For a fuller understanding of the nature and object of the present invention, reference is made to the accompanying drawings, wherein:
The present invention includes a number of aspects. First, the compact design that brings the person being inspected very close to the backscatter detector increases the usable x-ray flux in the image by about a factor of five without exceeding the government-mandated maximum dose of 10 μRoentgens by increasing the solid angle subtended by the backscatter detector. Second, the use of a plastic scintillator for the backscatter detector increases the usable flux by about a factor of four due to the improved stopping power of the thicker detector. Third, an additional factor of two in usable flux is a result of the uniform efficiency of light photon collection by the placement of the photo multipliers away from the electron beam. Fourth, the use of a shaped slot and shaped slits to form the x-ray pencil beam, described below, results in an additional factor of nearly two increase toward the outer edges of the person.
There is a fundamental difference between the requirements of a baggage inspection system and a body scan system which makes the present invention possible. In the case of baggage inspection, the requirement is to detect contraband anywhere in the package. In the case of a body scan system, the requirement is to detect any contraband on the surface of the body. This means that the lower of the dual-energy x-rays need only penetrate the clothing and identify the material behind the clothing and on the surface of the body. This allows the use of a much lower peak energy for the low energy beam of the two energies. Therefore, a peak energy for this lower energy beam that is only capable of penetrating the clothing twice, on the way in and backscattered on the way out, is used. This energy must be less than 30 KeV. The only way to discriminate between the low Z (atomic number) materials that make up both innocuous plastics and explosives is the slight differences in Z and the slight differences in density. The lower peak energy provides greater contrast due to the fact that the absorption of the x-rays due to the photoelectric effect is enhanced for low Z materials at lower x-ray energies.
The present invention employs a system 10 for scanning a subject 12 via x-rays. The basic design of the system 10 is shown in
The detector 20 uses a plastic scintillator that is about 5 cm thick, rather than the 1 mm thick scintillating screens of the prior art,. The detector 20 is in two sections, with the scintillator 66 between the x-ray source 14 and the subject 12 and the detectors 68 at the ends of the scintillator 66, allowing the pencil beam of x-rays to reach the body by penetrating only the scintillator. The large thickness of the plastic scintillator relative to a scintillating screen results in a detection efficiency of about 60%, a large improvement over the prior art scintillating screens. The small 5 cm total thickness of the detector 20 is important to the design of the system. Prior art total detector systems are quite thick, on the order of 30 cm. The small thickness of the detector 20 used in the present invention allows the distance from the x-ray source to surface of the person to be much shorter.
The detector 20 is curved, as shown in
In the prior art systems, the x-ray flux decreases significantly at the edges of the image due to the increased distance traveled by the pencil beam. Likewise, the angle, Θ, of the pencil beam 26 with the normal to the rotating disk 18 and the fixed slit 40 that forms the pencil beam 26 increases at the edges as shown in
Another feature of the invention is to place the x-ray source 14 as close as possible to the subject 12 being inspected. The design concept begins with the maximum width of a person to be inspected which is taken to be about 70 cm. From the outer edges of that widest person, two lines 32 are drawn at 45° angles from center 34, as shown in
A typical implementation of the invention uses an off-the-shelf, inexpensive x-ray source 14 with a spot size of about 1 mm. With this choice of x-ray source 14, the widths of the slit 40 and slot 42 must also be a minimum of 1 mm in order that the x-ray flux in the pencil beam 26 is not reduced. In the present design this width is about 3.1 mm, as described below. The rotating disk 18 with the slots 42 and the sheet 16 with the fixed slit 40 must also be sufficiently close to the subject 12 so that the pencil beam 26 is less than 4 mm across at the subject 12. The Nyquist theorem permits a sample every 2 mm, thereby achieving a 2 mm spatial resolution.
Placing the x-ray source 14 as close as possible to the subject 12, as described above, reduces the power required to reach the federally-mandated maximum permissible dose of 10 μRoentgen. In the design shown in the figures, the required power of the x-ray source 14 is about 200 watts, less than corresponding systems with equivalent performance. Despite the reduction in power, the system of the present invention performs better than prior art systems.
The system is capable of providing 2 mm pixels. Therefore a person that is 198 cm tall and 60 cm wide has approximately 300,000 total 2 mm pixels over the entire inspection area.
The present system uses a dual-energy x-ray source with a high peak x-ray energy of 100 KeV and a low peak x-ray energy of 30 KeV. The higher energy can be anything above 50 KeV without substantially changing the performance of the invention. A key element of the invention is that the lower of the two energies be as low as possible to be able to differentiate between the innocent low Z materials and contraband that may also contain only low Z materials. In addition the lower energy must be of sufficient intensity and peak energy to penetrate the clothing without too much absorption relative to the contraband and the body. Further this must be accomplished without exceeding the current federal limit of 10 μRoentgens per inspection, as defined in the federal regulations.
