Tomographic Scanner

This application pertains to computerized axial tomographic scanners and more particularly to traverse and rotate type scanners. Traverse and rotate type computerized axial tomographic scanners are well-known in the art, see for example U.S. Pat. No. 3,919,552. Also well-known in the art are the electronics and mathematics for translating the radiation attenuation through the object scanned into a visual representation of the cross section of the object scanned. See for example "The Fourier Reconstruction of a Head Section", Shepp and Logan, IEEE Transaction on Nuclear Science, June, 1974.

In a traditional traverse and rotate system, a beam of radiation and a receiving detector were scanned linearly at a constant speed across a patient in a scan circle during which time attenuation data was taken. After scanning the patient, the radiation source and detector were slowed to a stop in an area outside of the scan circle. The carriage carrying the source and detector were subsequently rotated a few degrees, and accelerated to the desired linear scan speed before reaching the scan circle. A further set of attenuation data was taken as the radiation source/detector traversed the patient in a generally opposite direction but shifted by the few degrees the carriage was rotated. This traverse, then rotate procedure was often repeated on the order of 15 to 180 times to obtain one set of data.

In the past it was commonly believed that the carriage supporting the radiation source and detector system must be traversed at a constant speed. It was believed that nonconstant traverse speeds would produce irregularities in the transmitted attenuation data. As a result only drives for the carriage were used which produced constant speeds when traversing the scan circle. Constant speed drives of the type required are complex and expensive. Because the scan velocity was constant through the scan circle all acceleration and deceleration had to occur outside of the area. The time and space required for realistic changes in speed added to the size and cost of the machine and increased the scan times.

In order to minimize the amount of human organ movement during scans, faster scan times were desired. As a result shorter acceleration and deceleration times were necessary. However, the more sudden were the changes in speed, the greater were the undesirable forces on the scanner parts. Additional wear, vibration and maintenance occured as a partial result of increased forces due to shorter acceleration and deceleration times.

To achieve the linear speed across the patient with rapid slowing and acceleration and deceleration at the extremes of the traverse, nonlinear gear boxes were often used. Ball screw drives were also used to drive the source and detector carriage through the rapidly changing speeds. Both these mechanical gearings were sometimes a source of undesirable vibration in the system.

The present invention contemplates new and improved traverse and rotate scanning apparatus which overcomes all of the above referred problems and provides a scanner which is fast, simple, inexpensive, and produces high resolution.

In accordance with the present invention, a traverse and rotate scanner is provided in which the source of radiation and the detectors are traversed on a carriage at a nonlinear speed. Data samplings are taken at spacially controlled increments rather than time controlled increments.

In accordance with a more limited aspect of the invention, a continuously operated drive mechanism is used which accelerates and decelerates the source and detector carriage over the entire distance traversed.

In accordance with another aspect of the invention, a vibration insulating drive such as a flexible drive is used in producing rotational indexing between traverses.

In accordance with yet another aspect a flexible linkage with two degrees of freedom supports electrical cable connections to electrical components mounted on the carriages.

One advantage of the present system is the smoothness of operation. The source and detector carriage accelerates and decelerates gradually over the entire scan length. This more gradual change decreases the time and distance of traverse, thus increasing the scan speed. Smaller forces accompany the more gradual acceleration and deceleration, thus reducing or eliminating many sources of vibration.

Another advantage of the present invention is that it uses relatively small power sources to create the smaller forces. The use of smaller motors reduces cost, wear and stress.

A further advantage of the present invention is the high degree of resolution. Spacial samplings of the radiation detector may be taken at intervals of one millimeter or less. The sampling positions being spacially determined the sampling rate is spacially dependent and time and velocity independent. According, the Nyquist frequency is independent of detector velocity in the present invention.

A further advantage of the present invention is a reduction in vibrational interference. Vibration is reduced in a number of ways, such as a continuous speed traverse motor, gradual traverse speed changes, and a flexible drive to name a few.

Still another advantage of the present invention is the smooth cable handling operation. The cable handling allows flexible interconnection with electric components mounted on the carriages without cable tangling and vibrational or other interference with the carriage travel.

The invention may take physical form in certain parts and arrangements of parts, a preferred embodiment of which will be discussed in detail in this specification and illustrated in the accompanying drawing.

FIG. 1 is a front view of a traverse and rotate system in accordance with the present invention;

FIG. 2 is a side sectional view of a traverse and rotate system in accordance with the present invention with the radiation source and detector carriage and counter balance carriages centered;

FIG. 3 is a PLAN view of an alternate embodiment of a traverse and rotate system in accordance with the present invention with a gear motor power supply;

FIG. 4 is a front elevational view of another alternate embodiment of a traverse drive system in accordance with the present invention; and

FIG. 5 is a diagrammatic illustration of a signal handling system in accordance with the present invention.

