JP5089442B2 - Ultrasonic image processing apparatus and method, and program - Google Patents

Description

  The present invention relates to an ultrasonic image processing apparatus, method, and program, and more particularly to a technique for extracting speckle noise from image data of an ultrasonic diagnostic apparatus using phase information included in the data.

  Conventionally, it has been widely performed to obtain a tomographic image of a subject using ultrasonic waves and used for medical diagnosis, but in image data of an ultrasonic diagnostic apparatus, called speckle noise (speckle), There is low echo information generated by random interference of reflected ultrasonic waves from a plurality of reflectors in the subject. Since this speckle noise hinders image diagnosis, it is required to reduce speckle noise from an ultrasonic diagnostic image. On the other hand, the signal (speckle pattern) containing speckle noise changes depending on the condition of the tissue diagnosed by the subject and is considered to reflect the tissue characteristics. Technologies that can be extracted and used for diagnosis are also being considered.

For example, there is known one that analyzes speckle information included in the entire ultrasonic image and displays the analysis result (see Patent Document 1). Specifically, the amplitude information of the ultrasonic reception signal is converted into image data, and structure image data free from speckle noise is generated from the maximum value and the minimum value for the obtained original image data. The speckle image data with only speckle noise is calculated from the difference obtained by subtracting the structure image data without speckle noise from the data.
JP 2006-212054 A

  However, in the above conventional speckle extraction technology, the shape and echo level of speckle noise vary, and it is very complex, so it is very difficult to accurately extract the speckle noise part from amplitude information. There is a problem that there is.

  The present invention has been made in view of such circumstances, and an ultrasonic image processing apparatus, method, and program capable of accurately extracting speckles without depending on the shape of speckles or the surrounding echo level. The purpose is to provide.

  In order to achieve the above-mentioned object, the invention according to claim 1 is characterized in that an ultrasonic transmission / reception unit that transmits an ultrasonic wave toward a subject and receives an ultrasonic wave reflected from the subject; There is provided an ultrasonic image processing apparatus comprising: speckle component extraction means for extracting a speckle component from a phase component of an ultrasonic signal received by a sound wave transmitting / receiving means.

  As a result, speckles are extracted using phase information instead of amplitude information, so speckles can be extracted independent of speckle shape and surrounding echo levels.

  According to a second aspect of the present invention, the ultrasonic image processing apparatus according to the first aspect further includes a display unit that displays an extraction result extracted by the speckle component extraction unit.

  By displaying the speckle extraction result, utilization for diagnosis is promoted.

  In addition, according to a third aspect of the present invention, in the ultrasonic image processing apparatus according to the first or second aspect, the speckle component extraction unit uses a change amount of the phase component to specify a spec from the phase component of the ultrasonic signal. It is characterized by extracting the components.

  By using the amount of change in the phase component, speckle can be extracted more accurately.

  According to a fourth aspect of the present invention, in the ultrasonic image processing apparatus according to the first or second aspect, the speckle component extraction unit uses a change amount of at least two directions of the phase component. And

  In this way, the extraction accuracy can be increased by using the change amounts in a plurality of directions.

  Further, as shown in claim 5, in the ultrasonic image processing device according to any one of claims 1 to 4, the speckle component extraction means uses a discriminant function for discriminating speckle. Features.

  Thereby, a discriminant function according to the purpose of speckle extraction can be used.

According to a sixth aspect of the present invention, in the ultrasonic image processing apparatus according to the third or fourth aspect , the speckle component extracting unit sets a threshold value for determining speckle in comparison with the change amount of the phase component. It is characterized by using.

  This facilitates speckle extraction when speckle can be determined by comparison with a threshold value.

  The ultrasonic image processing apparatus according to claim 5, wherein the discriminant function is a feature amount generated from a phase component whose positions of speckle and non-speckle are known. It is characterized by being set by.

  Thereby, a discriminant function can be created using the feature amount according to the speckle extraction process.

  According to an eighth aspect of the present invention, in the ultrasonic image processing apparatus according to the seventh aspect, the feature amount is a change amount of a phase component.

  According to this, since the change amount of the phase component is directly used as the feature amount for the speckle extraction, the extraction process is simplified.

