KR20030022567A - System for automatically optimizing input ultrasound image - Google Patents

Abstract

PURPOSE: A system for optimizing an ultrasound image is provided to reduce the process required to an ultrasound image diagnosis by optimizing automatically all the parameters affecting to an input ultrasound image. CONSTITUTION: A memory(10) stores standard image data and measured parameters having a plurality of first feature parameters and an image parameters about each ultrasound image. A feature extracting unit(20) extracts the feature of an input ultrasound image to output it as a second feature parameter. A similarity calculating unit(30) compares the second feature parameter with the first feature parameters and calculates the similarity among them to extract the first feature parameter having the highest similarity. A determining unit determines an internal organ name and an internal scan direction corresponding to the first feature parameter having the highest similarity. An image control unit(40a) controls the ultrasound image by using the image parameter corresponding to the internal organ name and the internal scan direction.

Description

Ultrasonic Image Optimization System {SYSTEM FOR AUTOMATICALLY OPTIMIZING INPUT ULTRASOUND IMAGE}

The present invention relates to a system for analyzing an ultrasound image input to an ultrasound imaging system and automatically optimizing all parameter elements affecting an input ultrasound image using the analysis result.

A conventional method for optimizing ultrasound images, a method in which a user arbitrarily selects some preset numerical values for image parameter elements such as brightness and contrast, and the input ultrasound image is related to a diagnosis site. There is a method of forcibly adjusting to a plurality of specific image parameter elements defined in advance. However, this method does not fully consider the characteristics of various organs in the human body. From the user's point of view, it is difficult to optimize the ultrasound image, and a standard value is manually selected or an ultrasound image is diagnosed to diagnose each organ in the human body. There is a hassle to force adjustment.

For example, if you want to measure a specific part of your body, such as BPD (Bi-Parietal Diameter), you need to designate the location you want to measure on the actual screen using several steps of button operation and pointer. However, there is a problem that the measurement work is cumbersome because the BPD item must be selected from a plurality of items.

Accordingly, the main object of the present invention is to provide a system for analyzing the input ultrasound image, and automatically optimizing all parameters affecting the input ultrasound image based on the analysis result, thereby simplifying the process necessary for the diagnosis of the ultrasound image. Is in.

1 is a schematic block diagram of an ultrasound image optimization system according to a preferred embodiment of the present invention.

Fig. 2 shows each feature parameter stored in the database DB of the present invention.

3 is a diagram illustrating each image parameter stored in a database of the present invention.

4 shows each measurement parameter stored in a database of the present invention.

5A to 8B are exemplary experimental photographs for explaining the difference between the image obtained by the conventional method and the image obtained by the present invention, respectively.

<Explanation of symbols for the main parts of the drawings>

10; Standard video DB

10a; Feature Parameter DB

10b; Image Parameter DB

11; Measurement Parameter DB

20a; Preprocessor

20b; Video converter

20c; Feature detector

30; Similarity Calculator

40; Video applications

40a; Video controller

40b; Automatic measuring unit

In order to achieve the above object, according to a preferred embodiment of the present invention, in the system for optimizing the input ultrasound image, the first feature parameter and the image parameter for each ultrasound image classified for each organ or organ scan direction in the human body A memory storing standard image data and measurement parameters, a feature extracting means for extracting a feature of an input ultrasound image and outputting the feature as a second feature parameter, the second feature parameter and the plurality of first feature parameters, Similarity calculation means for calculating a similarity between them to extract a first feature parameter having the highest similarity, determining means for determining a long term name or long term scan direction corresponding to the first feature parameter having the highest similarity; The image wave corresponding to the determined organ name or organ scan direction The ultrasound image optimization system including a video adjustment means for adjusting the ultrasound image is provided by the meter.

In addition, the ultrasound image optimization system according to the present invention, by using the automatic measurement function stored in the database for the input ultrasound image BP (Bi-Parietal Diameter), AC (Abdominal Circumference), FL (Femur Length), M The apparatus may further include an automatic measuring unit measuring at least one of a Mode (Motion Mode) and a Doppler item and outputting the measurement result to a display.

Hereinafter, with reference to the accompanying drawings, preferred embodiments of the present invention will be described in detail.

1 is a schematic block diagram of an ultrasound image optimization system according to a preferred embodiment of the present invention.