Referring now to the drawings wherein the showings are for the purpose of illustrating the preferred embodiment of the invention only and not for the purposes of limiting same.

FIGS. 1 and 2 show a traverse and rotate scanner built in accordance with the present invention. At the center of the scanner is a stationary cylinder 10 in which the object to be scanned is positioned. In the preferred embodiment, the cylinder 10 would be about ten inches in diameter, or sufficient to encompass the region to be examined, i.e. scan circle. The cylinder may be sized to receive the head of a patient for brain scanning, although other sizes may be used. Supporting and attached to the stationary cylinder 10 are a pair of legs 12 and 14 which mount the cylinder and other apparatus on a large base 16. Adjustable feet, three of which are illustrated at 18, 20 and 22 operatively attached to the base 16, are used for leveling.

Mounted for rotation about the cylinder 10 is a frame or plate 24. The frame 24 is integral with cylinder 26 through which it is rotatably mounted upon cylinder 10. Means 25 for incrementally rotating the frame 24 and cylinder 26 includes a rotational drive comprising a motor 28 connected with a gear box, a sprocket and a flexible drive chain 30. Chain 30 in turn engages cylinder 26 by being pinned to it or, alternately, by a ring 31 of sprocket teeth circumscribing cylinder 26. As drive motor 28 is rotated incrementally, the frame 24 is caused to assume different angular orientations incrementally about the object in cylinder 10. The rotational drive is described in greater detail below. Mounted on the front of frame 24 are a pair of upper and lower bars 32 and 34, upon which a means for supporting a radiation source and radiation detectors is slidably mounted. In the preferred embodiment this supporting means takes the form of a carriage 36. Another pair of bars 76 and 78 are similarly mounted on the back of frame 24 to carry counter balance means such as a carriage assembly 38.

A means 39 for traversing the carriage includes a rotary power source such as a torque motor 40 mounted on a platform on frame 24 and a nonlinear drive means 43. The nonlinear drive means includes a central shaft 41 which in this embodiment is integral with the armature of motor 40 on the opposite ends of which are mounted lever arm means in the form of a pair of lever arms 42 and 44. Between the lever arm 42 and the carriage 36 is an engaging means for slidably coupling the lever arm and carriage for transferring movement from the arm to the carriage. In the preferred embodiment this engaging means includes a wheel or follower 46 mounted on the arm 42 and a track 48 on carriage 36 in which the follower rides. Similarly, there is another engaging means between arm 44 and counter balancing carriage 38. This engaging means includes follower 50 and a track 52 on the counter balance carriage 38 in which the follower rides. It can be seen that as power source 40 rotates, the followers 46 and 50 pull and push the carriages 36 and 38 back and forth along bars 32, 34, and 76, 78 respectively. This illustrated drive assembly is known as a Scotch yoke which it may be noted causes the carriages to oscillate back and forth with substantially simple harmonic motion when driven by a constant speed power source. Other drives which cause other forms of smoothly changing back and forth motion are equally applicable. Mounted on carriage 36 is a source of radiation 54 composed of an X-ray tube 53 and a collimator and shutter assembly 55, a radiation detection means 56, the housings for mounting the detection means and normal electrical connections. In the preferred embodiment, the source of radiation is collimated by assembly 55 to form three finger-like beams of radiation angularly displaced from each other on two degree centers. Other number of beams and spacings, however, may be chosen. In the detection means 56, opposite each of the three diverging pencil-like beams of radiation are three scintillation crystals 58, 60 and 62 which are connected to the three photoelectric transducers such as photomultiplier tubes 64, 66 and 68. Thus, as the carriage 36 traverses the object positioned in cylinder 10, the detection means 56 will produce three sets of attenuation data. For example, data may be obtained from a parallel sets of beams at zero degrees, two degrees, and four degrees relative to vertical during the first traverse.

FIG. 2 is a side view of the assembly of FIG. 1 with the carriages 36 and 38 centered. As shown more clearly in this view, counter weight carriage 38 may have added weights 80 and/or 82 to give carriage 38 the the same mass as carriage 36 with the radiation source and detection mechanisms attached. Rather than being add on weights as illustrated in FIG. 2 in phantom, the weights may be lead inserts or merely thicker construction of the counter balance carriage assembly.

