CA1205217A - Device for and method of x-ray image processing - Google Patents

Abstract

ABSTRACT:

A device for and a method of X-ray image pro-cessing in which a constant adjustable signal value is subtracted from a video signal of an X-ray image in order to correct the video signal for the effects of inherent scatter. The adjustable signal value depends on the X-ray tube voltage, the nature of the object to be examined, the field size etc. so that it is preferably dependent on the X-ray exposure parameters. The video signal is preferably converted into a logarithmic value by means of a logarith-mic table which is dependent on the adjustable signal value (or on the actual exposure parameters).

Description

PHN `lQ.380 l 9-2_1983 "~evice for and method of X-ray image processing".

The inven-tton relates to a method of processing X-ray images of an object in which -the object is irradiated by X-rays, the radiation after passing through the object being converted into an electric video signal whose loga-rithmic value is forrned for display on a display device.
The invention also relates to a device for processing X-ray images of an object, comprising an image-forming device for generating a video signal, a converter circuit for forming a logarithmic value of the video signal, and 10 a display device for displaying the image converted into logarithmic values.
Such a device and method are known from "Optical Engineering", ~X~ No. 6, 1978, Nov./Dec. pages 652-657.
It is known that X-rays scattered by the obiect (scatter) have an adverse effect on the quality of` an X-ray image.
Notably in cases where quantitative calculations are made on the basis o~ X-ray images (densitometryg functional imaging, measurement of blood flow), the effect of scatter is to reduce the accuracy of the calcula-tions. It has also i 20 been found that the quality of a differential image ob-tained by X-ray image sllbtraction as described in said pwblication, is also adversely affected by the occurrence ~f scatter.
It is the objec~ of the invention to provide a 25 method and a device which produce X-ray images or diffe-rential X~ray images in which the effec-t of scat-ter is sub-s-tantially reduced.
To this end, a method in accordance with the invention is characterized in that the logarithmic value 30 -is :~orrned of -the video signal reduced by a constant, ad-jus-tab:Le value. ~ device in accordance with the inven-tion is characterized in -that -the device comprises arithrnetic means for forrning the logaritllmic value of` the video signal ~2~52~'7 PHN 10.380 2 9-2-1983 reduced by a constant adjustable signal value. In accor-dance with -the invention, a constant, adjustable signal value is subtracted frorn the video signal before the lo-garithmic conversion thereof is performed. ~onsequently, tha-t part of the video signal which is produced by scatter will be decreased, if not substantially reduced with re-spect -to the remaining video signal. The magnitude of the constant signal value to be provided depends inter alia on the hardness of the X-rays (KVp), the size of` the irra-10 diated field~ properties of the object to be examined(thickness, composition), the distance between the patient and the image-forming device, sca-ttered-radiation grid e-tc.
To achieve -this, a preferred version of a method in accordance with the invention is characterized in that 15 for the formation of X-ray images there is provided a pro-gram selection arrangesnen-t in which enables a simultaneous selection and adjustment to be made of several exposure parameters which are dependent on an objec-t to be imaged, the constant, adjustable value being dependent on the ex-20 posure parameters. When such a method is used, -the constant signal value to be subtracted can be chosen optimally for each type of examination, so that that part o~ the video signal which is produced by scattered radiation can be substantially completely eliminated, while the actual video 25 signal which represents the shadow image of -the object will not be a~ected.
Therefore, a preferred embodiment of a device in accordance wi-th the invention is characterized in that the device comprises selection and adjustment means f`or the 30 simultaneous selection and adjustment of a group of object-dependent exposure parameters~ one of the parame-ters being the adjustable signa~ value.
It has been found that the preferred embodiment of -the devlce preferably cornprises at leas-t one video 35 memory for the s-torage o~ an image which llas been forlrled frolrl at :Least one X-ray image, the ou-tput of -the memory being connected to an input of -the arithmetic means which cornpr:ises an adjustable logari-thrnic converter, and is - ~Z85~2:~7 PHN 10.3SO 3 S-2-1983 characterized in that the logarithmic converter is a read/write memory in which there ls stored a logarithrnic conversion table, the outpu~ of the video memory being connected to the address input of the read/write memory.
In a device of this kind it is possible to store, for example, depending on the adjusted exposure parameters, a suitable adapted logarithmic conversion table in the read/write memory. A further advantage is that the video signals are s-tored withou-t alteration (including thc part l0 which is due to scattered radiation)in the video memory~