The major process by which low atomic number (Z) elements remove x-rays from the primary x-ray pencil beam is Compton scattering, a fraction of which is scattered backwards into the detector, and Compton absorption. This process is proportional to the total number of electrons per unit mass of the material and is very nearly the same, or independent of Z, for all materials in the low Z range except for hydrogen which has twice the density of electrons than the other materials. In addition, there is a small attenuation of the x-ray beam due to the photoelectric (PE) effect which removes the x-ray completely from the beam. This effect is proportional to Z2 and is the main differentiation between materials. Since the K edge for absorption for low Z elements is less than 1 KeV and the PE absorption coefficient falls off approximately as the cube of the x-ray energy above this value, it is important to use a lower peak energy that is as low as possible.
The following items are calculated below (The difference between the effective dose in Rem and measured exposure in Roentgen is ignored in this calculation, which is a conservative assumption, and, in any event, does not significantly affect the result of the calculation.):
1. The power, W, of the x-ray tube required to achieve a dose, Finc, of less than 10 μRoentgen, at the subject. This power is 200 Watts.
2. The efficiency, eff, of the backscatter detector. This efficiency is eff=0.63, about four times greater than prior art systems.
3. The number of standard deviations (SD's), or the contrast, between a thin phantom hidden on the surface of the subject and the surrounding soft tissue. This number, 11.5 SD's, is a measure of the effectiveness of the system.
4. The average solid angle, Ω, subtended by the x-ray backscatter detector at a point on the surface of the subject. This solid angle is Ω=5 steradians, about ten times greater than prior art systems.
5. The width of the fixed slit and each of the rotating slots as a function of the distance along their lengths. This variation in width allows the pencil beam x-ray flux to remain constant as the beam moves across the subject.
The calculations start with the following parametric values based on the current system design. Dslit is the distance from the x-ray source to the slit/slot and is Dslit=12 cm. Lslit is the length of the fixed slit and is Lslit=24 cm. Φmax is one half of the angle between neighboring slots in the plane of the rotating disk and is (Φmax=45×(π/180) radians. Dper is the distance from the x-ray source to the surface of the subject and is Dper=20 cm.
The total area, Atot, to be inspected is determined by the area as defined by the length of the fixed slit, Lslit=24 cm, and the height of the scan, Hscan=198.12 cm (78 inches), at the distance of the slit/slot, Dslit, from the source. So
A
tot
=L
slit
×H
scan=4.755 ×103 cm2.
The area of the pencil beam at the subject is
A
pb=0.42 cm =0.16 cm
at the subject. If it is assumed that the x-ray tube operates at W=200 Watts and V=100 kVp, then
I=W/V=2×103A.
The Health Physics and Radiologic Health Handbook, Revised edition, page 195, gives the emitted flux, F, in R/mA/min, as a function of kVp for a typical x-ray tube at 1 meter. For a peak voltage of 100 kVp, the emitted flux at 1 meter is F1,1=0.564 R/min/mA. So the incident flux at the front of the slit/slot at 100 kVp is
or
F
incslit=1.306 R/sec.
F
incpersec
=F
incslit
×A
pb×(Dslit/Dper)R/sec/pencil beam.
The total exposure time, Tpb, for each pencil beam is the total inspection time, Tinspect divided by the number of pencil beams, Nbeams. The number of pencil beams is
N
beams
=A
tot
D
per/(Apb×Dslit)=4.953×104.
T
pb
=T
inspect
/N
beams=6.057×10−5 sec.
If the time for the inspection is Tinspect=3 sec, the total incident flux at the subject, Finc, is
F
inc
=F
incpersec
×T
pb=7.591×10−6R=7.591 μR.
or
F
reg=3.212×10−6R=3.212 μR.
If the average energy of a photon is V/2*0.001 MeV, the number of photons in the backscattered pencil beam, Npb, (using the formula in the American Institute of Physics Handbook, third edition, page 8-305) is
or
N
pb=5.223×104 photons/pencil beam.
N
back=3.917×103 photons/pencil beam.
If the thickness of the backscatter detector is Tdet=5 g/cm2 and the mass absorption coefficient at 45 keV is μdet=0.2 cm2/g, the average efficiency of the backscatter detector is
eff=1−e−Tdet×μdet=0.632.
N
detpix
=N
back×eff=2.476×103 detected photons per pixel.
Using the Nyquist theorem, the pencil beam can be sampled twice for each passage of the beam over the slot and twice as the slot moves down across the subject. Thus the pixel size, pix, at the front of the subject is one half of the pencil beam size at the front of the subject or
or
pix=2 mm.
