SCANNING SYSTEMS

The present invention relates to scanning systems. It has particular application in scanning systems for cargo, but can also be used in scanners for other applications such as security and high energy medical scanners.

There exists a requirement to screen containers for the presence or otherwise of illegal materials and devices. Often such containers are large (for example a 40 ft x 8 ft x 8ft ISO container) and may contain dense loads such as machinery. In order to provide a non-intrusive and fast inspection process, a number of imaging systems have been developed that use high energy X-ray sources to penetrate through even large containers.

Typically these systems combine a high energy X-ray linear accelerator source with a series of X-ray detectors arranged in a linear fashion to create a one- dimensional inspection line through the object. By scanning the object through the X-ray beam at a controlled velocity, a set of one-dimensional projections are collected which can then be re-formatted into a two-dimensional image for subsequent operator review.

In all such known systems, the image collecting process is inherently undersampled according to the Nyquist sampling theorem which states that the data should be sampled at twice the rate of the highest bandwidth in the signal. Also, X-ray scanners for cargo inspection typically comprise a high energy X-ray source (usually based on an X-ray linear accelerator) with a beam quality of 1 MeV to 9 MeV. The X-ray output from the X-ray linear accelerator is then collimated down to a narrow fan-beam of radiation which is shone through the item of cargo under inspection. A linear array of X-ray detector elements is then positioned opposite to the X-ray source such that it is irradiated by the fan-beam of radiation after attenuation of the X-ray beam by the object under inspection.

In some embodiments of the present invention, an enhanced detection array is described which provides full Nyquist sampling of the image data. In further embodiments, the detection array is further modified to provide materials discrimination capability which can provide significant assistance to the operator in their decision making.

One aspect of the invention provides a scanning method for scanning an object moving in a first direction comprising: irradiating the object with radiation; providing a first detector region and a second detector region wherein the second detector region is arranged to receive radiation that has passed through the first detector region; detecting the radiation after it has interacted with or passed through the object in order to provide information relating to the object; wherein detecting the radiation comprises: detecting radiation at the first detector region; receiving radiation that has passed through the first detector region at the second

detector region; detecting radiation at the second detector region; wherein at the first detector region, the second detector region, or both, detecting the radiation comprises capturing a plurality of sets of detection data using different groups of detection elements, each detection element being arranged to provide information relating to a corresponding portion of the object, and each set of detection data being captured at a different time such that each group of detection elements is arranged to provide information on the same region of the object; and wherein the corresponding portions covered by each group are overlapped in a second direction having a component normal to the first direction.

The independent claims define aspects of the invention for which protection is sought. The dependent claims define preferable inventive features. Any of the features of the dependent claims may be used in combination with the features of other claims or other aspects of the invention, even if they are not explicitly dependent upon them - this will be clear to a person skilled in this field.

Where a feature is claimed in one category (e.g. method, system, detector arrangement, etc.) protection is sought for that feature in other categories even if not explicitly claimed.

Embodiments of the invention will now be described, by way of example only, with reference to the accompanying drawings, in which:

Figure 1 is a flowchart outlining a method according to an embodiment of the invention;

Figure 2 is a flowchart outlining another method according to an embodiment of the invention; Figure 3 schematically shows a scanning system according to an embodiment of the invention;

Figure 4 schematically shows an off-set staggered detector array; Figure 5 schematically illustrates how data capture is co-ordinated in an embodiment of the present invention;

Figure 6 schematically illustrates a detector module used in an embodiment of this invention;

Figure 7 schematically illustrates a stacked detector arrangement according to an embodiment of this invention;

Figure 8 is a graph illustrating different characteristics of high and low atomic mass objects as seen by the scanning system of an embodiment of this invention;

Figure 9 is a graph illustrating the change in response relative to the energy of received radiation;

Figure 10 illustrates a detector arrangement according to an embodiment of the invention;

Figure 11 illustrates a detector arrangement according to another embodiment of the invention; Figure 12 illustrates a scanning system according to an embodiment of the invention;

Figure 13 schematically shows an example of an array of detectors arranged according to an embodiment of the invention;

Figures 14a and 14b illustrate how image data is displayed to an operator in an embodiment of the invention;

Figure 15 graphically illustrates an output radiation profile from a radiation source used in an embodiment of the present invention;