so that a visual check of the correction for the effects of scatter is possible, so that overcorrection (the sub-traction of too great a signal value, so tha-t -the relevant video signal is cut off below the black leveL) becomes 15 visible and correctable.
The invention will be described in detail here-inafter with reference to embodiments which are shown in the drawing; therein:
Fig. 1a and b show a device in accordance with 20 the invention;
Fig. 2a, b, c and d show a theoretical object, the effect of the object thickness on the video signal, a differential image signal, and the effect of scatter on the video signa;L;
Fig. 3 shows the characteristic of a logarithmic converter in accordance wi-th the sta-te of the art, and charac-teristics of converters in accordance wi-th the in-vention;
Fig. 4 shows a preferred embodiment o~ a device 30 in accordance with the invention, and Fig. 5 shows a modi~ication of -the device shown in Fig. 4.
The X-ray examination device for genera-ting diffe-ren-tially enhanced X-ray images which is shown in Fig. 1 35 cotnprises a high-voltage source G for powering an X-ray tube B, an image intens:ifier II, a pick-up tube PU, a diff`erential amplifier OA, an analog--to-digital ~onverter ~DC1, ancl two processing devices which are connected to ~2~7 P~N 10.380 ~l 9-~-1983 logari-thmic converters L1 and L2 and which are connected to the ou-tput of -the analog--to-digital converter ~CD1 via switches S1 and S2. The outpu:ts of the logarithmic conver-ters L1 and L2 which supply the logarithmic value of the video conten-t stored in the memory MM, are connected to arithmetic means, actually a subtraction circuit V1 (Fig.1a), an outpu-t of which is connected, via a digital-to-analog converter DAC1 and an adder circuit Al, to a disp~ay device MON.
The video signal which is digitized by the con-verter ADC1 is applied, via swi-tch S1 or S2, to one of the processing devices RF1 or RF2; Fig. 1b snows a block dia-gram RF thereof. The processing device RF comprises a re-cursive filtsr consisting of a sub-traction circuit V, a l5 multiplier M, an adder circuit A, and a memory MM. Using -the subtraction circuit V1, a subtraction image is formed from the images stored in the memories MM of the two pro-cessing circuits, said differen-tial image being displayed on the monitor, via the converter DAC1, af-ter the addition 20 of a so-called grey pedestal C to the differential image by means of the adder circuit A1.
- The processing circuit RF (Fig. 1b) comprises a recursive :f`ilter which receives tlle digitized video image.
From the images presen-ted, the recursive filter provides 25 a composite image which consists of a weigh-ted sum of the X-ray images presented in accordance wi-th the f~rmula:
u(i~1) K thVi(i) + (1-K) . V~l(i), i(i) is the i image presentedj VU(i) is the i cornposite image stored in the 30 memory MM~
VU(i~1) is -tile (i~1) composi-te image stored in the rnemory MM, and K is a weigh-ting fac-tor (O-~ K .1).
In order -to ob-tain -thls resul-t, -the i compcsite 35 image is read f`rom the memory M~l and applied -to the sub-tractiorI circult V wh:ich also receives -the i h ~ideo image.
The dif:ference :is muLtiplied by a factor Ii by a rnul-tipl:ier M, after wlIich i-t is added to -tZle :i h composite image in - ~L2a~52:~LPII~
PHN 1o.380 5 9-2 1983 an adder circuit A which is connected to -the ou-tputs of the multiplier circuit M and -the memory ~IM for this purpose.
For correct (weighted) summing of the video images, video synchronisa-tlon pulses SYNC are applied to an address coun-ter AT (Fig. 1a) which, consequent]y, always assigns -the same pixel of the video image to the same address in the memory M~ (for both processing circuits RF1 and RF2). It is to be noted that for the sake of simplici-ty a simple address counter AT is described in order to illustrate the 10 operation of the processing circuit RF. Ho~ever, it will be apparent that some time is required for recalling video informa-tion from an address in the memory M~ and for pro-cessing this information and the video information from a next X-ray image to be added thereto (by -the circui-t ~, ;15 M and A). Therefore so-called pipe line processing -techni-ques have to be used for the processing circuit RF.
In order to generate a differential image, for example, first a short series of X-ray images (for example, ~four) is applied, via S1, to the processing circuit RF1 ;20 which forms a composite image (mask without contrast medium) from the short series. Subsequent:Ly, S1 is closed and S2 is opened. From each subsequent series of X~ray images (after injection of contrast medi-um in5 for example, the cardiovascular system) a composite image is determined by 25 -the processing circuit RF2. The differential images are formed sequentially from the successive, changing composite images from RF2 and the mask from RF1, so that in principle only the cardiovascular system with contrast medium becomes visible on the moni-tor.
I-t is -to be no-ted that -the sequence of successive X-ray images :torming a shor-t series can also be simply summed (for example, in order to reduce the noise effect in an X ray image, see Op-tical Engineering, 17, No. 6, Nov./Dec. -1978, pages 652_657). For sumrning, the proces-35 Siilg circll:it RF (Fig. -Ib) comprises a swi-tch S so -that~
when the weighting factor I~ has the value 1 and -the swltch S is opened, -the X-ray images are surnmed.
In order -to reduce the effect of` scatter on the

