JP3744041B2 - Radiation imaging device - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an imaging apparatus for radiographic images including X-ray images used for medical diagnosis and the like.
[0002]
[Prior art]
As an apparatus that obtains an image of an object using radiation, there is a radiation imaging apparatus that makes a radiation bundle including subject information from a radiation source enter an array sensor and expresses intensity information on a screen such as a CRT. Also, in this type of imaging apparatus, a so-called photon counting method is conventionally employed in which pulse signals generated from a sensor upon incidence of radiation photons are counted and the counted value is used as image density information.
[0003]
By the way, since a radiation sensor used for taking such a radiographic image is required to have excellent sensitivity to radiation, density resolution and spatial resolution, a compound semiconductor such as CdTe is generally used as a sensor material. Has been. For this reason, some array sensors lack temporal stability and their characteristics change greatly each time they are measured. When an image is captured by scanning a scanning array sensor having such a temporal sensitivity change, even if the sensor sensitivity between adjacent channels deviates by about 1%, artifacts in the scanning direction The image will be visible.
[0004]
In order to solve this problem, conventionally, a calibration processing method has been used in which a phantom having a known radiation absorption coefficient and thickness is photographed, the characteristic curve of each sensor of the array sensor is obtained using the photographed data, and the original image is calibrated. It is taken.
[0005]
[Problems to be solved by the invention]
However, when calibrating the original image using the data obtained by phantom shooting, if you want to obtain a smooth calibration image that prevents the effects of artifacts, the time difference between phantom shooting and original image shooting must be extremely short. For this reason, every time an original image is taken, it is necessary to take phantom photography and take calibration data.
[0006]
The present invention has been made in view of such a situation, and radiation imaging that makes it possible to obtain a smooth calibration image in which artifacts are not visible even if data for sensor calibration is not collected every time an original image is captured. The purpose is to provide a device.
[0007]
[Means for Solving the Problems]
In order to achieve the above object, the radiation image capturing apparatus of the present invention collects the output of each sensor while moving an array sensor in which a plurality of radiation sensors are arranged stepwise on the imaging region, and collects the collected data. In an apparatus for obtaining a radiological image based on the above, an original image obtained from raw data collected by scanning the array sensor based on data collected by the array sensor or data collected by a reference image sensor And an image creation unit that creates a reference image that is smoother than the original image, and data of channels that correspond to each other between the reference image obtained by the image creation unit and the original image A calculation unit that obtains a characteristic curve of each sensor by obtaining a curve that approximates the correlation and uses the characteristic curve. It characterized by that it comprises a calibration processor to calibrate the original image data Te.
[0008]
With the above configuration, a smooth calibration image can be obtained without impairing the spatial resolution and density resolution of the original image.
The principle will be described below with reference to FIGS. 3 and 4 show an example of calibration of data collected from one channel (x-ch) sensor among images taken by scanning the array sensor.
[0009]
The original image (Fig. 3 (A)) obtained from the raw data collected by scanning the array sensor has good spatial resolution and density resolution, but has not been subjected to sensitivity correction. The image will be visible.
[0010]
Therefore, in the present invention, first, a smooth reference image [FIG. 3 (B)] with an emphasis only on temporal stability is created with an image corresponding to the original image. Next, the data of points PA and PB corresponding to each other between the created reference image and the original image are correlated, and the correlation is approximated to an appropriate curve m as shown in FIG. Is obtained.
[0011]
At this time, the equation of the approximate curve is arbitrary, but if a polynomial with a coefficient as a parameter is used, the linear least square method can be adopted, and the calculation of the equation is only addition and multiplication. For example, a cubic equation [a + b · (count) + c · (count) ² + d · (count) ³ ... (1)] having an intercept is used as a calculation formula. The coefficients a, b, c and d are obtained.
[0012]
The curve of the equation (1) obtained by the above processing becomes the characteristic curve of the sensor of the channel (x-ch), and the coefficients a, b, c and d of the terms of the equation (1) are the characteristics of the sensor. It becomes a parameter. Therefore, the original image data of one channel (x-ch) sensor can be calibrated by substituting the value of the original image data into (count) of each term of equation (1), and these calculation processes Is executed for all the channel sensors, and a smooth configuration image can be obtained.