The incremental rotation means 25 is also illustrated more clearly in FIG. 2. In the preferred embodiment, the cylinder 26 is rotated incremently in three degree steps, although other increments are within the scope of the invention. When an odd number of radiation beams N is used and the beams are 2.degree. apart, then rotating the cylinder 26 in increments of N degrees produces data sets representing traverses at 1.degree. rotational increments and after 60 traverses enables collection of 180 unique sets. Incremental rotation may be achieved by using a displacement measuring means 84, for measuring the rotation of cylinder 26 relative to cylinder 10. Alternately, a means 86 which measures the angular rotation of the drive shaft for driving chain 30 could also be used. As a third alternative, a mechanism for measuring the linear displacement for a number of links of chain 30 could be used. A specific incremental measuring means is the model LM-14.0-500-IX 4 manufactured by Disc Instruments of Costa Mesa, California. The placement of the rotation drive 28 remote from the area of examination and the interconnection with a flexible drive 30, is effective for isolating vibrations in the rotation drive from the rest of the structure. Other linkages between motor 28 and cylinder 26 may, however, be used.

A spacial displacement detection means 74 is mounted between carriage 36 and frame 24. This means measures the spacial displacement of carriage 36 relative to frame 24 as it slides along bars 32 and 34 during a traverse movement. The detector means 74, includes a spacial gradient or increment means 88 bearing spacial increment marking extending along frames 24 and reading means 90 for reading the incremental markings on 88. A specific spacial displacement detector which may be used is model LM-14.0-500-IX 4 manufactured by Disc Instruments of Costa Mesa, California.

In operation, the patient is positioned such that the planar slice of the body to be examined is positioned between the radiation source 54 and detection means 56. Having positioned the patient, the tomographic scan may be commenced.

The carriage 36 starts displaced to one side such as in FIG. 1. The motor means starts accelerating carriage 36 to the left, subjecting the planar slice of the body to radiation. At spacially regular intervals, for example every millimeter, displacement detector 74 causes the photomultiplier tubes 64-68 to be sampled. Upon decelerating to a momentary stop at the extreme left position, the photomultiplier tubes will have produced three sets of scan data--one set representing the attenuation along a series of vertical parallel lines, a second set representing the attenuation along a series of parallel lines displaced 2.degree. from vertical, and a third set representing attenuation data along parallel lines displaced 4.degree. from vertical.

The incremental rotation means 25 is actuated causing frame 24 and carriage 36 to be rotated by 3.degree.. Because bars 32 and 34 have been rotated 3.degree. from horizontal, on the return traverse, attenuation data along parallel lines displaced 3.degree., 5.degree. and 7.degree. from vertical will be collected. This traverse and rotate operation is repeated until at least 180 sets of attenuation data are collected. In the preferred embodiment this takes about two minutes.

It will be observed that counter balance carriage 38 is caused to traverse in the opposite direction from carriage 36. Because at any given instant the velocity of these two carriages of the same mass is equal but opposite, the net momentum is substantially zero.

It will further be observed that when motor 40 rotates at a constant speed, the Scotch yoke causes the carriage to traverse back and forth with a varying velocity. Because the velocity of the traverse is varying and the data is sampled at spacially regular intervals, the data samplings are time independent. Thus, the samplings are made at temporally varient but spacially constant intervals.

FIG. 3 is a top view of an alternate embodiment of a traverse and rotate scanner in accord with the present invention for illustrating an alternate means for traversing the carriage. Common elements with the embodiment of FIGS. 1 and 2 are marked with like reference numerals followed by a prime ('). The power source, a gear motor 92, is mounted on carriage supporting frame 24' and has its output connected to a gear box 94. The nonlinear drive means includes gear box 94 and central shaft 41' to which arms 42' and 44' for traversing carriages 36' and 38'. In this embodiment the central shaft is integral with output shaft of gear box 94. In the preferred embodiment, motor 92 is run at substantially constant speeds for efficient use of space, time and funds and to minimize vibration. However, as indicated above because the radiation source and detectors do not need to traverse at a linear speed, motor 92 may be allowed or caused to operate at a constant or fluctuating angular velocity. Similarly, gear box 94 may be a straight forward mechanical linkage which causes arms 44' and 42' to rotate at a constant angular velocity. Again, because the carriage need not, and in the preferred embodiment does not, traverse at a constant linear speed, gear box 94 may be most any linear or nonlinear linkage. For example, a gear box may be chosen with nonlinear gearing which causes the radiation source and detector to traverse the object being examined at a substantially linear constant speed or any one of a number of fluctuating traverse speeds. Gear box 94 may also adjust the angular velocity of arms 42' and 44' so that the radiation source and detectors traverse with simple harmonic motion or with fluctuations of speed. Thus, in this embodiment gear box 94 can be chosen to cause the nonlinear drive means to traverse the source and detectors relative to the object at a constant speed as in the prior art scanners or at a nonconstant speed.