  Similarly, in order to achieve the object, the invention according to claim 9 is an ultrasonic transmission / reception means for transmitting ultrasonic waves toward the subject and receiving ultrasonic waves reflected from the subject. And speckle component extraction means for extracting a speckle component from the combination of the amplitude component or envelope component of the ultrasonic signal received by the ultrasonic transmission / reception means and the phase component. An image processing apparatus is provided.

  According to this, speckle extraction accuracy is further improved by further combining amplitude information with phase information.

  Further, as shown in claim 10, the ultrasonic image processing apparatus according to any one of claims 1 to 9, wherein the phase component has a data resolution in a direction perpendicular to a traveling direction of the ultrasonic signal. It is more than the arrangement interval of the element which transmits / receives the said ultrasonic signal of the said ultrasonic transmission / reception means, It is characterized by the above-mentioned.

  This improves the resolution and further improves the accuracy of speckle extraction.

  Similarly, in order to achieve the object, the invention according to claim 11 transmits an ultrasonic wave toward the subject, receives an ultrasonic wave reflected from the subject, and receives the received ultrasonic wave. Provided is an ultrasonic image processing method characterized by extracting a speckle component from a phase component of a sound wave signal.

  According to this, since speckles are extracted using phase information rather than amplitude information, speckles can be extracted independent of speckle shape and surrounding echo levels.

  Similarly, in order to achieve the above object, the invention according to claim 12 is characterized in that an ultrasonic wave is transmitted to the subject and an ultrasonic wave reflected from the subject is received. An ultrasonic image processing program is provided that causes a computer to realize a function of extracting a speckle component from a phase component of an ultrasonic signal obtained in this manner.

  As a result, speckles are extracted using phase information instead of amplitude information, so speckles can be extracted independent of speckle shape and surrounding echo levels.

  As described above, according to the present invention, speckles are extracted using phase information rather than amplitude information, so that speckles can be extracted independent of speckle shape and surrounding echo levels. .

  Hereinafter, an ultrasonic image processing apparatus, method, and program according to the present invention will be described in detail with reference to the accompanying drawings.

  FIG. 1 is a block diagram showing a schematic configuration of an embodiment of an ultrasonic diagnostic apparatus including an ultrasonic image processing apparatus according to the present invention. This ultrasonic diagnostic apparatus captures and displays an ultrasonic image of a diagnostic region of a subject using ultrasonic waves, and mainly includes an ultrasonic probe 10, a signal processing unit 20, and image processing. The unit 30 and the display unit 40 are included.

  The ultrasonic probe 10 transmits ultrasonic waves toward a diagnosis site in the body of the subject and receives ultrasonic waves reflected in the body of the subject. In other words, since the ultrasonic waves are reflected at the boundary between regions having different acoustic impedances, such as the boundary between structures, an ultrasonic beam is transmitted into a subject such as a human body and an ultrasonic echo generated within the subject is received. The contour of the structure existing in the subject can be detected by obtaining the reflection point and reflection intensity at which the ultrasonic echo is generated.

  The ultrasonic probe 10 includes, for example, a plurality of ultrasonic transducers constituting a one-dimensional ultrasonic transducer array, and each ultrasonic transducer is formed by a vibrator in which electrodes are formed at both ends of a piezoelectric element such as PZT. Although configured, it is not particularly limited. For example, in addition to a linear array probe in which a plurality of ultrasonic transducers are arranged in one dimension in this way, a sector probe that scans the inside of a subject in a fan shape, a convex array probe in which a plurality of ultrasonic transducers are arranged on a convex surface, Alternatively, a two-dimensional array probe in which a plurality of ultrasonic transducers are two-dimensionally arranged may be used. Alternatively, it may be a mechanical radial probe that performs radial scanning in an ultrasonic endoscope.

  The ultrasound probe 10 transmits an ultrasound beam to the subject under the control of a control unit (not shown), and is scanned by a scanning method such as linear scanning, sector scanning, convex scanning, or radial scanning. Scan the examiner. The ultrasonic waves generated by the ultrasonic probe 10 are reflected by a reflector present in the body of the subject, and the reflected ultrasonic waves are received by the ultrasonic probe 10. The ultrasonic reception signal received by the ultrasonic probe 10 is converted into an electric signal and then delivered to the signal processing unit 20 via a transmission / reception unit (not shown).

