GB2043391A - Electronic device for compensating non-linearity in a gamma camera output - Google Patents

Abstract

In a gamma scintillation camera 10, the output signals Ux, Uy represent non-linear functions of the geometrical coordinates x, y of the location of a scintillation event in the crystal. The outputs Ux, Uy of the camera 10 are fed respectively to one input of corresponding high gain differential amplifiers 11 and 12 whose outputs V and W are applied to corresponding inputs of a quadripole non-linear device 13 (Figure 4 not shown) as orthogonal deflection signals for a cathode ray tube. The location of a spot deflected thereby is sensed by an assembly of photodetectors 23 having the same distribution relative to the cathode ray tube screen as the photomultipliers of the camera 10 relative to the scintillation crystal. A weighting circuit 24 generates output signals Uv and Uw representing the barycentre of the photodetector outputs, which are fed to the respective other inputs of the differential amplifiers 11 and 12, thus completing a servo feedback loop. When the servo loop reaches an equilibrium state, the signals V and W provide corrected output coordinate values linearly related to x and y. <IMAGE>

Description

SPECIFICATION Electronic device for compensating non-linearity in a gamma camera output The invention relates to an electronic device for compensating errors due to non-linearity, particu marly for compensating linearity errors in a gamma camera.

Nuclear medicine senses and measures the assimilation of small quantities of radioisotopes by healthy and diseased tissues. The nuclear analysis devices customarily used for detecting the emitted gamma rays, after administration of said radioisotopes to a patient, and for revealing the differences in assimilation of said radioisotopes in healthy or diseased tissues, are gamma cameras, notably gamma cameras of the Anger type described in United States Patent Specification No.

3,011,057.

The principle of operation of these gamma cameras is known and several variants thereof are described in numerous documents, such as United States Patent Specification Nos. 3,683,185 and 3,919,566, French Patent Application No.75.32.121 in the name of N.V. Philips' Gloeilampenfabrieken, published under No.2,288,987 on May 21st, 1976, and particularly French Patent Application No.

76,28.075 in the name of "ELSCINT LTD", published on April 15th, 1977, under No.2,325,053.

The latter document gives a detailed description of the construction and operation of a gamma camera, in this case of the Anger type, but also focuses attention on a fundamental problem of such a camera, that is to say the geometrical non-linearity thereof, which causes scintillation positions to be displayed which are nearer to the optical axis of the photomultiplier tubes (which serves to convert the emitted light into electric signals) than the positions in which scintillations actually take place.

For the reduction or removal of said non-linearity, numerous solutions have been proposed which are dependent on one or other of the following two principles: suppression ofthe non-linearity in an optical manner (see the French Patent Applications published under the Nos. 2,219,423 and 2,325,053) by changing the distribution of the light produced by the scintillator and transferred to the photomultiplier tubes, or electronic suppression of the non-linearity by a variation of the output signals of the camera (see French Patent Application No.2,219,424, French Patent of Addition No.2,243,450, and also United States Patent Specification No. 3,980,886).Subject to a distinction which will be more specifically described hereinafter, the invention described herein relates to the second category and realizes an electronic correction of said linearity errors. A comparative study of the three latter documents clearly reveals a common characteristic of the various methods or devices proposed thus far for the elimination or reduction of said linearity errors: these solutions always imply a modification of the camera via the interposition of the correction element or the correction elements in the processing network for the signals detected by said camera, usually at the expense of the geometrical resolution.

The invention has for an object, however, to provide an electronic correction device which does not involve modification of either the camera signal processing networkorthe internal optical assembly within the camera, but only a correction of the electrical output thereof. The linearity errors in the output signals of the camera are not suppressed but are compensated for.

According to the invention there is provided an electronic device for compensating errors due to non-linearity occurring at the output of a gamma camera in the two output coordinate signals (Ux, Uy) which correspond, on an output display surface corresponding to the detection field of the camera, for example on the output screen of the camera, to the position coordinates (x, y) of scintillations associated with the gamma rays detected by the camera characterized in that the device comprises an electric quadripole with two inputs receiving signals (v, w) and two outputs supplying feedback signals (Uv, Uw), the transfer function thereof being substantially the same as the corresponding transfer function of the gamma camera, relating the position coordinates (x, y) to the output coordinates (Ux, Uy) and two differential amplifiers, the first differential amplifier receiving a first coordinate signal (Ux) on a first input with a given polarity and a first feedback signal (Ux) on a second input with an opposite polarity, said first differential amplifier supplying at its output a first compensated input signal v for the quadripole, the second differential amplifier receiving on a first input with said given polarity the second coordinate signal (Uy) and on its second input of opposite polarity a second feedback signal (Uw), the output of said second differential amplifier supplying a second compensated input signal w for the quadripole, the arrangement being such that the signals v and w substantially linearly represent the corresponding coordinatex, y under equilibrium conditions.