Figure 16 schematically illustrates a detector arrangement according to an embodiment of the invention. Referring to figures 1 and 3, an embodiment of the invention provides a method 10 and system 30 for scanning an object 32 moving in a first direction. The system 30 comprises a radiation source 36 arranged to irradiate the object 32 with radiation as the object moves in the first direction. The radiation source is arranged to provide radiation having a peak energy of at least 900 keV. The scanning system 30 also comprises a detector arrangement 38. The detector arrangement is arranged to detect radiation after it has interacted with or passed through the object 32 in order to provide information to scan the object. The detector arrangement 38 comprises a first detector 40 and a second detector 42. The first detector 40 has a thickness of at least 2 mm. In this embodiment the thickness of the first detector is about 15 mm. In other embodiments the thickness may be more or less and can be tuned as required by a skilled person. The second detector 42 has a thickness of at least 5 mm. In this embodiment the thickness of the second detector 42 is about 30 mm. Once again, it will be clear to the skilled person that this detector thickness can be varied by experimentation in order to tune the detector arrangement 38 as required. In this embodiment, referring to figure 3, the first detector 40 is positioned between the object 32 and the second detector 42. In other embodiments, the skilled person may envisage a different arrangement. In this particular embodiment, this arrangement provides a simple geometry in order to achieve the desired detector configuration such that radiation passing through the first detector 40 reaches the second detector 42 after it has interacted with the object 32.

The system 30 in its broadest embodiment does not include a movement sensor. In some embodiments, the system 30 does include a movement sensor 44 as shown in figure 3. The movement sensor 44 is arranged to measure any one or more of the position, speed, velocity or acceleration of the object 32. The role of the movement sensor 44 will be discussed in further detail below.

Referring to figure 1, the scanning method 10 comprises the step 12 of irradiating an object to be scanned with radiation having a peak energy value of 900 keV or more. The method 10 also comprises the step of providing 14 a first detector region having a thickness of at least 2 mm, and a second detector region having a thickness of at least 5 mm. The second detector region is arranged to receive radiation that has passed through the first detector region. The method 10 comprises detecting the radiation after it has interacted with or passed through the object in order to provide information relating to the object. Detecting the radiation comprises detecting 16 radiation at the first detector region by capturing a plurality of sets of detection data using different groups of detection elements and detecting 18 radiation at the second detector region by capturing a plurality of sets of detection data using different groups of detection elements.

Each detection element is arranged to provide information relating to a corresponding portion of the object and each set of detection data is captured at a different time such that each group of detection elements provides information on the same region of the object. Corresponding portions covered by each group are overlapped in a second direction, which has a component normal to the first direction.

Referring to figure 4, a portion of the first detector 40 is shown. Here, the portion of the first detector 40 is viewed from the point of view of the X-ray source and is seen to comprise two rows of X-ray detector elements 50; a leading row 46 and a trailing row 48. The object is scanned through the X-ray beam past the two detector rows in the direction of the arrow, A. Each row is shown with only a few detector elements for clarity although in a full-scale imaging system, each row will typically contain hundreds to thousands of detection elements. In this embodiment, the elements are identical. In other embodiments, they may be different (sizes, shapes etc) as required.

Information generated by the X-ray detector elements must be read out in a time correlated manor with the scanning of the object; the faster the object is scanned, the faster each row of data must be read out. The relationship between the scanning speed of the object and the rate at which the detector must be read out is shown in figure 5. The vertical arrows at the top of the figure indicate the times at which the X-ray source produces bursts (at a, b, c, d, e, f) of signal and the data from the detector is read out prior to the start of the next X-ray burst. The X-ray pulse labelled "a" will illuminate both the leading and trailing detector arrays. The next X-ray pulse (labelled "b") will also illuminate both leading and

trailing arrays. This means that the data collected for the trailing array in pulse "b" fits exactly in line with the data that was collected for the leading array during pulse "a". To this end, in some embodiments, the method 10 comprises sensing 20 the rate of movement of the object in order to co-ordinate timing of the capturing of the sets of detection data. In such embodiments, there may be a control mechanism in order to adjust the timing of the X-ray bursts from the source 36 taking into account the movement of the object 32 as sensed by the movement sensor 44. For example, if the object 32 comprises a lorry, the movement sensor 44 is able to judge the speed at which the lorry is travelling and adjust the timing of the X- ray bursts from the X-ray source 36 accordingly in order to gather detection data such that each group of detection elements provides information on the same region of the object.

In this embodiment, since the centres for the detector elements in the leading detector array are offset along the length of the array by exactly half a detector element from those in the trailing array, it can be seen that the vertical sampling of the X-ray signal is twice what it would have been had a single detector row been used as is currently the case in known cargo inspection systems.

The spatial frequency bandlimit of the detector array is determined by the physical aperture of the individual detector elements. It can be shown that the arrangement of detectors in figure 4 exactly meets the Nyqist sampling theorem requirement in the vertical direction when sampled in the time domain as described above.

Using the same offset staggered detector array as is shown in figure 4, it is also possible to satisfy the Nyquist sampling theorem in the horizontal direction. In this case, the X-ray beam pulse rate must be doubled while maintaining the same scan velocity as shown in figure 5. These additional pulse are shown as a', b\ c' and so on in figure 5.

