JPH1091779A - Digital x-ray image processor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To suppress the halation by using the gradation conversion means having different linear conversion characteristics at each stage of input data, detecting the luminance in a region of interest out of the data received from an A/D conversion means, and setting the linear conversion characteristic of each stage according to the detected luminance. SOLUTION: For instance, the X-ray video signals sent from a camera of a medical X-ray TV system are inputted to an A/D converter 11. The converter 11 has 12 output bits and also has the density resolution higher than a clinically useful gradation range by several times. Thus no halation is caused by the converter 11. In other words, 9 bits of gradation are usually required for the fluoroscopic images and the digital radiography and accordingly the conversion 11 has the reproducibility up to the signal intensity of 8 times as much as the normal intensity with no halation. Furthermore, the halation is suppressed down to the signal intensity of 4 times for a digital subtraction image that requires just 10 bits.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an X-ray image signal processing apparatus used for medical diagnosis and the like, and more particularly to a digital X-ray image processing apparatus for digitizing and processing an X-ray image signal.

[0002]

2. Description of the Related Art An X-ray image has a very wide density distribution from a dark portion to a bright portion. Therefore, it is necessary to prevent the halation by compressing the contrast of the bright part. Conventionally, gradation conversion (input / output conversion) of an X-ray video signal is performed by an analog circuit to suppress halation.

[0003]

However, in the conventional circuit, there is a problem that it is difficult to enhance contrast in a low luminance (low density) portion in order to obtain an effect of suppressing halation. Therefore, in the case of an X-ray video signal having a large halation portion, the contrast of a relatively low-brightness portion where clinically useful data exists is reduced, and adverse effects such as a problem in reading the useful data occur. Would.

In view of the above, the present invention provides a digital X-ray system which performs so-called γ-characteristic conversion for enhancing the contrast in a low-brightness portion and suppresses halation by compressing the contrast in a high-brightness portion. It is an object to provide an image processing device.

[0005]

In order to achieve the above object, a digital X-ray image processing apparatus according to the present invention provides a practical density (luminance) resolution for converting an input X-ray video signal into digital data. A / D conversion means having a higher number of gradations (density resolution), and data from the A / D conversion means are inputted, and different linear conversion characteristics are obtained at each stage where the dynamic range of the input data is divided into multiple stages. And a means for detecting the luminance in the region of interest from the data from the A / D conversion means, and a means for setting the linear conversion characteristics at each of the stages according to the detected luminance. It is characteristic.

[0006] The A / D conversion means is used for an X-ray fluoroscopic image, a digital radiographic image or a digital subtraction image, or when performing other image processing. The X-ray video signal is converted into digital data with the number of gradations (the number of output bits) several times as large as the number of tones. Therefore, halation does not occur in the A / D conversion part. The digital data is subjected to gradation conversion (input / output conversion) by gradation conversion means, which has different linear conversion characteristics at each stage of dividing the dynamic range of the input data into multiple stages. Therefore, a non-linear conversion can be performed in a pseudo manner as a whole. Therefore, it is possible to suppress the halation by compressing the contrast of the high luminance portion while performing the conversion of the γ characteristic for enhancing the contrast of the low luminance portion. Moreover, since the luminance in the region of interest is detected and the linear conversion characteristics at each of the above stages are set according to the detected luminance, it is possible to change the characteristics of the pseudo non-linear conversion according to the change in the image for each frame. Yes, Dynamic Height Control (DH
C) can be realized.

[0007]

Next, embodiments of the present invention will be described in detail with reference to the drawings. In FIG. 1, an A / D converter 11 includes, for example, a medical X-ray TV.
An X-ray video signal sent from a camera of the system is input. Since the A / D converter 11 has 12 output bits and has a density resolution several times higher than a clinically useful gradation range, halation occurs in the A / D conversion. There is nothing.
That is, since the number of tones normally required for an X-ray fluoroscopic image or digital radiography is usually 9 bits (512 tones), it is possible to reproduce the signal intensity up to eight times the signal intensity without halation. Further, when used as a digital subtraction image or when other image processing is performed, only 10 bits (1024 gradations) are required, so that up to four times the signal intensity can be reproduced without halation.

