Anomalousness determination method, anomalousness determination apparatus, and computer-readable recording medium

1. A non-transitory computer-readable recording medium storing therein an anomalousness determination program that causes a computer to execute a process comprising:

sensing a region in an object wherein a plurality of regions are included in a structure of the object and the object is supposed to be included in each of a plurality of ultrasound examination images using an object sensing technique, the sensed region being represented by a probability of a position of the region in the object, the object sensing technique being a learning model where the regions relating to the structure of the object are learned;

based on the probability and the structure of the object, acquiring a rate of detection of each of the plurality of regions in the object; and

determining anomalousness in the object based on the acquired rate of detection of each of the regions.

2. The non-transitory computer-readable recording medium according to claim 1 , wherein the ultrasound examination images are images selected as determined to include the object.

3. The non-transitory computer-readable recording medium according to claim 1 , wherein determining includes determining anomalousness in the object based on whether the object is sensed to exist in the ultrasound examination images.

4. The non-transitory computer-readable recording medium according to claim 1 , wherein the ultrasound examination images are images obtained by scanning the object.

5. The non-transitory computer-readable recording medium according to claim 4 , wherein the scanning is performed in a single direction.

6. The non-transitory computer-readable recording medium according to claim 4 , wherein the ultrasound examination images are images that are selected such that parts of return of the scanning are not contained.

7. A non-transitory computer-readable recording medium storing therein an anomalousness determination program that causes a computer to execute a process comprising:

detecting frames over a predetermined time period wherein the frames are scan cross-sections included in a video in which an object is supposed to be included and changes in pixel values of the frames over the predetermined time period are more than a predetermined threshold;

performing image recognition for detecting regions included in a structure of the object in each of the detected frames, the image recognition using a learning model where the regions relating to the structure of the object are learned;

acquiring a rate of detection of each of the regions in the object based on the image recognition and the structure of the object; and

detecting anomalousness in the object based on the acquired rate of detection of each of the regions.

8. An anomalousness determination method comprising:

sensing a region in an object wherein a plurality of regions are included in a structure of the object and the object is supposed to be included in each of a plurality of ultrasound examination images, using an object sensing technique, the sensed region being represented by a probability of a position of the region in the object, the object sensing technique being a learning model where the regions relating to the structure of the object are learned;

based on the sensing probability and the structure of the object, acquiring a rate of detection of each of the plurality of regions in the object; and

determining anomalousness in the object based on the acquired rate of detection of each of the regions, by a processor.

9. An anomalousness determination method comprising:

detecting frames over a predetermined time period wherein the frames are scan cross-sections included in a video in which an object is supposed to be included and changes in pixel values of the frames over the predetermined time period are more than a predetermined threshold;

performing image recognition for detecting regions included in a structure of the object in each of the detected frames, the image recognition using a learning model where the regions relating to the structure of the object are learned;

acquiring a rate of detection of each of the regions in the object based on the image recognition and the structure of the object; and

detecting anomalousness in the object based on the acquired rate of detection of each of the regions, by a processor.

10. An anomalousness determination apparatus comprising:

a processor configured to:

sense a region in an object wherein a plurality of regions are included in a structure of the object and the object is supposed to be included in each of a plurality of ultrasound examination images, using an object sensing technique, the sensed region being represented by a probability of a position of the region in the object, the object sensing technique being a learning model where the regions relating to the structure of the object are learned;

based on the probability and the structure of the object, acquire a rate of detection of each of the plurality of regions in the object; and

determine anomalousness in the object based on the acquired rate of detection of each of the regions.

11. An anomalousness determination apparatus comprising:

a processor configured to:

detect frames over a predetermined time of period wherein the frames are scan cross-sections included in a video-an in which object is supposed to be included and changes in pixel values of the frames over the predetermined time period are more than a predetermined threshold;

perform image recognition for detecting regions included in a structure of the object in each of the detected frames, the image recognition using a learning model where the regions relating to the structure of the object are learned;

acquire a rate of detection of each of the regions in the object based on the image recognition and the structure of the object; and

detect anomalousness in the object based on the acquired rate of detection of each of the regions.