[0013]
Here, in the present invention, a method for creating a reference image is different from, for example, a sensor that collects data of an original image, and for a reference image in which spatial resolution and density resolution are rough and importance is placed only on temporal stability. Data is collected by the sensor and a reference image is obtained from the collected data, or the original image data collected by scanning the array sensor is processed, and the spatial resolution and density resolution are not so precise. A method for obtaining a smoother reference image than the image is exemplified.
[0014]
Note that the smooth image referred to in the present invention refers to an image having no extreme change in density between consecutive pixels.
[0015]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, embodiments of the present invention will be described with reference to the drawings.
First, as shown in the conceptual diagram of FIG. 2, the radiation imaging apparatus to which the present invention is applied includes a radiation source 10 for irradiating a subject T and radiation sensors S ·· S in a two-dimensional form (n columns × multiple rows). And an array sensor A arranged in a step-like manner and moving the array sensor A in a step shape behind the subject T to obtain an image.
[0016]
FIG. 1 is a block diagram showing the configuration of the embodiment of the present invention.
The radiation imaging apparatus apparatus of FIG. 1 includes an original image photographing unit 1, a reference image creating unit 2, a characteristic parameter calculating unit 3, a calibration processing unit 4, and a storage unit 5 in which an information table for creating a reference image is set. It is configured.
[0017]
The original image capturing unit 1 includes the array sensor A shown in FIG. 2 and an amplifier, a comparator, and a counter (none of which are shown) connected to each of the radiation sensors S ·· S, and the photon counting method is used. The image data is collected based on the image data. The original image data output from the original image photographing unit 1 is guided to the reference image creating unit 2 and the characteristic parameter calculating unit 3.
[0018]
The reference image creation unit 2 is a functional unit that creates the reference image corresponding to the original image by processing the original image data from the original image photographing unit 1.
The creation process first reads out the sensor characteristics for each channel of the array sensor A from the information table set in the storage unit 5, the position information of a specific sensor having a low sensitivity in the array sensor A, and the like. This process is executed by converting the original image data based on the information, and this process provides a smooth reference image with better spatial stability and better temporal stability than the original image. .
[0019]
Here, among the information set in the information table of the storage unit 5, for the characteristic information of each channel, for example, image data collected by scanning the array sensor A in the original image photographing unit 1 is used, and the data is stored in a predetermined number. These pixels are created by performing a simple addition for each pixel group, obtaining an addition average, and performing linear interpolation. Further, the position information of the unique sensor is obtained by a process of searching for image data collected by the above photographing that has an abnormally bad sensitivity. However, these pieces of information do not need to be collected every time an original image is taken, and may be performed at a frequency of about once a year, for example.
[0020]
Note that the above reference image creation processing method is merely an example, and other methods may be adopted.
On the other hand, the characteristic parameter calculation unit 3 correlates the corresponding data of the reference image created by the reference image creation unit 2 and the original image from the original image photographing unit 1, and the correlation is expressed by a curve [a + b · (count) + C · (count) ² + d · (count) ³ ... (1)] (see FIG. 4), and a calculation process such as obtaining a, b, c, d, which are sensor characteristic parameters, It is configured so as to obtain a characteristic curve for each channel by performing the process for the sensor of the channel.
[0021]
The calibration processing unit 4 is a functional unit that calibrates the original image data from the original image photographing unit 1 using the characteristic curve calculated by the characteristic parameter calculation unit 3. The process of substituting the value (count) of the original image data into (count) of the term is executed for the sensors of all the channels, and a smooth calibration image is obtained by the process in the calibration processing unit 4. can get.
[0022]
The curve equation used when fitting the correlation between the reference image data and the original image data in the characteristic parameter calculation unit 3 is not limited to the above equation (1), and any nonlinear equation or the like is used. However, if you take into account the fact that the linear least squares method can be used, and that the calculation can be done with only simple addition and multiplication, a polynomial with a coefficient like equation (1) as a parameter Is preferably used. In addition, when such a polynomial is used, in order to obtain a smooth calibration image without impairing the spatial resolution and density resolution of the original image, a cubic or quartic expression having an intercept or an intercept can be used. It is appropriate to use a cubic or quartic equation.
[0023]
Here, in the calculation processing of the calibration curve described above, when calculation is performed using all of the data (count) collected by the original image photographing unit 1, the size of the image is increased and the sample per channel When the number increases, a lot of calculation time is required to obtain a calibration curve. In addition, as shown in FIG. 6B, when the original image sample point value Qc and the reference image sample point value Qs are concentrated in a place where the count is large, the curve fitting is a portion where the count is large. In the portion where the count of (Qc, Qs) is small, an error such as an error that the curve and the sample value do not match may occur.