The A / D converter 11 has a sampling frequency of, for example, 40 MHz, and corresponds to a video signal of an X-ray image having a spatial resolution of about 1000 horizontal scanning lines and a resolution of about 1000 horizontal scanning directions. I can do it. The signal sampled at this frequency is output as 12-bit digital data. In other words, this data is obtained by expressing the value (density value, luminance value) of each pixel by 12 bits.

The digital data from the A / D converter 11 is sent to a gradation conversion circuit 12 where the digital data is subjected to gradation conversion, and 12-bit data is converted to 10-bit data. The gradation conversion circuit 12 is configured to realize a non-linear conversion characteristic as a whole by combining a large number of linear conversion circuits as shown in FIG. 2, for example.

In FIG. 2, a multiplier 21 performs an operation of multiplying the input digital data by a predetermined coefficient.
Thereafter, the adder 22 adds a predetermined addition value. Therefore, the multiplication coefficient corresponds to the gain of the amplifier, and the added value corresponds to the offset value (initial value at input 0). As described above, the multiplier 21 and the adder 22 constitute a linear conversion circuit, and here, 32 linear conversion circuits are provided in parallel. Then, each of the 32 kinds of gain and offset values is set to the γ coefficient setting circuit 15.
(FIG. 1). The outputs of the 32 adders 22 are selected by the output selector 23 and output.
This output selector 23 is used for the γ coefficient selection circuit 13 (FIG. 1).
The output of the adder 22 is selected according to the range in which the value of the input digital data falls.

Therefore, the gradation conversion circuit 12 has polygonal conversion characteristics A, B, C, D, E,... As shown in FIG.
Will have. The conversion characteristics A, B,
The slope of each of the many line segments constituting each of C, D, E,... Is determined by the gain given to the multiplier 21, and the initial value at the input value 0 of the line segment is added to the adder 22. It is determined by the offset value. The gamma coefficient selection circuit 13 monitors which range of the input value I falls within the range of 0 to I1, I1 to I2,..., and accordingly has a linear characteristic having a conversion characteristic (expressed by a line segment) of the range. The output selector 23 is controlled so as to select the output of the conversion circuit (the output of the adder 22).

As described above, since the gradation conversion circuit 12 has a polygonal conversion characteristic, it is desirable to suppress the halation in the high luminance portion while performing the γ characteristic conversion for enhancing the contrast in the low luminance portion. Any transformation of the properties that seems to be possible can be performed. By giving a polygonal conversion characteristic represented by 32 types of straight lines as in this example, it is possible to simulate a smooth conversion of a nonlinear characteristic having a smooth curve.

Here, the reason why the polygonal conversion characteristics are provided like A, B, C, D, E,... In FIG. 3 is because the optimum conversion characteristics are used depending on the state of the image. It is. The conversion characteristic is determined by the γ coefficient setting circuit 1
5, the gamma coefficient setting circuit 15
It operates according to the output of.

Digital data from the A / D converter 11 is sent to the integration circuit 14, and data of the region of interest set by the region of interest setting circuit 16 is extracted from the data and included in the region. The integrated value of the data of all pixels to be obtained is obtained. If this integral value is small (integral value a), the average luminance of the region of interest is low, so the conversion characteristic should be such that the contrast enhancement of the low luminance portion like the conversion characteristic A should be performed. Conversely, if the integral value is large (integral value a) Value e) Since the average luminance of the region of interest is high, a conversion characteristic such as the conversion characteristic E that does not significantly compress the contrast in a high luminance portion should be used.

Therefore, here, the switching points of the straight lines (range boundaries) I1, I2, I3,... Are fixed (the range monitored by the γ coefficient selection circuit 13 is fixed).
The output value O at the switching points I1, I2, I3,.
Is changed according to the integral value as shown in FIG. For example, the integral a
, The output value Oa1 is obtained at the switching point I1, and the output value Oa2 is obtained at the switching point I2. When the output value O at the switching point I1 increases to the integral values b, c, d, e,.
Each line segment of the polygonal conversion characteristic is determined so as to gradually decrease as b1, Oc1, Od1, Oe1,. That is, the gamma coefficient setting circuit 15 obtains the gain given to each of the many multipliers 21 and the offset given to each of the many adders 22, and then sets these values in the gradation conversion circuit 12.