In another embodiment of the present invention, an offset staggered detector array is proposed with both vertical and horizontal sampling modes selectable during system operation.

Vertical sampling to the Nyquist criteria is typically always utilised. Horizontal sampling can be operated with full Nyquist sampling when best image quality is required and at other lower sampling rates when low dose imaging is required.

In another embodiment of the invention, a detector module may be fabricated such as the example shown in figure 6. Here an offset staggered detector array 60 is formed from an array of scintillation crystals. A scintillation crystal emits light under exposure to X-radiation and this light is transmitted through the bulk of the scintillation material to a photosensitive device which converts the input optical

signal to an electrical current for subsequent integration by an electronic circuit. The integrated signal can then by converted to digital form for transfer to a digital circuit which is able to render the data to a human readable form such as an image and is also able to store the data to a digital archive such as a computer disk. Typical scintillation detectors should have high density and high atomic number and would typically be selected from the set of materials that include CsI, NaI, BGO, CdW04, LSO, GSO, LYSO and LaBr3. Often the photosensitive device will be a semiconductor device such as a photodiode, an avalanche photodiode or a silicon photomultiplier or it may be a vacuum tube device such as a photomultiplier tube or a more exotic hybrid device such as a micro-channel plate with photodiode readout.

Other detector configurations are possible, for example by using a gas filled ionisation or proportional detector such as one filled with pressurised Xenon gas. To achieve the required detection efficiency the gas would normally be pressured to at least 5 bar. An alternative would be to use a high density wide bandgap semiconductor material such as CdTe, CdZnTe, HgI and so on.

Exactly the same offset staggered detector array principle can be used with other imaging probes including thermal neutrons and fast neutrons which can provide additional diagnostic benefit.

In another embodiment of this invention, a hybrid imaging system is envisaged in which an offset staggered X-ray sensitive detector array is situated proximate to an offset staggered neutron sensitive detector array to allow simultaneous X-ray and neutron imaging inspection of the same container.

The primary equation that governs X-ray attenuation in matter (e.g. as the radiation from the source pulses through the object and each detector) is

I{E) = I ₀ [E)^-I μ{E)dl^ (1)

where KE) = intensity of radiation transmitted through the object at energy E, Io (E) = intensity of radiation emitted by the source at energy E, μ(E) = linear attenuation coefficient of object at energy E and / = line taken by the (pencil) beam of radiation through the object.

The X-ray output from an X-ray linear accelerator is polychromatic having an energy .distribution substantially as shown in figure 15. The maximum X-ray energy (E _p ) results from those electron interactions in the target of the linear accelerator where all of the electron energy is transferred to a single X-ray photon. Typically, less than the full electron energy is transferred to a photon resulting in the broad range of X-ray energies in the X-ray beam. At low energy, the peaks shown in figure 15 are due to fluorescence interactions between the

electrons and the target atoms so resulting in X-rays which are characteristic of the target material.

It is customary to use an integrating detector to measure the X-ray signal that is described in equation 1. In this case, the detected signal can be written as

where l _d - detected signal, μ/E) = linear attenuation coefficient of the detector material at energy E and 5 = path length of the X-ray beam through the detector.

It is therefore clear that l _d retains no knowledge of the energy distribution of the incoming X-ray beam, only of the cumulative effect of all X-ray energies. However, it can also be seen that unless the path through the detector material, s, is very large indeed, some energy will be transmitted through the detector (i.e. it will not have a 100% detection efficiency). Referring to figure 16, if a second detector is placed at the output of the first detector, then the energy transmitted through the first detector has a chance of being absorbed in the second detector. In this case we can write:

l _dB = }/(£)exp - 1 -exp - (3)

where l _iB - intensity recorded in detector B, μ _M (E) = linear attenuation coefficient of detector A material at energy E, μ _iB (E) - linear attenuation coefficient of detector B material at energy E and t - path taken by the X-ray beam through detector B.

Inspection of equation 3 shows that the energy spectrum that is incident on detector B is not the same as the energy spectrum that is incident on detector A. Therefore, detector A can be thought to have retained some energy information even though the integrated output alone is not sufficient to tell what this energy information is. The same is true of detector B. In another aspect of this invention, it is recognised that the measurements that are produced by detector A and detector B are spatially and temporally correlated and that the ratio of the intensity recorded in detector A to that recorded in detector B will necessarily provide some information about the energy distribution of the incident X-ray beam, i.e. - ^l f- = /{/(£)} (4)

where f{ } - function operator.

It can further be seen through inspection of equation (1), that the ratio of detector measurements also includes a factor that is due to attenuation in the object.

