significance in the situation in which the one or more lesions are related to the anatomical location information comprising:

changing a confidence score determined in the step of detecting the one or more lesions, based on the occurrence frequency of lesion for each of the divided regions which is obtained from clinical statistical information; and

generating the clinical information, based on the changed confidence score and based on the degree of risk of lesion for each of the divided regions which is obtained from the clinical statistical information.

9. A server, comprising:

a processor including one or more cores;

a network unit for receiving a body medical image; and

a memory,

wherein the processor is configured to:

detect one or more lesions from the received body medical image using a lesion detection model;

extract a body region having anatomical significance from the received body medical image using an anatomical analysis model;

generate anatomical location information for the one or more lesions by matching the detection result for the one or more lesions with the extracted body region; and

generate a diagnosis result for the received body medical image, based on the anatomical location information for the one or more lesions,

wherein the body region having the anatomical significance is a body organ that is likely to be a point of occurrence of the one or more lesions or a body organ that is a target of the detection of the one or more lesions, and

wherein the generating the diagnosis result for the received body medical image, based on the anatomical location information for the one or more lesions comprising:

generating clinical information indicating clinical significance in a situation in which the one or more lesions are related to the anatomical location information,

wherein the clinical information comprising:

an occurrence frequency of lesion for each of regions divided from the body region having the anatomical significance; and

a degree of risk of lesion for each of the divided regions,

wherein the generating the clinical information indicating the clinical significance in the situation in which the one or more lesions are related to the anatomical location information comprising:

changing a confidence score determined in the step of detecting the one or more lesions, based on the occurrence frequency of lesion for each of the divided regions which is obtained from clinical statistical information; and

generating the clinical information, based on the changed confidence score and based on the degree of risk of lesion for each of the divided regions which is obtained from the clinical statistical information.

10. The server of claim 9 , wherein the detecting the one or more lesions from the received body medical image using the lesion detection model comprising:

determining type information of the one or more lesions included in the received body medical image;

determining a confidence score corresponding to the determined type information of the one or more lesions; and

generating contour information of the detected one or more lesions.

11. The method of claim 1 , wherein the generating the diagnosis result for the received body medical image, based on the anatomical location information for the one or more lesions further comprising: generating a readout using a readout generation model, based on the generated clinical information.

12. The method of claim 11 , wherein the generating the diagnosis result for the received body medical image, based on the anatomical location information for the one or more lesions further comprising:

modifying the generated readout by reflecting a modification input of a user; and

generating the diagnosis result for the received body medical image based on the modified readout.

13. The server of claim 9 , wherein the processor is further configured to:

generate a user interface including information on the diagnosis result, and

wherein the user interface comprising:

a first area for displaying the detected one or more lesions on the received body medical image;

a second area for displaying summary information about the detected one or more lesions; and

a third area for displaying a readout corresponding to the diagnosis result.

14. The server of claim 13 , wherein the user interface further comprising:

a notification area for providing additional information according to a degree of risk of lesion, and

wherein whether to display the notification area is determined based on at least one of the anatomical location information for the one or more lesions or a confidence score for the one or more lesions.

15. The server of claim 9 , wherein the extracted body region is divided into a plurality of sub regions, and

wherein the generating the anatomical location information for the one or more lesions comprising:

determining a location in which the one or more lesions exist in the divided sub regions by matching the detection result for the one or more lesions with the plurality of sub regions for the extracted body region; and

generating the anatomical location information indicating that the one or more lesions exist in the determined location.

Assignments (1)

Priority Claims (1)

Continuity (1)

References Cited (28)

Cited By (1)

US 12,725,706