In the device in accordance with the invention, the camera has associated with it a quadripole which simulates the operation of the camera and which, as a result of its transfer function which is substantially identical to the transfer function of the camera, enables compensation of the linearity errors in conjunction with a differential amplifier stage connected in series with the camera. Moreover, said compensation has no effect on the geometrical resolution of the gamma camera. The characteristic of the invention leads to the said distinction hereinbefore referred to in that a device in accordance with the invention enables errors in linearity to be corrected electronically, but does not apply said electronic correction to the camera circuits themselves in contrast to devices used hitherto which have always involved modifications within the camera itself.

An embodiment of the invention will now be described by way of example, with reference to the accompanying diagrammatic drawing; of which: Figure 1 shows the essential elements of a gamma camera, Figure 2a and 2b are diagrams indicating the form of linearity errors to be compensated, Figure 3 illustrates a gamma camera including a devise fc ,ompensating errors due to non-linearity in accordance with the invention, and Figure 4 is a more detailed view of a transfer function simulation quadripole which can be used in the device in accordance with the invention shown in Figure 3.

A gamma camera serves to deliver information concerning intensity and position of gamma radiation occurring, and supplies amplitude information and position information; to achieve this, a gamma camera as shown in Figure 1, comprises a scintillation crystal 1 in which scintillations occur at the points of interaction between gamma rays and the crystal, a light conductor 2 which constitutes an optical coupling member between the crystal 1 and a distributed assembly 3 ofn photomultiplier tubes for detecting scintillation light and for generating corresponding electrical signals, and a circuit 4 for processing these signals.

Figures 2a and 2b illustrate the errors due to nonlinearity which have an adverse effect on the performance of a gamma camera. Thus, a scintillation with input position coordinatesx, y as shown in Figure 2a, a given output point will give rise to output voltages Ux and Ux, such that a non-linear relationship will in general existbetweenx, y and Ux, Uy, respectively. Figure 2b illustrates the image deformation of the raster grid shown in Figure 2a, which is caused by the presence of said errors due to nonlinearity.Errors of this kind cause the appearance of "concentration" zones in the image obtained at the output of the camera which are symmetrically situated with respect to the respective optical axes of the corresponding photomultipliertubes. The presence of these concentration zones can be demonstrated by a uniform irradiation of the scintillation crystal.

The image thus produced will not be uniform, but will exhibit concentrations of detected scintillations around centres which correspond to the locations of the optical axes ofthe respective photomultiplier tubes with respect to the scintillation crystal.

An embodimenfof the compensation device in accordance with the invention for compensating these errors, is shown in Figure 3; it comprises two differential amplfiers 11, 12 having an amplification coefficient K which are connected to respective outputs of a gamma camera 10, and a quadripole 13, the transfer function of which is substantially identical to that of the gamma camera 10. It is to be noted that it cannot in general be completely identical in practice, considering that the detection of scintillation events by the gamma camera 10 are of a random nature, and the output signals Ux and Ux which correspond to the location of said events will be adversely affected by drift and other fluctuations which set a limit to the geometrical resolution of the gamma camera.However, corresponding fluctuations will not in general occur simultaneously in the input and output signals of the quadripole 13.

As is shown in Figure 4, the quadripole 13 in the present embodiment comprises a cathode ray tube 21, a light conductor 22 which formr an optical cou pling member between the fluorescant screen of the cathode ray tube 21 and a distributed assembly 23 of light detectors, and a circuit 24 for processing the signals supplied by said assembly 23 in response to the detection of a light signal generated by the impact of an electron beam on the screen of the cathode ray tube 21 to provide corresponding coordinate signals Ux, Uw.The circuit 24 is substantially identical to the corresponding circuit4 employed in the camera 10 for determining the iocation of the scintillations occurring by performing the weighting calculation, for example in the camera by evaluating Ux and Uy using relationships of the type:

said relationships corresponding to the determination of the coordinates Ux, Uy of the barycentre of an assembly of point signal values.

