Image segmentation method and electronic device therefor

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

acquire a medical image;

receive a selection of a partial area of the medical image via an input interface;

transform the medical image into a transformed medical image in which the selected partial area is positioned in a center of the transformed medical image;

extract at least one feature from the transformed medical image, using a deep learning technique;

enhance at least one feature of the at least one extracted feature;

restore, as a restored medical image, at least one feature of the at least one enhanced feature; and

inversely transform the restored medical image to provide segmented medical images.

2. The non-transitory computer-readable storage medium of claim 1 , wherein the instructions further cause the computing device to:

enhance the transformed medical image; and

perform a medical image reverse enhancement of the restored medical image.

3. The non-transitory computer-readable storage medium of claim 1 , wherein the instructions further cause the computing device to downsize the medical image to prevent the medical image from being moved out of a display area when the medical image is transformed.

4. The non-transitory computer-readable storage medium of claim 1 , wherein the instructions further cause the computing device to extract the at least one feature by:

generating a feature map by applying a convolution function to the transformed medical image,

reducing a size of the feature map by applying a pooling function to the feature map, and

applying an activation function.

5. The non-transitory computer-readable storage medium of claim 1 , wherein the instructions further cause the computing device to enhance the at least one feature of the at least one extracted feature by adding a first classification result, in which the at least one extracted feature is classified, and a second classification result, in which the at least one extracted feature is classified by reducing a dimension of the at least one extracted feature.

6. The non-transitory computer-readable storage medium of claim 1 , wherein the instructions further cause the computing device to:

create a combined feature by combining the at least one extracted feature and at least one feature enhanced by spreading the at least one extracted feature; and

determine a class of the combined feature.

7. The non-transitory computer-readable storage medium of claim 2 , wherein the instructions further cause the computing device to enhance the transformed medical image by enlarging the center of the transformed medical image.

8. The non-transitory computer-readable storage medium of claim 1 , wherein the instructions further cause the computing device to:

extract a boundary based on the segmented medical images; and

display the extracted boundary in the medical image.

9. The non-transitory computer-readable storage medium of claim 1 , wherein the instructions further cause the computing device to restore, as the restored medical image, the at least one enhanced feature and the at least one extracted feature.

10. A medical image segmentation method comprising:

acquiring a medical image;

displaying the acquired medical image;

receiving a selection of a partial area of the acquired medical image via an input interface;

transforming the acquired medical image into a transformed medical image in which the selected partial area is positioned in a center of the transformed medical image;

extracting at least one feature from the transformed medical image, using a deep learning technique;

enhancing at least one feature of the at least one extracted feature;

restoring, as a restored medical image, at least one feature of the at least one enhanced feature; and

inversely transforming the restored medical image to provide segmented medical images.

11. The method of claim 10 , wherein the transforming of the acquired medical image comprises one of:

reducing a size of the acquired medical image by trimming a part from a display area of the acquired medical image, the part being generated when the selected partial area is moved to the center of the transformed medical image; and

downsizing the acquired medical image to prevent the acquired medical image from being moved out of the display area.

12. The method of claim 10 , wherein the extracting of the at least one feature comprises:

generating a feature map by applying a convolution function to the transformed medical image,

reducing a size of the feature map by applying a pooling function to the feature map, and

applying an activation function.

13. The method of claim 10 , wherein the enhancing of the at least one feature comprises:

a first classification operation comprising determining a class of the at least one extracted feature;

a second classification operation comprising reducing a dimension of the at least one extracted feature and determining a class of the dimension-reduced feature; and

adding a result of the first classification operation and a result of the second classification operation.

14. The method of claim 10 , wherein the enhancing of the at least one feature comprises:

creating a combined feature by combining the at least one extracted feature and at least one feature enhanced by spreading the at least one extracted feature; and

determining a class of the combined feature.

15. The method of claim 10 , further comprising:

enhancing the transformed medical image; and

performing a medical image reverse enhancement of the restored medical image.

16. The method of claim 15 , wherein the enhancing of the transformed medical image comprises enlarging the center of the transformed medical image.

17. The method of claim 10 , further comprising:

extracting a boundary based on the segmented medical images; and

displaying the extracted boundary in the medical image.

18. The method of claim 10 , wherein the restoring comprises restoring, as the restored medical image, the at least one enhanced feature and the at least one extracted feature.

19. An electronic device comprising:

a display configured to display a medical image;

at least one processor operatively connected to the display; and

a memory operatively connected to the at least one processor,

wherein the memory is configured to store instructions which, when executed, cause the processor to:

receive a selection of a partial area of the medical image via an input interface;

extract at least one feature from the displayed medical image, using a deep learning technique;

enhance at least one feature corresponding to the partial area of the at least one extracted feature;

restore, as a restored medical image, at least one feature of the at least one enhanced feature to provide segmented medical images.

20. The non-transitory computer-readable storage medium of claim 1 , wherein the instructions further cause the computing device to:

provide one or more boundary candidates of the partial area; and

identify a boundary of the selected partial area based on a selected boundary candidates among the one or more boundary candidates.

21. The method of claim 11 , further comprising:

providing one or more boundary candidates of the partial area;

identifying a boundary of the selected partial area based on a selected boundary candidates among the one or more boundary candidates.

22. The electronic device of claim 19 , wherein the instructions further cause the processor to:

provide one or more boundary candidates of the partial area; and

identify a boundary of the selected partial area based on a selected boundary candidates among the one or more boundary candidates.