Method for providing information on M-mode ultrasound image, and device for providing information on M-mode ultrasound image by using same

A method implemented by a processor, the method comprising:

• receiving an M-mode ultrasound image of a subject;
• segmenting a plurality of cardiac tomographic regions within the M-mode ultrasound image, respectively, by using a first model that is trained so as to segment the M-mode ultrasound image into the plurality of cardiac tomographic regions by inputting the M-mode ultrasound image;
• obtaining a plurality of post-processed cardiac tomographic regions by post-processing the plurality of segmented cardiac tomographic regions, the post-processing including at least one of
• removing segmented regions outside an M-mode region of the M-mode ultrasound image, or
• filling holes within the plurality of segmented cardiac tomographic regions;
• determining diastole and systole within the M-mode ultrasound image, by using a second model that is trained to predict the diastole and the systole of the M-mode ultrasound image by inputting the M-mode ultrasound image;
• and
• determining measurements for the plurality of segmented post-processed cardiac tomographic regions based on the diastole and the systole,
• wherein each of the first model and the second model includes a sampling stage and an upsampling stage, the first model and the second model sharing one sampling stage, and
• wherein the upsampling stage of the second model has a filter size smaller than a filter size of the upsampling stage of the first model.