  The signal processing unit 20 performs predetermined signal processing on the input received signal, phase matching addition for adding a plurality of received data with delay from the received ultrasonic received signal, and the received data as IQ data (complex Signal) to generate IQ data and output it to the image processing unit 30.

  Note that the amplitude A and phase θ of the received data are calculated from the IQ data by the following equations.

A = √ (I ² + Q ² ), θ = arctan (Q / I)
Here, the symbol √ (X) represents the square root of X. Further, when taking the envelope of the waveform of the received data, the amplitude must be 2A by doubling the above value.

  The signal processing unit 20 may use a method of generating data for each line. For example, as described in Japanese Patent Application Laid-Open No. 2003-180688, the reception data of all elements is stored in a memory, A method of generating data by processing may be used. With this method, the azimuth resolution (element direction resolution) can be further improved. For example, high-resolution phase information is used in the direction perpendicular to the traveling direction of the ultrasonic signal so that the resolution of the phase information is higher than the element interval in the direction perpendicular to the traveling direction of the ultrasonic signal. , Speckle random phase changes can be more accurately distinguished.

  Further, in a general ultrasonic B-mode image, an image is generated from amplitude and phase information is not used. Therefore, ultrasonic image data in the existing ultrasonic image processing technique (B mode) indicates amplitude data.

  Next, the relationship between signal interference and phase will be described with reference to the drawings. Interference includes constructive interference and canceling interference. Constructive interference occurs when the phase difference between the waves is small, and destructive interference occurs when the phase difference is close to π.

  FIG. 2 shows an example of constructive interference. FIG. 2A shows the state before interference, and three waves with small phase differences of 0.2 (rad), 0.4 (rad), and 0.6 (rad) are used for the reference wave. To interfere. FIG. 2B shows the state after interference, and the interference wave is represented by a solid line with respect to the reference wave represented by a broken line. When the phase difference is small in this way, an intensifying interference wave is obtained. As can be seen from FIG. 2B, the peak of the peak of the interference wave is close to the peak of the peak of the reference wave, and the interference wave having a small phase difference has a small phase difference from the reference wave even after the interference.

  FIG. 3 shows an example of destructive interference. FIG. 3A shows the state before interference. In this case, a phase difference of 3.0 (rad) and a phase difference of 3.2 (rad) in addition to the phase difference of 0.2 (rad) with respect to the reference wave. Is close to π (rad), causing large waves to interfere. At this time, as shown in FIG. 3B, a destructive interference wave is obtained as represented by a solid line with respect to a reference wave represented by a broken line. When waves having such a large phase difference are caused to interfere with each other, destructive interference waves are obtained. As can be seen from FIG. 3B, the peak of the peak of the interference wave is separated from the peak of the peak of the reference wave, and the interference wave having a large phase difference has a large phase difference from the reference wave even after the interference. .

  Previously, speckle noise was said to be caused by random interference of reflected ultrasonic waves from a plurality of reflectors in the subject, but interference with a large phase difference as shown in FIG. 3 generates speckle noise. Interference.

  Further, the phase obtained from the IQ data represents the phase difference between the wave after interference and the wave after detection. Since the phase of IQ detection includes an error, it is a result of interference having a large phase difference from the wave after interference. In order to confirm whether the result is small interference, it is only necessary to look at a continuous phase change of IQ data (for example, a difference between adjacent pixels). This is because when a wave having a small phase difference continues, the phase change is small, and when the speckle noise portion is reached, the phase changes suddenly and the phase change increases.

  In this way, speckle noise can be determined by using the feature that the phase changes suddenly.

  The image processing unit 30 performs image processing on the IQ data obtained by the signal processing unit 20 and displays the IQ data on the display unit 40. The center of processing in the image processing unit 30 is to extract speckles by extracting phase information from the IQ data of the signal processing unit 20 and discriminating speckles.

  FIG. 4 shows the flow of speckle extraction processing using phase information in the image processing unit 30.

  As illustrated in FIG. 4, the image processing unit 30 includes a phase change amount extraction unit 32 and a speckle determination unit 34.