2~7 PHN 10.380 6 9-2-1983 images stored in -the memories ~IM of the processing cir-cuits RF1 and/or RF2, an adjustable constant signal value XS is applied to an input of the differential amplifier OA, said signal value XS being subtracted from the v~ideo sig-nal. The adjustable signal value X~ is determined ~in amanner yet to be described) on the basis of -the exposure parameters used -to form an X-ray image, to be processed.
Even though the descri'bed device 10 and the de-vices yet to be described hereinafter are intended for 10 generating differentially enhanced images, the invention ca~ also be used per se in X-ray diagnos-tic devices in which an X-ray image is displayed (and/or electronically stored) by means of electronic means (image intensifier pick-up tube and TV monitor).
Fig. 2a shows a theoretical object O which is irradiated by X-rays in the direction of the arrow X. One half of the object O has a thickness of 15 cm. (tissue, ~ no bones) and the other half has a thickness of 20 cm.
I (tissue, no bones). Each half contains the same blood ves-20 sel 1, 2, respectively. The X-ray image provides, as a function of the location x, a video signal which is shown ~ in Fig. 2b. The amplitude I1 of the video signal is assumed ;` to ~e 100 "in the region of" -the tissue layer having a thickness of 15 crn. The amplitude I1 in the other half 25 then a~ounts to, for example, 33. When -the blood vessels 1 and 2 are filled with a (diluted) contrast medium, the amplitude I1 exhi'bi-ts negative (small) peaks P1 and P2 in the region o~ the blood vessels 'I and 2. If no con-trast medium were present in the blood vessels, -the negative 30 peaks P1 and P2 ~ould be absent. The ampli-tuda I1 in the region of the peaks P1 and P2 amounts to 97 ancl 32, respec-tively (0.97 x 100 and 0.97 x 33, respectively). The diffe-rence ampli-tude I2 (-the dif`ference between the video sig-naLs obtained with con-tras-t medium and wi-thou-t con-trast 35 m~d:ium, respec-tively) is shown in Fig. 2c. The diff`erence amplitude I2 amoun-ts -to .'~ = 3 for the blood vessel 1 and l\ = 1 for -the 'b:Lood vessel 2, even though tlle blood vessels 1 and 2 are the sarne. This unsatisfactory sltuation can be ~ ~2~æ~
PHN 10.380 7 9-2-1983 eliminated by amplifying the amplitude I1 mcasured in the region of the thick objec-t parts more than the ampli-tude I1 measured in the region of thin object parts.
Consequently, in accordance with the state of the ar-t, the amplitude I1 is amplified in a logarithmic manner (the gain characteristic is denoted by the curve LT1 in Fig. 3, Vi being the in~ut amplitude (Il) and Vu the output ampli--tude ) .
The amplitude diffe:rence I2 shown in Fig. 2c 10 then clearly departs from the difference between the loga-rithms o~ the amplituda I1. The difference in the region of the blood vessel 1 can be expressed in a formule as follows:
log 100 - log 97 = 0.0132 and in the :region of the blood vessel 2 as follows:
log 33 - log 32 = 0.0~32.
An amplifier which operates exactly in accordance with the logarithm in -thaory thus solves said problem, because the same signal intensities are obtained for the blood vessels 2~ 1 and 2.
Should scatter occur during the formation of the X-ray images (which is always the case when X-rays encoun-ter matter), the amplitude will vary as shown diagrammati-cally in Fig. 2d. It is assumed that the ratio of scatte-25 red X-rays and non-scattered9 direct X-rays is, for example, 1 : 3.3. It can be deduced therefrom that for an amplitude I3 of 130, the contribution o~ the direct radiation is 100 and that of scattered radiation is 30. When an object 0 as shown in Fig. 2a is irradiated, an X-ray image is ob-30 tained with an amplitude I3 which may be considered toconsist of a "clean signal" (amplitude I1, Fig. 2b) to which a noise amplitude I has been added. After logarithmic amplification (Fig. 3, curve ~T1) of -the ampli-tude I39 the dirference in the region of the blood vessel 1 produces 35 the value :Log 130 - log 127 = 0.0101 and in the region o:t` -the blood vessel 2 log 63 - log 62 = 0. oo6g .