To estimate the performance of the system, an object phantom is assumed to have an area Aphan=2.5 cm2, and a thickness of dphan=0.5 cm. It is resting on a large plate (equivalent to the soft tissue of the human body) that is dsur=5 g/cm2 thick. This is a very challenging phantom. All materials are low Z. The number of detected backscatter photons from the phantom, Fphan, is
or
F
phan=3.064×104 photons
scattered backward and detected. And the number of photons from a similar area, Aphan, of the surround, Fsur, is
or
F
sur=2.786×104.
or
N
SD=11.517 standard deviations
to see the small piece of explosive. Thus an explosive 5 mm (0.2 in) thick and 2.5 cm (1 in) on a side placed over an area of the subject with soft tissue 5 cm (2 in) thick would be seen with 11.5 standard deviations of contrast.
Below is a calculation of the solid angle subtended by the point where the pencil beam hits the surface of the subject, on the backscatter detector. Referring to
or
Ω=5.544 steradians.
Now the width of the slit/slot that allows the flux to the person to be uniform as the pencil beam moves transversely from side to side is calculated. The cross-section of the pencil beam must increase because both the distance from the x-ray source to the slit/slot and the angle of the pencil beam with the normal to the plane of the rotating disk increase as the pencil beam moves from the center to the edge of the fixed slit and from the top to the bottom of the slot as the disk rotates. These effects are compensated for by increasing the area of the pencil beam cross-section. Preferably, one half of the correction is made by the slit and the other half is made by the slot.
The area of the pencil beam is Apb=0.16 cm2 at the person, so the minimum area of the pencil beam at the slit is
A
pbslitmin
=A
pb×(Dslit/Dper)=0.96 cm2.
Δminslit=√Apbslitmin=0.31 cm.
As indicated above, the length of the slit Lslit=24 cm, the distance from the x-ray source to the slit Dslit=12 cm, and one half of the separation angle of the rotating slots Φmax=45×(π/180) radians. The radial distance, R, from the axis of rotation of the disk to the end of the fixed slit is
R=L
slit/(2 sin Φmax))
and the radial distance, r, from the axis of rotation of the disk to the center of the fixed slit is
r=R cos (Φmax)
If R=16.971, then r=12 cm. Θ is the angle of the x-ray beam from the beam normal to the fixed slit. For ten locations, i=0 . . . 9, along the fixed slit,
Φi=45i×(π/(180×9)) radians
and
Θi=a tan (r×tan(Φi))/Dslit) radians.
A
pbsliti
=A
pbslitmin/(cos(Θi))3.
The cube of the cosine includes a square of the cosine to take into account the increase in distance from the x-ray source and an additional cosine to take into account the angle between the pencil beam and the normal to the disk. The width of the slit at location i is
Δi=√Apbsliti.
λi=r/(cos((Φi))−r
and the position of location i along the slit is
L
i/2=Dslit×tan(Θi).
These values for each of the ten locations i=0 . . . 9 is shown in TABLE I and graphically in
Note in
A detailed calculation has been performed to produce the x-ray backscatter flux from carbon and plastic at the two peak energies of the x-ray source, 100 KeV and 30 KeV. This calculation uses similar values for the flux as those above. The calculation uses each of the appropriate x-ray energy intervals and then sums the results. This process is known to people of average skill in the art.
The system of the present invention employs technologies that are well-known to those versed in the art. Dual-energy x-ray sources are now common in the industry, and the algorithm for dividing two images pixel by pixel is quite simple and can be done while the data is being collected.
The present invention has application to x-ray systems used in any application where the aim is to identify materials that are not shielded significantly by overlying material. An example of another application is radiology of the skin to evaluate potential tumors. The requirement is to examine the region just below the surface of the skin behind the lesion and perhaps identify the nature of the chemistry in this region using backscatter and dual energy or even multiple energies to identify the elemental composition and thus the molecular composition, including the concentration, of different elements and molecules.
Thus it has been shown and described a method a apparatus for using dual-energy x-rays in a backscatter scan system which satisfies the objects set forth above.
Since certain changes may be made in the present disclosure without departing from the scope of the present invention, it is intended that all matter described in the foregoing specification and shown in the accompanying drawings be interpreted as illustrative and not in a limiting sense.
The applicant wishes to claim the benefit of U.S. Provisional Patent Application No. 60/837,838, filed Aug. 15, 2006 for BACKSCATTER PERSONNEL SYSTEM in the name of Martin Annis, and U.S. Provisional Patent Application No. 60/875,630, filed Dec. 19, 2006 for METHOD TO USE A DUAL ENERGY X-RAY SOURCE IN A BACKSCATTER BODY SCAN SYSTEM in the name of Martin Annis.
Number | Date | Country | |
---|---|---|---|
60837838 | Aug 2006 | US | |
60875630 | Dec 2006 | US |