FIG. 4 shows another alternate embodiment of the means for traversing the carriages (like elements marked with a double prime). In this embodiment the nonlinear drive means includes a flywheel 140 which is mounted on central shaft 41" journalled in frame 24". The engagement means includes a pair of rollers or followers 46" and 50" interconnected with shaft 41" at a 180.degree. offset and tracks 48" and 52" on the carriages in which the followers ride. In this drive system, one or more flywheels rotate with shaft 41" and lever arm means connect the shaft with the followers. Each of the followers 46" and 50" may be mounted on an individual lever arm, or each may be mounted on a flywheel making the lever arm means an integral part of the flywheel, or one may be mounted on the flywheel and one on an individual lever arm. Further, flywheel 140 may either be mounted on the shaft or connected to it by gears or belts.

The flywheel 140 is driven by the power source such as an electric motor 144 with the motor having a surface 146 on its armature shaft for engaging and driving flywheel 140. The motor is mounted on a plate 148 which is pivoted at pivot 150. A positioning means 152 adjusts the degree of physical engagement or pressure between shaft surface 146 and flywheel 140.

An encoder or tachometer 154 indirectly monitors the speed of the shaft 41" and flywheel 140 by monitoring the speed of motor 144 to which it is connected by a pulley 156 and belt 158. The tachometer is used to control the speed of motor 144, holding it, the flywheel, and shaft 41" to a constant speed. The constant speed from the inertia of the flywheel and the constant monitoring of motor 144's speed, causes followers 46" and 50" to traverse the carriages very smoothly. This smoothness reduces vibrations and thereby does not impair spacial resolution.

In operation these alternate embodiments function essentially the same as the embodiment of FIGS. 1 and 2, although these alternate embodiments may rotate shaft 41 with a smoother more constant angular velocity than the embodiment of FIGS. 1 and 2 or with nonconstant angular velocities. The sampling of the radiation detectors, as explained above, being dependent only on spacial position and independent of time or traverse velocity, the resultant attenuation data sets will be the same as in the embodiment of FIGS. 1 and 2.

The cable system includes an electric cable 200 for supplying power to X-ray tube 53 and for controlling the shutter and collimator of assembly 55. A series of clips 202 anchors cable 200 to carriage 36. The end of cable 200 opposite the radiation source is connected to a junction box or electrical housing 204.

A second cable 210 having at least leads 212, 214 and 216 connects photoelectric transducers 64, 66 and 68 respectively with electrical junction box 204. Also connected with the junction box is a combined cable means 220 which cables the leads of cables 200 and 210 into a single larger cable. Alternately a plurality of cables may be used.

A flexible cable support means 230 supports cable means 220 between moveable carriage 36 and frame 24. Cable support means 230 is flexible in the plane which is horizontal in FIG. 1 and rigid in the plane which is vertical in FIG. 1. The cable support means follows a generally arcuate path from the carriage to the frame. As the carriage traverses the cable support means flexes such that the arcuate path segment is shifted in space.

In the preferred embodiment, the cable support means is made up of a series of links--alternating main links 232 and connecting links 234. The main links each have a generally rectangular top plate or preferably pair of plates 240 and 242; a generally rectangular bottom plate or preferably pair of plates 244 and 246. Each of the plates is drilled with three round holes 250 and one elongate hole 252 to receive connecting pins 254.

The top and bottom plates are maintained in a spaced relationship by spacing means 256 such as strips connecting the plates or shoulders on pins 252, which engage the top and bottom plates. The spacing means further serve to guide cable means 220 and hold it within the support.

The connecting links are made up of a pair of plates 260 and 262 each of which have four round holes for receiving pins 254. The elongate holes 252 allow the main links to pivot relative to the connecting links but sliding contact of the main and connecting links prohibit movement out of the plane of the link plates. This allows only single planar flexing of support means, i.e. two degrees of freedom.

Support means 230 arcs around to the rear of frame 24 where cable means 220 is attached to the frame. Any convenient means may be used for connecting the cable means from the frame to the power supply, processor and other electrical equipment.

In FIG. 5 a diagram of the attenuation data receiving and handling system is shown. There are three attenuation data receiving circuits 100, 102 and 104 in the preferred embodiment--one corresponding to each of the combination scintillation crystal and photomultiplier tube detectors. Accordingly, the exact number of data receiving circuits will vary with the detection means chosen.