The referencesx, y denote the geometrical coordinates (in the assembly formed by the figures) of a scintillation occurring due to a detected gamma ray (phenomenon coordinates). A relationship exists between the mean value of the output signals Ux and Ux determined by the camera 10 for displaying the coordinatesx, y on a display surface which represents and corresponds to the detection field of the camera (for example, a screen), and the said coordinates x, y said relationship being given by the for mulae: Ux = F (x, y) and Ux = G (x, y). Because the transfer function of the quadripole 13 is substantially identi cal to that of the camera 10, the relationship U-F c-)andUw=G(c-c7) Vcc cc exists between the output signals Ux, I Uw formed by the quadripole, and the input signals v, w.

The ratiosand and - represent the values c c of the coordinates to which the voltage values v and w correspond (the letter c represents a constant).

The non-inverting (positive) input of the first differential amplifier 11 receives the output signal Ux of the camera 10, and its inverting (negative) input receives the output feedback of the quadripole 13.

The input signal vforthe quadripole 13 is the output signal of said differential amplifier 11. The noninverting input of the second differential amplifier 12 receives the output signal Uy of the camera 10 and its inverting input receives the output feedback signal Uw of the quadripole 13. The input signal w for the quadripole is the output signal of the differential amplifier 12.

The electronic device for compensating errors due to non-linearity in accordance with the invention, constitutes a system which continuously searches for a state of equilibrium. The operation will be described hereinafter. It is assumed that the functions F and G, determining the relation between Ux and Uy on the one hand andx andy on the other hand, are monotonous functions, for example, monotonously increasing functions; for a given value of this means that U, is largerthan Uxt ifx2 is larger than x1, and for a givenx value that U, is also largerthan Uy1 if y2 is largerthan y,.

At the instant at which the compensation device is connected to the output of the camera 10, a general state of equilibrium will tend to occur in the system on the one hand because the four said relations as regards Ux, Ux, Ux and Uw have to be satisfied, and on the other hand because the equalities pursued by the feedback relationship set up between the camera 10 and the compensation device have also to be satisfied: namely v = K (Ux-Ux),and w = K (Uy - Uw).

If the value of K, denoting the amplification coefficient of each differential amplifier 11, 12, is large, it can be deduced, because v and ware finite values, thatthe differences (Ux- Uv) and (Uy- Uw) are small.

This means that the values of Ux and Ux will be substantially equal to one another, as also will be the values of Uy and Uw. The unambiguous character of the relationships realized by the transfer functions of the camera 10 and of the quadripole 13 of the compensation device, ensures that the values of the ratio and of the parameterx will also be substantially c equal to one another, as also will be the values of the ratio cWand of the parametery. It will be clear that c these equivalences will become more precise as the amplification coefficient K is increased. Therefore, the following can be written: v = c.x, and w = c.y.

The two input signals v and w of the quadripole 13 will therefore be linearly dependent on the values of the coordinates x andy which it is desired to provide.

The output signals Ux and Ux which contains errors due to non-linearity generated in the camera, do not exhibit this linearity. The said input signals v and w thus form compensated output signals of the assembly formed by the electronic compensation device and the camera 10.

The state of equilibrium of the system formed by the camera 10 and the compensation device, moreover, must be made completely stable. This stability will ensure that a new state of equilibrium will be reached for any scintillation occurring with newx andy coordinates, thus generating new Ux and Uy values.When it is assumed that, starting from a state of equilibrium xO, yO, Uxo, Uyo, the next coordinates occurring will assume the values x1 and y1 and that the corresponding signals will assume the values Uxi, U,, x1 andy1 being greater than x0 and yO, respectively, the value of the difference (Ux - Uv) will become greater, i.e. the value of the ratio vk will became greater while the effect of the transfer function of the quadripole 13 will be to cause the value of Uv correspondingly to increase.The increase in the value of Uv will cause a reduction in the difference (Ux Ux) which would otherwise have tended to increase. The feed back system thus attempts faithfully to follow the variations in the output signals of the gamma camera 10.If it is assumed that starting from a state of equilibrium xO, yO, Uxo, Uyo, Uxo, Uwo, the value of (for example) UVO changes by a positive amount dUvo, the value of the difference (UXO- - UVO) will be reduced and become equal to (UXO - UVO dUvo), i.e. (vJK) - dUxo. If the amplitude of the input signal of the quadripole 13 is decreased, the characteristic of monotony relating to the functions F and G ensures that the amplitude of the output signal Uxo will also decrease, contrary to the initially positive amount dUVo. A state of equilibrium will thus be reached for each group of values xj, y1, Uxi, Uxi, Uvi, Uwi.