  Phase data is input to the phase change amount extraction unit 32 from the IQ data obtained by the signal processing unit 20. As described above, the phase is given by the equation θ = arctan (Q / I). The phase change amount extraction unit 32 extracts the phase change amount from the phase data. The extracted phase change amount data is input to the speckle determination unit 34. The speckle determination unit 34 extracts speckles using the phase change amount data and outputs a speckle extraction result.

  FIG. 5 shows an example of processing in the phase change amount extraction unit 32.

  When the phase change amount extraction unit 32 obtains the phase data, first, the phase change amount in the azimuth direction (lateral direction) and the phase change amount in the distance direction (depth direction), that is, the pixel difference (azimuth direction difference) in each direction. And the distance direction difference).

  In the case of a two-dimensional B-mode image, it is desirable to determine the amount of phase change in the azimuth direction and the distance direction in this way. This is because when speckle noise exists parallel to either the azimuth direction or the distance direction, the phase change is small in that direction, but the phase change is large in the direction orthogonal thereto.

  Next, a certain amount of phase change due to detection deviation is removed. Note that the speckle phase change changes abruptly compared to other parts, so only the components that increase the amount of change, such as differential data from the filter (low-pass filter, median filter) and secondary differentiation of the phase data, are extracted. By doing so, it is possible to obtain data having characteristics that make it easier to discriminate speckle.

  Then, a component whose phase changes suddenly in each direction is extracted, and phase change amount data in each direction is calculated. The amount of change is obtained from the difference between pixels in each direction. At this time, it is most preferable to use a difference between adjacent pixels, but a difference between pixels thinned out as appropriate may be used.

  In this manner, speckle / signal determination can be performed based on information obtained by removing a fixed amount of phase change in each of the vertical and horizontal directions.

  Note that the phase change amount is not limited to this information, and a difference average of a plurality of pixels or a difference in an oblique direction may be used. Furthermore, instead of separate change amounts, the data may be converted into one data such as a vector component.

  In the speckle discrimination unit 34, the obtained phase change data is subjected to binary discrimination as to whether it is speckle noise or multi-level discrimination as to how much speckle noise is included. Results are output according to purpose. Of course, the phase change amount itself may be output as data indicating speckle-likeness.

  The azimuth direction phase change amount data and the distance direction phase change amount data extracted by the phase change amount extraction unit 32 are input to the speckle determination unit 34. The speckle discriminating unit 34 discriminates speckles based on these phase information data.

  FIG. 6 shows a configuration example of the speckle determination unit 34.

  As shown in FIG. 6, the speckle discrimination unit 34 includes a discrimination function creation unit 341. The discriminant function creating unit 341 creates a discriminant function for discriminating whether or not speckle noise is generated from the phase change amount data.

  The discriminant function creating unit 341 prepares in advance phase data in which the speckle or non-speckle position is known in the amplitude image, and creates a discriminant function based on the feature amount created from the phase change amount. That is, a predetermined feature amount is calculated in the feature amount conversion unit 342 from the phase change amount data 344a known to be speckle and the phase change amount data 344b known to be non-speckle, From this, a speckle discriminant function is created.

  Here, the feature amount may be a single pixel of phase change amount data. However, in the case of a single pixel, when the phase change amount is obtained by taking the difference, a slight pixel shift occurs. When the phase change amount is obtained from the vertical and horizontal directions, the phase of the center of the crossing portion is obtained. Since there is a problem that the change does not increase, it is desirable to use a plurality of data such as the value of the neighboring pixel of the target pixel and the calculation result with the neighboring pixel. At this time, since a plurality of data are multidimensional, the dimension may be lowered by performing PCA (principal component analysis) or the like so that the threshold value can be easily designed.

  FIG. 9 shows an example of the speckle discriminant function.

  Here, the vertical direction phase change amount and the horizontal direction phase change amount are defined as a feature amount (1) and a feature amount (2), respectively, non-speckle noise is represented by ◯, and speckle noise is represented by x. In the example shown in FIG. 9, the discriminant function is set as a straight line that separates the non-speckle noise O and speckle noise x regions. A function (or threshold) for discriminating between speckle noise and non-speckle noise is designed based on the result converted into the feature quantity in this way. The discriminant function is not limited to such a linear function.