As shown in FIG. 1, the ultrasound image optimization system 100 of the present invention includes a standard image DB 10, a measurement parameter DB 11, a feature extractor 20, a similarity calculator 30, and an image application ( 40).

The standard image DB 10 stores standard image data for each organ in the human body, and includes a feature parameter DB 10a for storing standard feature parameters for each image and an image parameter for storing standard image parameters for each organ. DB 10b is provided. Image parameters stored in the image parameter DB 10b include brightness values, contrast levels, edge intensities, transmission / reception frequencies, average speeds of ultrasonic waves, and the like. The measurement parameter DB 11 stores an automatic measurement function necessary for measuring each image. The automatic measurement functions stored in the measurement parameter DB (11) are BPD (Bi-Parietal Diameter), AC (Abdominal Circumference), FL (Femur Length) and M Mode ( It is a function that automatically measures items such as Motion Mode and Doppler, the mode that shows the flow of blood flow based on the Doppler effect. Can be used.

The feature extractor 20 includes a preprocessor 20a, an image converter 20b, and a feature detector 20c. The preprocessor 20a performs normalization and noise removal on the input ultrasound image obtained from an ultrasound probe (not shown) to select meaningful regions within the input ultrasound image.

The image converter 20b deforms the shape or distribution of the image so as to easily extract the features of the image. That is, the image transform unit 20b uses an ordinary fast Fourier transform (FFT), a discrete cosine transform (DCT), a wavelet transform, a Gabor transform, and the like. The system converts each frequency component in the image more systematically by converting it into an area, and performs the function of emphasizing the characteristics of the image in the spatial domain through image segmentation or edge enhancement. In this way, the image classified and highlighted by the image converter 20b is input to the feature detector 20c.

The feature detector 20c detects a feature representative of an image and performs a function of numerically representing it. That is, the feature detector 20c defines an image of the image using energy, moment, local brightness distribution, and directionality of each frequency component in the frequency domain, and brightness distribution of the image in the spatial domain. The characteristics of the image are defined by combining edge position and intensity information and region segmentation information. The feature of the image defined as described above is supplied to the similarity calculator 30 as a first feature parameter for long term recognition.

The similarity calculator 30 may include a first feature parameter supplied from the feature detector 20c and a plurality of second feature parameters input from the feature parameter DB 10a of the standard image DB 10 as shown in FIG. 2. By comparing the two, and calculating the similarity between them and extracting the second feature parameter having the highest similarity among the calculated similarities, thereby outputting the classification name to which the extracted second feature parameter belongs as the recognition organ name. As shown in FIG. 2, since each feature parameter stored in the feature parameter DB 10a is classified for each organ, a second feature parameter having a feature most similar to the first feature parameter is recognized as a long-term name of the input ultrasound image. do. In addition, although not shown in FIG. 2, each feature parameter is classified into a long-term scan direction, so that the feature parameter having the highest similarity may be regarded as the long-term scan direction of the input ultrasound image through the similarity calculation of the feature parameter as described above. Can be.

The long term name data or long term scan direction data output from the similarity calculator 30 is input to the image application 40 having the image adjusting unit 40a and the automatic measuring unit 40b. As illustrated in FIG. 3, the image adjusting unit 40a of the image application 40 stores image parameters for each organ such as gain, contrast, edge, etc. stored in the image parameter DB 10b. The input ultrasound image is adjusted based on the output, and the adjusted image is output to the display. The automatic measuring unit 40b of the image application 40 is stored in the measurement parameter DB 11 as shown in FIG. 4, for example, Ob (Obstetric), M mode (M Mode), Doppler (Doppler). The measured image data is output on the display so that the user can perform automatic measurement on each item necessary for measurement or directly perform the function corresponding to each item by using the measurement parameter of each type.

5A to 8B are exemplary experimental photographs for explaining the difference between the ultrasound image obtained by the conventional method and the ultrasound image obtained by the present invention.

5A is an ultrasound image photograph of a liver obtained by a conventional ultrasound diagnosis system, and FIG. 5B is an ultrasound image photograph of a liver whose gain is adjusted according to the present invention. As can be seen from Figure 5b, compared with Figure 5a it was possible to obtain an ultrasound image of the liver adjusted to the gain of the appropriate brightness.