In practice, typical integral values (a, b, c,
d, e,...) are stored in a table, and the output values at the switching point corresponding to the intermediate integral value are obtained by interpolation calculation such as the least square method. From the values, the slope (gain) and initial value (offset) of each line segment are obtained. Here, after calculating an integral value for an input video signal of a certain frame, calculating each gain and each offset, and in a blanking period before the next frame starts, the obtained gain and offset are converted to a gradation conversion circuit. 12 is given. That is, the optimum conversion characteristic for the region of interest image is determined in real time with a delay of one frame. X-ray images usually have temporal continuity and are almost the same before one frame,
There is no problem in setting the conversion characteristic with a delay of one frame. Of course, in a special case where it is necessary to avoid such a one-frame delay, the gradation conversion circuit 1
A delay circuit for delaying the data input to 2 by one frame may be provided before the gradation conversion circuit 12 (between the A / D converter 11).

As described above, the pseudo-nonlinear conversion characteristic is dynamically changed in real time in accordance with the pixel value (luminance) of the region of interest. Halation suppression can be combined with balance.

The conversion characteristic shown in FIG. 3 is an example, and the density resolution may be secured without changing the conversion characteristic of the low luminance portion as shown in FIG. 5, for example. Also in FIG. 5, in the range larger than the input value I1,
As the integral value increases, A, B, C, D, E,
.. And the conversion characteristics are changed.

In addition, the specific configuration can be variously changed. For example, as shown in FIG. 2, the gradation conversion circuit 12 provides a number of combinations of the multiplier 21 and the adder 22 in parallel, operates them in parallel, and outputs one of the outputs according to the input value. Although the selection is made by the selector 23, it is also possible to provide a single system of the multiplier 21 and the adder 22, and to switch the gain and the offset for each pixel by the γ coefficient selection circuit 13. in this case,
For example, assuming that the non-linear conversion is approximated by 32 straight lines in the same manner as described above, the 32 kinds of gain and offset values obtained according to the integrated value one frame before are written in a memory or the like, and the current When pixel values (digital data) of a frame are sequentially input, a corresponding value is read out according to each of the pixel values (in accordance with which range is entered), and given to the multiplier 21 and the adder 22. What is necessary is just to be a structure.

[0020]

As described above, according to the digital X-ray image processing apparatus of the present invention, gradation conversion is performed by a polygonal conversion characteristic composed of multi-stage straight lines. Thus, while performing the conversion of the γ characteristic for enhancing the contrast of the low luminance portion, the contrast of the high luminance portion can be compressed to suppress the halation, and the ideal gradation conversion can be performed. Further, since the pseudo-nonlinear conversion characteristic is dynamically changed depending on the state of the image of the region of interest for each frame, even in the case of an X-ray image that changes over time like a moving image, the characteristic of the region of interest is not changed. , And more appropriate gradation conversion can be performed by following the fluctuation in real time.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an embodiment of the present invention.

FIG. 2 is a block diagram showing a part of the gradation conversion circuit of FIG. 1 in further detail;

FIG. 3 is a graph showing gradation conversion characteristics.

FIG. 4 is a graph showing a relationship between an integral value of a pixel value in a region of interest and a conversion characteristic.

FIG. 5 is a graph showing another gradation conversion characteristic.

[Explanation of symbols]

 Reference Signs List 11 A / D converter 12 Gradation conversion circuit 13 γ coefficient selection circuit 14 Integrator circuit 15 γ coefficient setting circuit 16 Region of interest setting circuit 21 Multiplier 22 Adder 23 Output selector

Claims (1)

[Claims] 1. An A / D converter for converting an input X-ray video signal into digital data, which has a number of gradations exceeding a practical density resolution.
A / D conversion means, gradation conversion means to which data from the A / D conversion means is inputted, and which have different linear conversion characteristics at each stage where the dynamic range of the input data is divided into multiple stages; A digital X-ray image processing apparatus comprising: means for detecting a luminance in a region of interest from data from a converting means; and means for setting the above-described linear conversion characteristics at each stage according to the detected luminance.