9b~D52~1~7 PHN 10.380 ~ 9-2-1983 Thus, in splte of an ideal, e~act logarithmic amplifier, the signal value o~ the di~erence is not the same ~or the same two blood vessels (dif~erent by a ~actor 1.~). In spite o~ the comparative:Ly ~avourably selec-ted ratio o~
scatter and non-scattered radiation, an undesirable discre-pancy occurs in the imaging o~ the same par-ts o~ an ob-ject. In practice, said ratio will even be more unfavourable (more scatter).
Said problem is solved in accordance with the 10 invention by the subtraction o~ an adjustable signal value XS whose magnitude is determined by the amount o~ scatter occurring during the formation of an X~ray image. In Fig. la, the signal value XS is applied to the di~eren-tial ampli-~ier OA which also receives the video signal from -the pick-15 up tube PU. Using the curve LT2, Fig. 3 shows how -the conversion o~ the video signal (I3, Fig. 2d) is actually per~ormed. After subtraction o~ the adjustable signal value XS, the "cleaned" video signal (I1, Fig. 2b) is converted into a logarithmic value. The adjustable signal value XS
20 depends inter alia on3 - the X-ray tube vol-~age KVp used ~or an X-ray exposure, - the size of the irradiated ~ield, - the na-ture o~ the object to be examined (bones, tissue), - the distance used between X-ray tube, object and image intensifier, - the scattered radiation grid used between the object and the image intensi~ier In X-ray examination devices o~ering program selection (known as organ~rogrammed X-ray apparatus) in 30 which several objectdependent groups o~ X-ray exposure parame-ters (of the decribed kind) can be simultaneously seLected and adjusted by means o~ a selector, i-t is advan-tageous to provide at the same time an adjustable signal value XS (to be determined experimentally) which is applied 35 -to the di~:~erentia:L ampli~ier OA.
F:Lg. ~I shows a pre:~erred em~odime-nt o~ a device IIO in accordance with the invention in which -the correc-tion i`or -the eL`fect o~ sca-tter :is per~ollled in a further nlanner.