[0024]
An example of a method for eliminating such a point will be described below with reference to FIGS. 5 and 6A.
First, as shown in FIG. 5, the sample points are classified into several groups G1, G2,. Next, representative values (Qc1, Qs1), (Qc2, Qs2),..., (Qcn, Qsn) are obtained for each of the groups G1, G2,. As the representative value, for example, an average value or an intermediate value of sample values in the group is adopted.
[0025]
Next, as shown in FIG. 6A, the representative values (Qc1, Qs1), (Qc2, Qs2),..., (Qcn, Qsn) are correlated, and the correlation is expressed as follows. After fitting to the curve [a + b · Qc + c · Qc ² + d · Qc ³ (2)] to obtain sensor characteristic parameters a, b, c and d, the equation (2) By substituting the value of the original image data into Qc (count) of each term, the original image data of the sensor of one channel (x-ch) is calibrated, and these processes are performed for the sensors of all channels. To obtain a calibration image.
[0026]
Since the amount of calculation is reduced by performing the sample grouping process as described above, the calculation time can be shortened even if the image size is increased.
[0027]
In addition, when fitting a curve, sample points are not concentrated in a specific area, so that an error as shown in FIG. 6B can be avoided.
For example, when an edge-shaped object was photographed on a uniform subject and the original image was calibrated using all of the collected data (original image data), artifacts were found at the edge tip. However, such an error is also eliminated by the above-described sample grouping process.
[0028]
FIG. 7 is a block diagram showing the configuration of still another embodiment of the present invention.
This embodiment differs from the previous embodiment shown in FIG. 1 in that a reference image sensor 11 for photographing a subject is provided, and image data collected by scanning of the sensor 11 is used as a reference image creation unit. 12 to obtain a smooth reference image.
[0029]
In this embodiment, only the temporal stability is important for the reference image sensor 11 and the spatial resolution and the density resolution need not be very precise. For example, a semiconductor (Si Etc.) or a combination of a scintillator and a photodiode may be used.
[0030]
Further, as a method of collecting reference image data in this embodiment, for example, as shown in FIG. 8, a reference image sensor 11 is arranged at the rear of the scanning direction of the original image capturing array sensor A, The reference image may be taken at the same time as the original image.
[0031]
【The invention's effect】
As described above, according to the present invention, a reference image that is smoother than the original image is created with an original image obtained from the raw data collected by scanning the array sensor, and the reference image Since the original image is calibrated by correlating the corresponding data of the original image and obtaining the characteristic curve of each sensor, photographing (such as phantom photographing) for obtaining the characteristic of the sensor is performed. A smooth calibration image free of artifacts can be obtained without having to run each time.
[Brief description of the drawings]
FIG. 1 is a block diagram showing the configuration of an embodiment of the present invention. FIG. 2 is a conceptual diagram showing the overall configuration of a radiation imaging apparatus to which the present invention is applied. FIG. 3 is an explanatory diagram of calibration processing in the present invention. FIG. 5 is an explanatory diagram of another embodiment of the present invention. FIG. 6 is an explanatory diagram of another embodiment of the present invention. FIG. 7 is a configuration of still another embodiment of the present invention. FIG. 8 is a block diagram showing a configuration example of a reference image sensor used in the embodiment of FIG. 7.
DESCRIPTION OF SYMBOLS 1 Original image photographing part 2 Reference image creation part 3 Characteristic parameter calculation part 4 Calibration processing part 5 Storage part 10 Radiation source A Array sensor S Radiation sensor T Subject

Claims (1)

  An apparatus that collects the output of each sensor while moving an array sensor in which a plurality of radiation sensors are arranged stepwise on the imaging region, and obtains a radiation image based on the collected data. Based on the collected data or the data collected by the reference image sensor, the original image obtained from the raw data collected by the scanning of the array sensor and the channel have a corresponding relationship, and are smoother than the original image. An image creation unit for creating a simple reference image, and correlating data of channels corresponding to each other between the reference image obtained by the image creation unit and the original image, and obtaining a curve approximated to the correlation A calculation unit that obtains the characteristic curve of the sensor and a calibration processing unit that calibrates the original image data using the characteristic curve. The radiation imaging apparatus according to claim.

Images