Consequently, by the addition of the electronic compensation device in accordance with the invention to the camera, a stable system can be obtained in which the effect of the linearity errors of the camera 10 alone, can be compensated for.

One of the conditions to be satisfied for the effective compensation of the non-linearity of the camera 10 by the system thus formed is, as has already been stated, that the transfer function of the quadripole 13 should approximate to the transfer function of the gamma camera as closely as possible.A suitable approximation (disregarding statistical fluctuations) can be obtained if a) the network 23 formed by the light detectors is arranged so that the centres of the detectors form a geometrical arrangement which corresponds with the arrangement formed by then photomultiplier tubes in the gamma camera, and b) for each of said light detectors the response of the light detectors varies as a function of the distance between the point of incidence of the electron beam on the screen of the tube 21 and the longitudinal axis of a respective detector in a manner which corresponds to the equivalent variation in the response of the photomultiplier tubes as a function of the distance between a scintillation and the axis of the corresponding photomultipliertube.The processing circuits 4 and 24 may be substantially identical, the only difference being that the circuit 24 need not calculate the amplitude of the signal, since the spot brightness of the cathode ray tube 21 is maintained constant and the purpose of the circuit 24 is to compensate for the linearity errors in the detected positions. The light detectors can comprise either photodiodes or photomultiplier tubes.

It has already been stated that the degree of compensation increases as the magnitude of the amplification coefficient K of each differential amplifier 11, 12 is increased, because the approximation ofx by v/C and of y by w/C is made closer. The validity of this operation is enhanced by the fact that the errors due to non-linearity may themselves vary in dependence on the values of the coordinates disturbed by said errors. This means that these errors due to nonlinearity do not exhibit the same amplitude for a phenomenon having coordinatesx, y, detected by the camera 10, and for the signals v and w (approximately equal to Cx and Cy, respectively), which correspond to said coordinates at the input of the quadripole 13 when the system is in the state of equilibrium. The amplitude difference thus determined will be smaller, however, as the differences Ux - Ux and Uy - Uw are made smaller,i.e. as the amplification coefficient K is increased. Tests have demonstrated that, when -re amplification coefficient K has a value of approximately 100, a satisfactory degree of com pensation can be achieved. However, because it is desirable for the feedback system to operate in a stable manner and that it should be correctly damped, a value no higher than 1000 should be chosen for the amplification coefficient K, the value to be chosen being determined on the basis of practical tests.In order to limit the magnitudes of the input signals v and w of the quadripole 13 to a useful operational range, the differential amplifier 11 and 12 can be arranged so that the amplifiers saturate .beyond the limits of said range, which means that said amplifiers have the coefficient value K only within the desired dynamic range of the signals v andw.

Thus in one form of operation of the embodiment shown in Figures 3 and 4, the output signals Ux, Uy of the gamma camera are maintained, in the absence of a scintillation event, at a central value Uxo, Uyo representing the central coordinate x0 = o y0 = o, the absence of a scintillation being indicated by the absence, i.e. by the presence of a value below a minimum threshold, of an amplitude output signal (not shown) provided by the camera 10.Servo feedback, hereinbefore referred to, will cause the spot of the cathode ray tube 21 to be deflected to the centre of the screen by the applied deflection signals v0 = w0 = o which will give rise to corresponding com puted coordinate values UVO = o, UWO = o relating to the barycentre of photosignals produced by the centred light spot on the cathode ray tube screen, from the outputs of the circuit 24.