6A is an ultrasound image photograph of a liver obtained by a conventional ultrasound diagnosis system, and FIG. 6B is an ultrasound image photograph of a liver of which a dynamic range of an ultrasound input signal is adjusted according to the present invention. As can be seen from FIG. 6B, the ultrasound image of the liver is wider than the cloudy FIG. 6A, and the ultrasound image of the liver can be obtained by adjusting to an appropriate dynamic range compared to FIG. 6A.

7A is an ultrasound image photograph of the liver obtained by a conventional ultrasound diagnosis system, and FIG. 7B is an ultrasound image photograph of the liver with edge enhancement adjusted according to the present invention. As can be seen from FIG. 7B, compared to FIG. 7A, in which the difference between tissues is small and the discrimination power decreases, the ultrasound image of the liver with increased discrimination power may be obtained by emphasizing the difference between boundaries.

8A is an ultrasound image photograph of a liver obtained by a conventional ultrasound diagnosis system, and FIG. 8B is an ultrasound image photograph of a liver obtained through image optimization according to the present invention. As can be seen from FIG. 8B, by automatically adjusting each image parameter such as gain, dynamic range, and edge enhancement with respect to the input ultrasound image, an ultrasound image having a higher quality than that of FIG. 8A can be obtained.

The ultrasound image optimization system according to the present invention may be operated independently in the ultrasound diagnosis system, or may be operated by sharing a database used by each system by being connected to another ultrasound diagnosis system through a network.

As mentioned above, although preferred embodiment of this invention was described, those skilled in the art can change variously, without deviating from the claim of this invention.

Therefore, according to the present invention, a feature parameter of an ultrasound image of a human organ input from a medical ultrasound imaging system is extracted, and the extracted feature parameter is compared with a feature parameter for each organ in the human body already stored in a database. By extracting feature parameters having similar features, recognizing organ names corresponding to the extracted feature parameters as organ names of the input ultrasound image, and optimizing the images of the organs using image parameters stored in a database, The ultrasound image may be provided to the user.

In addition, according to the present invention, by applying a pre-stored automatic measurement algorithm to the ultrasound image of good quality obtained as described above, the manual manual operation, such as button operation, positioning operation, etc. during the measurement of the ultrasound image is simplified. You can.

Claims (6)

In the system for optimizing the input ultrasound image, A memory for storing standard image data and measurement parameters having a first feature parameter and an image parameter for each ultrasound image classified by an organ or an organ scan direction in a human body; Feature extraction means for extracting a feature of the input ultrasound image and outputting it as a second feature parameter; Similarity calculating means for comparing the second feature parameter with the plurality of first feature parameters, calculating similarity between them, and extracting a first feature parameter having the highest similarity; Determining means for determining a long term name or a long term scan direction corresponding to the first feature parameter having the highest similarity; Image adjusting means for adjusting the ultrasound image using the image parameter corresponding to the determined organ name or organ scan direction Ultrasound image optimization system comprising a. The method of claim 1, The memory automatically measures at least one of Bi-Parietal Diameter (BPD), Abdominal Circumference (AC), Femur Length (FL), M Mode (Motion Mode), and Doppler items for the input ultrasound image. A measurement function, the image parameter having at least one of a brightness value, contrast, edge intensity, transmit / receive frequency, and average speed of the ultrasound. The method of claim 1, The feature extraction means Normalization means for normalizing at least one of brightness, contrast, and edge intensity with respect to the input ultrasound image; Image conversion means for classifying each frequency component in the input ultrasound image by converting the normalized ultrasound image from a spatial domain to a frequency domain; In the frequency domain, the characteristics of the input ultrasound image are defined using at least one of energy, moment, local brightness distribution and directionality of each frequency component; In the case of the spatial region, a feature of the ultrasound image is defined by combining at least two of brightness distribution, edge position and intensity information of the input ultrasound image, and region segmentation information, and the defined feature is defined as the second feature parameter. Ultrasonic image optimization system comprising a feature defining means for outputting as. The method of claim 3, wherein And the image converting means emphasizes a feature of the input ultrasound image in the spatial region through image segmentation and edge enhancement. The method of claim 4, wherein And the image conversion means uses at least one of a fast Fourier transform (FFT), a discrete cosine transform (DCT), a wavelet transform, and a Gabor transform. The method of claim 1, Among the items of Bi-Parietal Diameter (BPD), Abdominal Circumference (AC), Femur Length (FL), Motion Mode (M Mode), and Doppler items, the automatic measurement function stored in the memory is applied to the input ultrasound image. Means for measuring at least one and outputting the measurement result to a display device.