  The discriminant function setting method is not particularly limited. For example, a statistical method using known data (learning data) such as SVM (support vector machine) (for example, Nero Christianini as a reference). And a linear or non-linear discriminant function used for known classification such as “Introduction to Support Vector Machine” by John Taylor, etc.). Of course, speckle may be determined only by the threshold value if it can be determined only by giving a threshold value for each feature amount. In addition, a discrimination function may be set with data representing a phase change, such as pixels of continuous phase data, as a feature quantity without being converted into a phase change amount.

  FIG. 7 shows an example of a speckle extraction discriminant function generation process in the feature quantity conversion unit 342 using SVM (support vector machine).

  As shown in FIG. 7, first, the speckle part and the non-speckle part are manually labeled from the phantom image to create known data for creating a speckle discriminant function. When labeling speckles and non-speckle parts, it is not necessary to perform labeling for ambiguous parts. Next, a feature amount used for discrimination of speckle is extracted from the speckle portion and the non-speckle portion of the known data.

  Here, as a feature amount, as shown in FIG. 8, a central pixel c of 3 × 3 pixels is set as a target pixel, and the target pixel c and four neighboring pixels a, b, d, and e above, below, left, and right are vertically A total of ten feature quantities, ie, (distance) direction phase change and lateral (azimuth) direction phase change are used.

  Next, a discriminant function (speckle discrimination) is applied by applying SVM (support vector machine) with the phase change amount of a total of 10 positions of the labeled pixels (the target pixel and its vicinity) in the vertical direction and the horizontal direction as feature amounts. Generator). Of course, the data and features used for labeling may be changed to optimize the discrimination results.

  In this manner, the discriminant function creation unit 341 creates a speckle discriminant function by extracting feature amounts from known data that are known to be speckle and non-speckle in advance. In actual ultrasonic diagnosis, speckle extraction is performed based on phase information obtained from IQ data generated by the signal processing unit 20 from data obtained by scanning the ultrasonic probe 10.

  That is, in FIG. 6, when the azimuth (transverse) direction phase change amount data 346a and the distance (vertical) direction phase change amount data 346b are input, the feature amount conversion unit 343 converts them into feature amounts used for speckle discrimination. Then, using the speckle discriminant function 347 created in advance by the discriminant function creating unit 341, it is discriminated whether it is speckle, and speckle extraction is performed. Then, the speckle extraction result 348 is output and displayed on the display unit 40. In this display, only speckles may be displayed or may be displayed so as to be superimposed on the amplitude image that is the original image.

  The discrimination result may not only indicate whether the speckle is binary, but also output the difference from the threshold as speckle-likeness indicating how much speckle noise is included in multiple values. . In the case of multi-value output, the value may be further adjusted using a LUT (Look Up Table) or the like.

  Note that the speckle discriminant function changes depending on the condition of ultrasonic transmission / reception, and therefore, in the case of an actual apparatus, it is desirable to set the discriminant function for each condition.

  As described above, in this embodiment, speckles are extracted using phase information instead of amplitude information. However, it is difficult to distinguish speckles depending on the shape of speckles and the echo level. In some cases, however, speckles cannot be accurately extracted, but by using phase information, speckles can be extracted independent of speckle shape and surrounding echo levels.

  Thus, speckles can be extracted only by phase information, but speckles are extracted by combining amplitude information (amplitude component or its envelope component) and phase information, such as adding amplitude information to a feature quantity. It may be. In this case, for example, in FIG. 6, phase information (azimuth direction phase change data 346a and distance direction phase change data 346b) is input to the feature value conversion unit 343. Speckle extraction may be performed by inputting to the conversion unit 343 and adding the amplitude information (pixel value) to increase the dimension of the feature amount (two dimensions in the example shown in FIG. 9).

  Thus, speckle extraction using both amplitude information and phase information enables more accurate speckle extraction.

  Further, the ultrasonic image processing apparatus of the present embodiment is controlled by an ultrasonic image processing program stored in a memory attached to a control unit (not shown). That is, an ultrasonic image processing program is read from the memory by the control unit, and according to the ultrasonic image processing program, an ultrasonic wave is transmitted to the subject and an ultrasonic wave reflected from the subject is received. The function and the function of extracting the speckle component from the phase component of the ultrasonic signal obtained by reception are executed.

