Three-dimensional medical image anonymization method and apparatus

1. A three-dimensional medical image anonymization method comprising:

determining a skin region of a three-dimensional medical image;

generating a human mask based on a human tissue region segmented from the three-dimensional medical image through segmentation algorithm, the human tissue region including bones and organs inside a human body, excluding the skin region, wherein the human mask represents a human tissue region to be protected from anonymization;

generating a skin expansion region in which the skin region of the three-dimensional medical image is expanded;

generating an anonymization region obtained by removing a region corresponding to the human mask from the skin expansion region; and

changing brightness values of voxels corresponding to the anonymization region in the three-dimensional medical image to a predefined value or an arbitrary value,

wherein the determining of the skin region comprises determining, as the skin region, a region of a predefined thickness inwards into the human body with respect to a boundary portion of the human body in the three-dimensional medical image,

wherein the generating of the skin expansion region comprises expanding the skin region in both an outward direction and an inward direction of the human body with respect to voxels of the skin region, and

the skin expansion region comprises all of the skin region, a region expanded outwards from the human body, and a region expanded inwards into the human body.

2. The three-dimensional medical image anonymization method of claim 1 , wherein the three-dimensional medical image comprises a computed tomography (CT) image.

3. The three-dimensional medical image anonymization method of claim 1 , wherein the determining of the skin region comprises determining, as the skin region, a region of a predefined thickness inwards into the human body with respect to a boundary portion of the human body in the three-dimensional medical image or determining the skin region by using an artificial intelligence model for segmenting the skin region from the three-dimensional medical image.

4. The three-dimensional medical image anonymization method of claim 1 , wherein the generating of the human mask comprises:

segmenting human tissue regions including bones and organs from the three-dimensional medical image by using a human tissue segmentation algorithm; and

generating the human mask by filling empty regions between the segmented human tissue regions.

5. The three-dimensional medical image anonymization method of claim 1 , wherein the generating of the skin expansion region comprises expanding a region in up, down, left, and right directions with respect to each voxel of the skin region.

6. A three-dimensional medical image anonymization apparatus comprising:

a skin region extractor configured to determine a skin region segmented from a three-dimensional medical image through segmentation algorithm;

a mask generator configured to generate a human mask based on a human tissue region of the three-dimensional medical image, the human tissue region including bones and organs inside a human body, excluding the skin region, wherein the human mask represents a human tissue region to be protected from anonymization;

an expansion unit configured to generate a skin expansion region with respect to voxels constituting the skin region of the three-dimensional medical image;

an anonymization region generator configured to generate an anonymization region obtained by removing a region corresponding to the human mask from the skin expansion region; and

an information changing unit configured to change brightness values of voxels corresponding to the anonymization region in the three-dimensional medical image to a predefined value or an arbitrary value,

wherein the determining of the skin region comprises determining, as the skin region, a region of a predefined thickness inwards into the human body with respect to a boundary portion of the human body in the three-dimensional medical image,

wherein the expansion unit is configured to expand the skin region in both an outward direction and an inward direction of the human body with respect to voxels of the skin region, and

the skin expansion region comprises all of the skin region, a region expanded outwards from the human body, and a region expanded inwards into the human body.

7. A non-transitory computer-readable recording medium having recorded thereon a computer program for executing the method of claim 1 .