`` ~;2~52~7 PHN 10.380 9 9-2-1983 The video signal (I3) originating frorn the pick-up tube PU is applied in its entirety, after amplification by the amplifier OA (in which no correction is performed by means of the signa] XS i.Il this case), to -the converter ADC1 in order to be processed by the processing circuits RF1 and/or RF2 in the manner described with reference to the Fig. 1a and b. However, in this case a logarithmic characteristic is stored both in the logarithmic converter L1 and in the logarithmic converter L2, said characteristics correspon-l0 ding to the curve LT2 of Fig. 3. Because the signal valueXS must be adjustable, -the converters L1 and L2 are formed by read/write memories, so that the conversion tables s-tored in the memories (for each input signal Vi (address) and output signal ~u (content at the address Vi) can be re-15 written in accordance with the na-ture of the X-ray exposure.
When a group of exposure parameters are selected for a given type of X-ray examination, one of the selector-buttons 42 is depressed on a console 41. Simultaneously with exposure parameters (.I~V, thickness of the patient, 20 distance between the patient and the imaging device, etc.), the adjus-table signal value XS which is stored in a memory (not shown) in the control console 41, together with the other parameters, are recalled. The exposure parameters are also manually adjustable by means of knobs 44 by means 25 of which the X-ray tuhe vol-tage, patient thickness, dis-tance between pa-tient and image intensifier, etc. can be selected. Using selection means, an address can be formed from the values of thc adjusted parameters, said address being used to address the appropriate adjustable signal 30 value XS in a table in a memory. The signa:l. value XS is presented -to a memory control unit l~6 which modifies the contents of the converter memo:ries L1 and L2 on the basis thereof. The controlunit 46, therefore, is to 'be connec-ted to ';he address inpu-ts of -the converter memories Ll and L2 35 via switches Sl~, and to t'he da-ta input/ou-tput of the con--verte:r memories L1 and L2 v:ia a fur-ther connection. De-pending on the signal vallle XS or on the group of exposure parameters, a curve LT (Fig. 3) is completely re~ritten ~2~5;~
PHN 10.380 10 9-2-1983 in the converter memories L1 and L2; it should be noted that the dynamic range of the output signals ~Vu) is in no way restricted (ranging between black level and white level on the display monitor). It is also possible to shift the values stored at the addresses in dependence on the magnitude of the signal value XS by one, two or more ad-dress posit:ions further (the curve LT2 in Fig. 3 is silifted towards LT1 when the signal value XS decreases, and away therefrom when the signal value XS increases). The dynamic lO range of the output signal Vu (Fig. 3) might, as a result, be sligh-tly res-tricted in the "white" level.
It has l~een found that the curve LT2 preferably does no-t have an abrupt "beginning" as shown in Fig. 3, but rather a "s-tarting phase" as denoted by a broken line 15 LT3 in Figo 3. In the case of such a curve LT3, any cut-off of the black level of the video signal (which occurs when too large an adjustable signal value XS has been chosen in relation to the actual amount of scatter present) will be less abrupt.
Fig.-5 shows a modification 50 of the device ~0 shown in Fig. L~. The signal value XS determined on -the control console 41 by means of the selectors L~2 or adjust-ment knobs 44, is presented directl~J to -the subtraction circuit AA1 and AA2 which also receive the content o:~ the 25 memories MM of the processing circuits RF1 and RF2. The difference between the signal value XS and the con-tent of the memories MM is presented to the converters L1 and L2, respectively. The converters L1 and L2 may now be read-only memories (ROM) in which the curve LT1 (Fig. 3) or a slight 30 modification thereof (in the sense of the curve LT3) is permanen-tiy s-tored.
I-t is -to be noted that in the de-vice 10 (Fig. la) -there is a risk that some video informa-tion in the blacl~
level will be removed from the video signal by subtrac-ting 35 the signa:L vaLIle XS :~rorn the video signaL. In -the p:referred embodirnen-t of -the device llO (Fig. ~) and tlle modlfication 50 (Fig 5) thereof, this phenomenon is avoide~, because tlle processin~ dev:ices RFI and RF2 s-tore -tlle entire video , 2~d~
PHN 10.38~ 11 9-2-1983 signal (or weighted sums thereof) and the correction for scatter is performed only at a later stage, followed by the display, so that adaption is possible.

Claims (14)

THE EMBODIMENTS OF THE INVENTION IN WHICH AN EXCLUSIVE
PROPERTY OR PRIVILEGE IS CLAIMED ARE DEFINED AS FOLLOWS:

1. A method of processing X-ray images of an object in which the object is irradiated by X-rays, the radiation after passing through the object being converted into an electric video signal whose logarithmic value is formed for display on a display device, characterized in that the logarithmic value is formed of the video signal reduced by a constant, adjustable value.

2. A method as claimed in Claim 1, characterized in that for forming the logarithmic value, a value is re-trieved from a logarithmic table on the basis of the video signal, an adjustable conversion characteristic being stored in said table.

3. A method as claimed in Claim 1, characterized in that the constant, adjustable value is subtracted from the video signal before forming the logarithmic value thereof.

