JPH0424673B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method for correcting spatial distortion of a scintillation camera.

Generally, scintillation cameras, including what is called an unger camera, have inherent spatial distortion (spatial nonlinearity), and various methods have been proposed to correct this spatial distortion. However, in the conventional correction method, it is necessary to measure the spatial distortion at each position using an orthogonal parallel line pattern or an arrangement pattern of a large number of small holes for each scintillation camera, which is extremely complicated.

The present invention is designed to prevent the spatial distortion of a scintillation camera from forming a hot pattern on the central axis of each photoelectric converter (the image shrinks on the central axis and expands elsewhere) or in a light guide. When a mask is attached to the center of the image, the main distortion is a cold pattern (on the contrary, the image expands on the central axis and shrinks in other parts), and the spatial fundamental frequency is the spacing of the photoelectric converter array. It is an object of the present invention to provide a spatial distortion correction method for a scintillation camera that makes it possible to more easily correct spatial distortion.

An embodiment of the present invention will be described below with reference to the drawings. In Figure 1, scintillator 1
A large number of PMTs (photomultiplier tubes) 3 are arranged on the back of the scintillator 1 via a light guide 2, and the light emitted from the scintillator 1 due to radiation incidence reaches each PMT 3 through the light guide 2, and the output current of each PMT 3 increases. After being converted into a voltage by each preamplifier 4, it is input to the arithmetic processing unit 5, where it is subjected to normal position calculations and normalization calculations using a resistance matrix, wave height analysis of the energy signal, etc. Luminance signal which is signal Z and pulse height analysis output
UNBLANK is output.

In the configuration of the detection section of such a normal underwater camera, the first step is to use, for example, a large number of
The output of the preamplifier 4 of the PMT 3 located at the center of the array of PMTs 3 is taken out through the integrating circuit 6 and the amplifier circuit 7, and the response of the output wave height V depending on the light emission position is measured using a multi-channel analyzer or the like. That is, for example, a point-like collimated radiation source 8 is placed on the surface side of the scintillator 1 at the center.
When the output wave height V is determined by changing the distance r from the central axis O of the PMT 3, response data as shown in FIG. 2, for example, is obtained. Thus one PMT
Response data for 3 is obtained, but other
Since the response data of PMT3 can be considered to be similar to this, the position Y is approximately determined, and the difference between this position and the incident position, that is, the distortion vector, is determined. Therefore, a distortion correction vector, which is the inverse vector thereof, is obtained as an initial value for the positions X and Y after each calculation.

Next, as a second step, a flat image (uniformly irradiated image) is collected using the above configuration, and the data is stored in a data storage device (not shown). that time,
In order to reduce the influence of sensitivity nonuniformity caused by variations in the energy signal Z depending on the position, the window width for wave height analysis is widened, or when a second wave height analysis process is performed by a computer. It is preferable to take measures such as correcting the relative relationship between the energy signal and the window level of the second wave height analysis depending on each position. When the data of the stored flat image is corrected (or coordinate transformed) according to the initial values of the correction vectors at each position, each event density after correction is statistically It should be uniform within a certain margin of error. Therefore, as described above, a corrected flat image is obtained using the initial value of the correction vector obtained first, the density gradient for each position is obtained, and the correction amount of the corresponding correction vector is obtained according to the density gradient. This corrects the initial value of the correction vector. Next, the flat image is corrected again using the correction vector modified in this way, and the amount of correction of the correction vector is obtained again by the same procedure as described above. Such iterative correction is repeated until a suitable degree of convergence is reached for an appropriate area unit.

In fact, instead of calculating the density after calculating the flat image corrected by the correction vector as described above, the count value of the uncorrected flat image within the area of each matrix element is calculated as a rectangle surrounding the matrix. The density of each position after correction can be obtained by dividing by the area formed by the correction vector of each vertex of , that is, the area of the matrix element after correction.

The correction vectors obtained in the first and second steps are stored for each coordinate position, and the spatial distortion is corrected, for example, as shown in FIG. In FIG. 3, X and Y (see FIG. 1) obtained for each scintillation event are passed through sample and hold circuits 91 and 92 to an AD converter 101,
The address of the correction coefficient memory 11 is specified using each of the obtained digital position signals. In this correction coefficient memory 11, the above-mentioned correction vectors are stored for each address related to each coordinate position, and the corresponding correction vector is read out by specifying the address as described above. Each read correction vector is input to DA converters 121 and 122, and analog correction amounts ΔX and ΔY are obtained. This correction amount ΔX,
ΔY is added to the outputs Xa and Ya of the sample and hold circuits 91 and 92 in addition circuits 131 and 132, respectively, to obtain corrected position signals Xc and Yc. The luminance signal UNBLANK is given on the one hand as a timing signal for signal acquisition and holding of the sample and hold circuits 91 and 92, and on the other hand is input to the unblank control circuit 14, and is used as a conversion start signal and a corrected position of the AD converters 101 and 102. Generates a luminance signal UNBLANKC with timing matched to signals Xc and Yc.

In addition, in the above example, the
Although we used response data of PMTs, we also used response data of PMTs arranged at other positions, or multiple PMTs such as a central PMT and a peripheral PMT.
It is also possible to combine response data. Furthermore, response data may be measured using a device other than a scintillation camera that has the same optical system. Furthermore, in addition to correcting with the configuration shown in Figure 3 above, for example, the accuracy of the correction vector can be further increased through interpolation, and local non-uniformity can be prevented by adding random number data. Correction is also possible. Further, instead of adding the correction amount to the position signal in analog form, it is also possible to add it digitally.

As described above with respect to the embodiments, according to the present invention, spatial distortion can be corrected, and therefore sensitivity non-uniformity that occurs due to this can be improved. In particular, when manufacturing a large number of scintillation cameras with the same optical system, the initial value of the correction vector is determined only for the first camera in the first step, and the initial value of the correction vector for the second and subsequent cameras is determined in the same step as for the first camera. Since it is only necessary to carry out the two-step iterative correction work, the correction work can be further simplified. Therefore, it is not necessary to measure the spatial distortion at each position using an orthogonal parallel line pattern or an arrangement pattern of a large number of small holes for each scintillation camera as in the conventional method.

[Brief explanation of drawings]

Figures 1 and 3 are block diagrams for explaining a correction method according to an embodiment of the present invention, and Figure 2 is a PMT.
7 is a graph showing response data of light emission position dependence of output. 1...Scintillator, 2...Light guide, 3
... PMT, 4 ... Preamplifier, 5 ... Arithmetic processing section, 6 ... Integrating circuit, 7 ... Amplifying circuit, 8 ... Radiation source, 91, 92 ... Sample hold circuit, 10
1,102...AD converter, 11...Correction coefficient memory, 121,122...DA converter, 131,
132...addition circuit, 14...unblank control circuit.

Claims (1)

[Claims] 1 Measure the response data of the emission position dependence of the output of a specific photoelectric converter in a scintillation camera, and calculate the spatial distortion at each coordinate position by simulation calculation using the response data and data regarding the array of photoelectric converters. The first step is to determine the initial value of the correction vector at each position so as to correct this distortion, and the data of the density gradient at each position after collecting a flat image and correcting the spatial distortion using the initial value of the correction vector. and a second step of iteratively correcting the correction vector from Spatial distortion correction method for scintillation cameras.