Granted Patent B2

US 12,456,193 · App. 18/180,965 · Granted Oct 28, 2025

Inventors: Kohtaroh Kusu (Ebina, JP); Yuki Sakaguchi (Fujisawa, JP)

Assignee: TERUMO KABUSHIKI KAISHA

G06T7/0012G06T7/12G06V10/25G06T2207/30021G06V2201/07

Abstract

Claims (61)

1. A non-transitory computer-readable medium storing a computer program executed by a computer processor to execute a process comprising:

acquiring a medical image generated based on a signal detected by a catheter inserted into a luminal organ;

detecting an object region included in the medical image by inputting the acquired medical image into a model trained for detecting the object region included in the medical image;

detecting the object region in units of pixels by semantic segmentation in the model;

acquiring an output of an activation function included in the model in each pixel included in a predetermined region of the medical image using the model;

calculating reliability of each of a plurality of boundary points located on a boundary between the detected object region and another image area in the medical image;

calculating the calculated reliability of a boundary point included in the predetermined region based on a ratio of pixels in which the output of the activation function to all the pixels included in the predetermined region of the medical image is within a predetermined range;

correcting the boundary based on a detection result of the object region by the model by performing interpolation using the boundary points at which the calculated reliability satisfies a predetermined standard; and

generating screen information that displays the detected object region.

2. The computer-readable medium according to claim 1 , further comprising:

generating screen information that displays a boundary based on a detection result of the object region by the model and a new boundary generated by correcting the boundary based on the detection result of the object region by the model in different display modes; and

outputting the generated screen information that displays the boundary based on the detection result of the object region by the model and the new boundary generated by correcting the boundary based on the detection result of the object region by the model in different display modes.

3. The computer-readable medium according to claim 1 , further comprising:

generating a new boundary for correcting the boundary based on a detection result of the object region by performing spline interpolation using the boundary points at which the calculated reliability satisfies a predetermined standard.

4. The computer-readable medium according to claim 1 , further comprising:

acquiring an output of an activation function included in the model; and

calculating the calculated reliability based on the acquired output of the activation function.

5. The computer-readable medium according to claim 1 , further comprising:

calculating the calculated reliability based on presence or absence of an artifact in the medical image.

6. An information processing device comprising:

a control unit comprising a processor configured to:

acquire a medical image generated based on a signal detected by a catheter inserted into a luminal organ;

detect an object region included in the medical image by inputting the acquired medical image into a model trained for detecting the object region included in the medical image;

detect the object region in units of pixels by semantic segmentation in the model;

acquire an output of an activation function included in the model in each pixel included in a predetermined region of the medical image using the model;

calculate reliability of each of a plurality of boundary points located on a boundary between the detected object region and another image area in the medical image;

calculate the calculated reliability of a boundary point included in the predetermined region based on a ratio of pixels in which the output of the activation function to all the pixels included in the predetermined region of the medical image is within a predetermined range;

correct the boundary based on a detection result of the object region by the model by performing interpolation using the boundary points at which the calculated reliability satisfies a predetermined standard; and

generate screen information that displays the detected object region.

7. The information processing device according to claim 6 , wherein the control unit is further configured to:

generate screen information that displays a boundary based on a detection result of the object region by the model and a new boundary generated by correcting the boundary based on the detection result of the object region by the model in different display modes; and

output the generated screen information that displays the boundary based on the detection result of the object region by the model and the new boundary generated by correcting the boundary based on the detection result of the object region by the model in different display modes on a display.

8. The information processing device according to claim 6 , wherein the control unit is further configured to:

generate a new boundary for correcting the boundary based on a detection result of the object region by performing spline interpolation using the boundary points at which the calculated reliability satisfies a predetermined standard.

9. The information processing device according to claim 6 , wherein the control unit is further configured to:

acquire an output of an activation function included in the model; and

calculate the calculated reliability based on the acquired output of the activation function.

10. The information processing device according to claim 6 , wherein the control unit is further configured to:

calculate the calculated reliability based on presence or absence of an artifact in the medical image.

11. An information processing method comprising:

acquiring a medical image generated based on a signal detected by a catheter inserted into a luminal organ;

detecting an object region included in the medical image by inputting the acquired medical image into a model trained for detecting the object region included in the medical image;

detecting the object region in units of pixels by semantic segmentation in the model;

acquiring an output of an activation function included in the model in each pixel included in a predetermined region of the medical image using the model;

calculating reliability of each of a plurality of boundary points located on a boundary between the detected object region and another image area in the medical image;

calculating the calculated reliability of a boundary point included in the predetermined region based on a ratio of pixels in which the output of the activation function to all the pixels included in the predetermined region of the medical image is within a predetermined range;

correcting the boundary based on a detection result of the object region by the model by performing interpolation using the boundary points at which the calculated reliability satisfies a predetermined standard; and

generating screen information that displays the detected object region.

12. The information processing method according to claim 11 , further comprising:

generating screen information that displays a boundary based on a detection result of the object region by the model and a new boundary generated by correcting the boundary based on the detection result of the object region by the model in different display modes; and

outputting the generated screen information that displays the boundary based on the detection result of the object region by the model and the new boundary generated by correcting the boundary based on the detection result of the object region by the model in different display modes.

13. The information processing method according to claim 11 , further comprising:

generating a new boundary for correcting the boundary based on a detection result of the object region by performing spline interpolation using the boundary points at which the calculated reliability satisfies a predetermined standard.

14. The information processing method according to claim 11 , further comprising:

acquiring an output of an activation function included in the model; and

calculating the calculated reliability based on the acquired output of the activation function.

15. The information processing method according to claim 11 , further comprising:

calculating the calculated reliability based on presence or absence of an artifact in the medical image.

16. The information processing method according to claim 11 , further comprising:

generating warning information when the object region is not detected in the acquired medical image because the ratio of pixels in which the output of the activation function to all the pixels included in the predetermined region of the medical image is equal to or greater than the predetermined range.

17. The information processing method according to claim 16 , wherein the warning information is one or more of a text or audio data.

Assignments (1)

Priority Claims (1)

Continuity (2)

References Cited (18)