Looking to data receiving circuit 100 as typical, scintillation crystal 62, which may be a thallium doped sodium iodide crystal receives radiation and produces photoluminessence in proportion to the amount of radiation impinging thereupon. Photomultiplier tube 68 transforms the intensity of photoluminessence into a current representative of the intensity. The output current of the photomultiplier tube is amplified by log amp 106. The output of the log amp is connected through a resistive element 108 to an output 110. Further, there is a capacitive element 109 connecting the output to ground. As the radiation source and the radiation detector traverse across the object examined, the amount of radiation striking the scintillation crystal will fluctuate with the degree of attenuation along the path through the body. As the amount of radiation fluctuates so does the signal at output 110.

Each of the data receiving circuits has an RC time constant. One way of expressing the RC time constant is as the time need for the signal at output 110 to decay to e.sup.-1 if the radiation impinging on the crystal were instantaneously stopped. In the preferred embodiment this time is about two milliseconds. It is apparent that if the output 110 were sampled twice in a time period shorter than the time constant, then the second sampled output would be composed of two components. One of the components is the signal indicative of the actual amount of radiation currently detected and the other component is the carried over signal from the previous sampling which had not yet decayed to zero. This carry over then is a source of error in the second sampling.

There are at least two solutions for the carry over error. One solution is to operate the system slowly enough that the time between samplings at the maximum transverse speed is as great or greater than the time constant. Another solution is to correct each sampled signal by the amount of the carried over component.

The first solution, it has been found is satisfactory under some acceptable operating conditions. The maximum speed may easily be determined by speeding up the traverse speed until there is a degradation in resolution. For example, if the spacial sampling interval is 1 millimeter, the maximum speed about 12 inches per second, and the RC time constant is two milliseconds, then no correction has been found to be needed.

The other solution is to subtract the carried over component from the output signal. The carried over component may be approximated multiplying the previous signal by the filter valve ##EQU1## where .DELTA.x is the constant spacial increments between sampling and .tau. is spacial equivalent of the RC time constant. .tau. is determined by the formula ##EQU2## where T.sub.RC is the known RC time constant and t.sub..DELTA.x is the elapsed time since the preceding sampling. As expressed above as the first solution, when t.sub..DELTA.x is large compared to t.sub.RC .multidot..DELTA.x then the correction becomes negligable.

To perform this correction an extrapolation filter means 112 is used. Filter means 112 may be a suitable microcomputer or suitably programmed digital computer. The computer has a memory 114 to store a first output signal from output 110. When the computer then reads a second output signal, it places the second signal in memory 114 and moves the first signal to multiplier 116. The time interval between the first and second sampling is determined by a counter 118, which is connected to a clock 120 and spacial increment sensor 74. A divider 122 connected with counter 118 determines .tau. by dividing the product of the RC time constant and .DELTA.x, a constant for the system, by the time from counter 118. A circuit 124 connected with divider 122 calculates the filter value ##EQU3## Multiplier 116 multiplies the first output signal by the filter value to produce a correction factor. An adder 126 combines the second output signal with the correction factor. An analogous procedure is followed for each sampling in each traverse.

It is apparent that if the traverse speed is slow, the output signal at output 110 is essentially the same as the signal at output 128. This for lower traverse speeds, the filter means 112 is superfluous.

Connected to the outputs of each of the filter means 112, 112', and 112" if used and if not used to outputs 110, 110', and 110" is a processor 130. The processing for transforming data from a traverse and rotate scanner into a visual representation of the radiation attenuation across a planar section of an examined object is well-known in the art. An example of a processing means which may be used is a programmable digital computer programmed to implement the algorithms set forth in "The Fourier Reconstruction of the Head Section" L. A. Shepp and B. F. Logan, IEEE Transactions on Nuclear Science, June, 1974. Alternately the algorithems set for in "Optimum Reconstruction of a Function from Its Projections" Logan and Shepp, Duke Mathematics Journal, Vol. 42, Vol. 4, December, 1975 can be implemented. As a third alternative the implementation shown in U.S. Pat. No. 3,924,129 is usable.

The processor is connected with the spacial displacement detector 74 which cues the processor to sample the outputs 128 or 110. Further, the processor is connected with a T.V. monitor 132 on which the visual representation produced by the processor is displayed.

It will be appreciated that a single computer can be programmed to perform the function of processor 130 and of a plurality of filter means 112.

The above illustrated preferred embodiment is by way of example an explanation only and is not intended to limit the invention to any of the specific details set forth above beyond limitations and restrictions of the claims which follow.