  Note that the ultrasonic image processing program is not limited to the one stored in the memory attached to the control unit in this way, and the ultrasonic image processing program is an ultrasonic image processing apparatus such as a PC card or a CD-ROM. It may be recorded in a memory medium (removable medium) configured to be detachable and read into the apparatus via an interface corresponding to the removable medium.

  Although the ultrasonic image processing apparatus, method, and program of the present invention have been described in detail above, the present invention is not limited to the above examples, and various improvements and modifications can be made without departing from the spirit of the present invention. Of course you can go.

1 is a block diagram showing a schematic configuration of an embodiment of an ultrasonic diagnostic apparatus including an ultrasonic image processing apparatus according to the present invention. It is a graph which shows the example of constructive interference, (a) represents before interference, (b) represents after interference. It is a graph which shows the example of the destructive interference, (a) represents before interference, (b) represents after interference. It is a block diagram which shows the flow of the speckle extraction process using the phase information in an image process part. It is a block diagram which shows an example of the process in a phase change amount extraction part. It is a block diagram which shows the example of 1 structure of a speckle discrimination | determination part. It is a block diagram which shows an example of the discriminant function production | generation process of the speckle extraction using SVM (support vector machine) in a feature-value conversion part. It is explanatory drawing which shows the example of the pixel used as a feature-value. It is explanatory drawing which shows the example of a speckle discriminant function.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 10 ... Probe for ultrasonic waves, 20 ... Signal processing part, 30 ... Image processing part, 40 ... Display part, 32 ... Phase change amount extraction part, 34 ... Speckle discrimination | determination part, 341 ... Discrimination function creation part, 342, 343 ... feature amount conversion unit

Claims (12)

An ultrasonic transmission / reception means for transmitting ultrasonic waves toward the subject and receiving ultrasonic waves reflected from the subject;
Speckle component extraction means for extracting a speckle component from a phase component of an ultrasonic signal received by the ultrasonic transmission / reception means;
An ultrasonic image processing apparatus comprising:   2. The ultrasonic image processing apparatus according to claim 1, further comprising display means for displaying an extraction result extracted by the speckle component extraction means.   3. The ultrasonic image processing apparatus according to claim 1, wherein the speckle component extraction unit extracts a speckle component from the phase component of the ultrasonic signal using a change amount of the phase component. An ultrasonic image processing apparatus.   3. The ultrasonic image processing apparatus according to claim 1, wherein the speckle component extraction unit uses a change amount of at least two directions of the phase component. 4.   5. The ultrasonic image processing apparatus according to claim 1, wherein the speckle component extraction unit uses a discriminant function for discriminating speckles. 6. 5. The ultrasonic image processing apparatus according to claim 3 , wherein the speckle component extraction unit uses a threshold value that discriminates speckle in comparison with a change amount of the phase component. Sonic image processing device.   6. The ultrasonic image processing apparatus according to claim 5, wherein the discriminant function is set by a feature amount generated from a phase component whose positions of speckle and non-speckle are known. Sonic image processing device.   8. The ultrasonic image processing apparatus according to claim 7, wherein the feature amount is a change amount of a phase component. An ultrasonic transmission / reception means for transmitting ultrasonic waves toward the subject and receiving ultrasonic waves reflected from the subject;
Speckle component extraction means for extracting a speckle component from a combination of an amplitude component or an envelope component and a phase component of an ultrasonic signal received by the ultrasonic transmission / reception means;
An ultrasonic image processing apparatus comprising:   10. The ultrasonic image processing apparatus according to claim 1, wherein the phase component has a data resolution in a direction perpendicular to a traveling direction of the ultrasonic signal, and the ultrasonic transmission / reception unit has the ultrasonic resolution. An ultrasonic image processing apparatus characterized by being at least an arrangement interval of elements for transmitting and receiving a sound wave signal. While sending ultrasonic waves toward the subject, receiving the ultrasonic waves reflected from the subject,
An ultrasonic image processing method, wherein a speckle component is extracted from a phase component of a received ultrasonic signal. A function of transmitting ultrasonic waves toward the subject and receiving ultrasonic waves reflected from the subject;
A function of extracting speckle components from the phase component of the ultrasonic signal obtained by reception;
An ultrasonic image processing program characterized in that a computer is realized.

Images