Anatomical structure identification apparatus and display method thereof

Inventors

PAUL, Anjan KumarLee, Hojun

Assignees

Samsung Medison Co Ltd

Interested in licensing this patent?

MTEC can help explore whether this patent might be available for licensing for your application.

Publication Number

US-12239481-B2

Patent

Publication Date

2025-03-04

Expiration Date


Abstract

An ultrasonic imaging apparatus and an ultrasonic image display method are disclosed. An ultrasonic imaging apparatus according to one disclosed embodiment can comprise: a display unit; a user interface; a memory for storing one or more instructions; and a processor for executing the one or more instructions so as to determine a reference point in an ultrasonic image, identify at least one anatomical structure of interest on the basis of a location relationship with the reference point with respect to the determined reference point, and control the display unit so that the identified anatomical structure of interest and information about the anatomical structure of interest are displayed.

Core Innovation

The invention relates to an ultrasonic 2-dimensional image display method in which an ultrasonic 2-dimensional image including a proximal carpal tunnel is acquired, the ultrasonic image showing a cross section of the proximal carpal tunnel. A reference point is determined in the ultrasonic 2-dimensional image by selecting one of a plurality of landmark anatomical structures included in the proximal carpal tunnel as the reference point, and at least one anatomical structure of interest is identified from among the plurality of landmark anatomical structures except the reference point based on a positional relationship with the reference point.

The positional relationship includes a 2-dimensional distance value in the ultrasonic 2-dimensional image, a direction in the ultrasonic 2-dimensional image, and a natural language description indicating the 2-dimensional distance value and the direction between the anatomical structure of interest and the reference point. The method displays the identified anatomical structure of interest of the proximal carpal tunnel and information about the anatomical structure of interest on the ultrasonic image or parallel with the ultrasonic 2-dimensional image, and the displayed information indicates the positional relationship between the anatomical structure of interest and the reference point.

The invention further includes identifying the anatomical structure of interest by applying reference-point and anatomical-structure data to a learning model trained on position context information. When an anatomical structure of interest is unidentifiable, an estimated position of the anatomical structure of interest on the ultrasonic image is acquired using the positional relationship to the reference point, and the estimated position can be accompanied by a probability displayed for the ultrasonic image.

Claims Coverage

The provided independent claims cover three aspects: an ultrasonic 2D image display method, an ultrasonic imaging apparatus implementing that method, and a non-transitory computer-readable medium storing program code to perform the method. Across these independent claims, the key inventive features revolve around reference-point selection from landmark anatomical structures, identification of anatomical structures of interest from 2D positional relationships expressed with distance, direction, and natural language, and display of the identified structure and positional-information on or parallel with the ultrasonic image.

Reference point selection from proximal carpal tunnel landmarks

Determining a reference point in the ultrasonic 2D image, wherein one of a plurality of landmark anatomical structures included in the proximal carpal tunnel is selected as the reference point.

Identification of anatomical structure of interest from 2D positional relationship

Identifying at least one anatomical structure of interest based on a positional relationship with the reference point on the basis of the determined reference point, wherein the at least one anatomical structures of interest is identified from among the plurality of landmark anatomical structures except the reference point, and the positional relationship includes a 2-dimensional distance value in the ultrasonic 2D image, a direction in the ultrasonic 2D image, and a natural language description indicating the 2-dimensional distance value and the direction between the anatomical structure of interest and the reference point.

Display of identified structure and positional-information relative to the reference point

Displaying the identified anatomical structure of interest of the proximal carpal tunnel and information about the anatomical structure of interest on the ultrasonic image or parallel with the ultrasonic 2D image, wherein the information about the anatomical structure of interest is information indicating the positional relationship between the anatomical structure of interest and the reference point.

Learning model-based identification using position context information

Identifying at least one anatomical structure of interest by applying reference-point and anatomical-structure data to a learning model trained using position context information.

Estimated position acquisition when anatomical structure is unidentifiable

Determining whether an anatomical structure of interest is identifiable, and if unidentifiable, acquiring an estimated position of the anatomical structure of interest on the ultrasonic image using positional relationships based on a reference point.

Probability display for an estimated position

Calculating a probability of an estimated position and displaying the calculated probability.

Overall, the independent claims consistently require selecting a reference point from landmark anatomical structures within the proximal carpal tunnel, identifying other anatomical structures of interest using a 2D positional relationship (distance, direction, and a natural-language description), and displaying the identified structure together with positional-information relative to the reference point. Dependent refinements add learning-model-based identification with position context information, fallback to estimated positions when structures are unidentifiable, and displaying probability for an estimated position.

Stated Advantages

Not explicitly described in patent.

Documented Applications

Not explicitly described in patent.

JOIN OUR MAILING LIST

Stay Connected with MTEC

Keep up with active and upcoming solicitations, MTEC news and other valuable information.