4. A method as claimed in Claim 1, 2 or 3, charac-terized in that for the formation of X-ray images there is provided a program selection arrangement which enables a simultaneous selection and adjustment to be made of several exposure parameters which are dependent on an object to be displayed, the constant, adjustable value being dependent on the exposure parameters.

5. A device for processing x-ray images of an object, comprising an image-forming device for generating a video signal, a converter circuit for forming a logarith-mic value of the video signal, and a display device for displaying the image converted into logarithmic values, characterized in that the device comprises arithmetic means for forming the logarithmic value of the video sig-nal reduced by a constant, adjustable signal value.

6. A device as claimed in Claim 5, characterized in that the device comprises selection and adjustment means for the simultaneous selection and adjustment of a group of object dependent exposure parameters, one of the para-meters being the adjustable signal value.

7. A device as claimed in Claim 6, characterized in that the device also comprises selection means for select-ing the adjustable signal value on the basis of the selected exposure parameters.

8. A device as claimed in Claim 5, 6 or 7, charac-terized in that the arithmetic means comprises at least one subtraction circuit which is connected to an output of the image-forming device and which receives the constant, adjustable value via a further input in order to form a corrected video signal which is applied to the converter circuit.

9. A device as claimed in Claim 5, 6 or 7, charac-terized in that the device comprises at least one video memory for the storage of an image which has been formed from at least one X-ray image, the output of the memory being connected to an input of the arithmetic means which comprise an adjustable logarithmic converter.

10. A device as claimed in Claim 5, 6 or 7, charac-terized in that the device comprises at least one video memory for the storage of an image which has been formed from at least one X-ray image, the output of the memory being connected to an input of the arithmetic means which comprise an adjustable logarithmic converter and wherein the logarithmic converter is a read/write memory in which a logarithmic conversion table is stored, the output of the video memory being connected to the address input of the read/write memory.

11. A device as claimed in Claim 5, 6 or 7, charac-terized in that the device comprises at least one video memory for the storage of an image which has been formed from at least one X-ray image, the output of the memory being connected to an input of the arithmetic means which comprise an adjustable logarithmic converter and wherein the logarithmic converter comprises a subtraction circuit and a read-only memory, the output of the video memory being connected to a first input of the subtraction cir-cuit, a second input of which receives the adjustable signal value, an output thereof being connected to the address input of the read-only memory in which a logarith-mic table is stored.

12. A device as claimed in Claim 5, 6 or 7, charac-terized in that the device comprises at least one video memory for the storage of an image which has been formed from at least one X-ray image, the output of the memory being connected to an input of the arithmetic means which comprise an adjustable logarithmic converter and the device comprises two video memories, each of which is connected to a logarithmic converter, and also comprises further arithmetic means to which the outputs of the log-arithmic converters are connected in order to form a differential image from the converted images stored in the video memories.

13. A device as claimed in Claim 5 r 6 or 7, charac-terized in that the device comprises at least one video memory for the storage of an image which has been formed from at least one X-ray image, the output of the memory being connected to an input of the arithmetic means which comprise an adjustable logarithmic converter and wherein the logarithmic converter is a read/write memory in which a logarithmic conversion table is stored, the output of the video memory being connected to the address input of the read/write memory and the device comprises two video memories, each of which is connected to a logarithmic con-verter, and also comprises further arithmetic means to which the outputs of the logarithmic converters are con-nected in order to form a differential image from the converted images stored in the video memories.

14. A device as claimed in Claim 5, 6 or 7, charac-terized in that the device comprises at least one video memory for the storage of an image which has been formed from at least one X-ray image, the output of the memory being connected to an input of the arithmetic means which comprises an adjustable logarithmic converter and wherein the logarithmic converter comprises a subtraction circuit and a read-only memory, the output of the video memory being connected to a first input of the subtraction cir-cuit, a second input of which receives the adjustable signal value, an output thereof being connected to the address input of the read-only memory in which a logar-ithmic table is stored and the device comprises two video memories, each of which is connected to a logarithmic converter, and also comprises further arithmetic means to which the outputs of the logarithmic converters are con-nected in order to form a differential image from the con-verted images stored in the video memories.