When a scintillation occurs, the values of Ux and Uy will change to the corresponding new values at a rate determined by the response times of the detectors 3 and of the computing circuits 4. Because of the high gain of the differential amplifiers 11 and 12, these amplifiers could rapidly reach an overload condition and are preferably arranged to limit. The rate of change of the outputs v and wwill depend ultimately on the slew rate of the respective amp lifier. This consideration together with the need to charge or discharge other circuit capacitances, for example the deflection circuit of the cathode ray tube 21, will determine the initial rate at which the cathode ray tube spot will be deflected across the screen towards a new position.As the spot moves across the screen, the outputs from the photocells 23 will change, and via the computing circuits 24 will cause the values Ux and UwtO change. In orderto follow the position of the spot, the response rates of the photocells 23 and of the computing circuits 24, must be sufficiently fast, and the after-glow of the cathode ray tube screen must be short and/or its effect reduced if possible, ag. by a suitable colour filter which discriminates in favour of the initial fluorescence due to electron impact.

Thus one favourable manner of operating the device shown in Figure 4 would be to ensure that the initial rate of deflection of the cathode ray tube spot is so related to the response time of he photocell network 23 and the circuit 24, that undue overshoot is avoided before the feedback control via the new values of Ux and Uw becomes effective at the other inputs of the amplifiers 11 and 12.

The occurrence of a scintillation, signalled by the commencement of an amplitude output signal (not shown) from the camera 10, can be arranged to start a delay period during which the transfer of the compensated outputs v and w to subsequent apparatus is inhibited to allowthese valuesto attain theirfinal stabilised values for the corresponding scintillation event. To ensure that the values Ux, Ux at the camera output, are maintained until these final values are attained and have been read out, the camera 10 can be provided with suitable output hold circuits whose operation and timing is suitably controlled with respect to the occurrance of the corresponding scintillation event.

The invention is not intended to be limited to the embodiment described in this specification; other embodiments can be realized equally well within the scope of the invention as defined by the claims. Notably the value range determinedforthe amplification coefficient K during the adjustment of the device in accordance with the invention is an optimum zone of operation, butthe lower and upper limit can be changed without the system ceasing to operate in accordance with the description given herein.

It will also be clear that the electronic compensation device realized in accordance with the invention can be connected to an arbitrary gamma camera whose output signals contain errors due to nonlinearity, optimalization of the amplification coefficient K and the adaption of the characteristics signal = f (distance) being the only important adjustments to be performed in any specific case.

Claims (6)

1. An electronic device for compensating errors due to non-linearity occurring at the output of a gamma camera in the two output coordinate signals (Ux, Uy) which correspond, on an output display surface corresponding to the detection field of the camera, for example on the output screen of the camera, to the position coordinates (x, y) of scintillations associated with the gamma rays detected by the camera, characterized in that the device comprises an electric quadripole with two inputs receiving signals (v, w) and two outputs supplying feedback signals (Uv, Uw), the transfer function thereof being substantially the same as the corresponding transfer, function of the gamma camera relating the position coordinates (x,y) to the output coordinates (Ux, Uy), and two differential amplifiers, the first differential amplifier receiving a first coordinate signal (Ux) on a first input with a given polarity and a first feedback signal (Uv) on a second input with an opposite polarity, said first differential amplifier supplying at its output a first compensated input signal v for the quadripole, the second differential amplifier receiving on a first input with said given polarity the second coordinate signal (Uy) and on its second input of opposite polarity a second feedback signal (Uw), the output of said second differential amplifier supplying a second compensated input signal w for the quadripole,the arrangement being such that the signals and Wsubstantially linearly represent the corresponding coordinatesx, y, under equilibrium conditions.

2. A device as claimed in Claim 1, characterized in that the quadripole comprises a cathode ray tube the beam deflection of which is controlled by the input signals v and w, a light conductor, a set of light detectors arranged in a distribution relative to the output screen of the cathode ray tube which corresponds to the distribution of photomultipliers relative to the scintillation crystal of the gamma camera whose output is to be corrected, and a circuit for processing the output signals from the light detectors and for supplying the output signals Uv and Uw.

3. A device as claimed in Claim 1 or 2, characterized in that the amplification coefficient of each differential amplifier lies in the range from 100 to 1000.

4. A device as claimed in Claim 1 or 2 or 3, characterized in that the said differential amplifiers are arranged to limit for a predetermined value of the differential input.

5. An electronic device for compensating errors due to non-linearity occurring at the output of a gamma camera, substantially as herein described with reference to the accompanying drawing.

6. A gamma camera, characterized in that an output of the camera is connected to a device for compensating linearity errors as claimed in any one of Claims 1,2,3,4 and 5.