Three object parameters will affect the ratio of detectors (equation 4) and these are the linear attenuation coefficient of the object, μ(E) , the path / taken by the X-ray beam through the object and the energy distribution of the primary beam, Io (E). In this situation, there are three unknowns and two measurements and therefore it is impossible to uniquely determine a value for the three object unknowns. The energy distribution of the primary beam, Io (E) is constant from one scan to the next and so an estimate of Io(E) to the first order can be made. Therefore Io (E) is assumed to be known and basic materials discrimination which results in an average atomic mass indication for the object being analysed.

The present invention is concerned with high energy scanning. At low energies (for example most medical scanners) , the photo-electric effect is a mechanism by which X-rays interact with matter within objects being scanned. In contrast, the present invention is concerned with much larger X-ray source energies - namely having a peak value of 900 keV or above. The predominant mechanism governing interactions of radiation within matter at these energies is Compton scattering. The attenuation in matter of X-rays affected by the photo-electric effect shows a dependence proportional to Z ⁴ (where Z = atomic number) . In contrast, Compton scattering produces a Z ¹ dependence. Some Compton scattering is also present at low energies. The detector regions of the present invention are configured such that in the front detector 40, there is approximately a Z ⁴ dependence arising from a combination of the photo-electric and Compton scattering effects. The second, rear detector 42 has a Z ¹ dependence. As a result there are significantly different considerations compared to low energy X-ray scanning, due to the different physical laws governing the interaction of matter. The inventor has realised that for high energy X-ray scanning applications, the front and rear detectors in the claimed arrangement are governed by different physical laws with regards to their interaction with high energy radiation. As a result of the different physical relationships, different detector arrangements are required, relative to low energy X-ray scanners. Accordingly, a first detector is specified as being at least 2 mm thick, whilst the second detector is specified as being at least 5 mm thick. Also, different signal processing is required to account for the combination of the photo-electric effect and Compton scattering occurring at the first detector, and the predominantly Compton scattering effect at the second detector. As a result conventional cargo scanners do not use a dual detector region arrangement as specified in this invention.

As a further embodiment of this invention, a dual-energy sensor array as shown in figure 7 is described. Here, the radiation beam 70 impacts through the long edge of a first offset staggered row of detector elements. The thickness of the first offset staggered detector array 72 is tuned to allow a fraction of the incident radiation to penetrate through the detector elements so that the radiation beam can then intercept a second offset staggered row 74 of detector elements.

In some embodiments, the X-ray beam is generated by using an X-ray Linear Accelerator. The beam has a wide energy spectrum. The maximum X-ray energy in the X-ray spectrum is equal to the peak accelerating potential of the linear accelerator (e.g. 6 MeV in some embodiments) but the mean X-ray energy in the spectrum is typically only one third of the peak (e.g. 2 MeV) and the most probable X-ray energy is much less than this still. In other embodiments, the peak energy is 900 keV, or any suitable higher peak energy.

By adopting the stacked approach shown in figure 7, the front offset staggered row 72 of detector elements pre-filter the X-ray beam and preferentially detect the low energy component of the X-ray signal. The X-ray signal that penetrates through to the rear offset staggered row of detector elements therefore has a higher net energy spectral content than that detected in the front offset staggered row of detectors. This change in energy response of the two sets of detectors is summarised in the diagram shown in figure 9. In figure 9 in this particular embodiment the front and rear detector thicknesses have been tuned to give a similar overall detection probability which is advantageous in the design of the subsequent data acquisition system.

By then comparing the ratio of the signal detected in the front detector element with that detected by the corresponding rear detector element, it is possible to perform some basic materials discrimination. The representative effect on the Front-to-Rear (F/R) detected signal ratio as a function of material thickness for high and low atomic number (Z) materials is shown in figure 8. At lower energies there is no significant differentiation between low and high atomic number materials since almost all radiation is absorbed at the front (F) detector regardless of atomic number of the material. At larger energies the differences between materials is more pronounced. At much larger energies, the differentiation becomes less significant. Taking the absolute values detected by the front and rear detector sets provides the system with an estimate of the total radiological thickness of the object under inspection. The F/R ratio then provides an assessment of whether the object in the image is a small thickness of high-Z material or a large thickness of low-Z material. In this way, a basis assessment of materials type can be performed automatically.

It is noted that the dual energy approach described here applies equally well to other detector configurations such as a single row linear array.

Another embodiment of this invention may use two independent data acquisition systems, one to read out the front detector array data and one to read out the rear detector array data as shown in figure 10 where just the detector component of the data acquisition system is shown. Here, an X-ray detector arrangement 100 is populated onto a circuit card 102 which provides mechanical support for the detector as well as providing the necessary electronic readout 104 and cable interface 106. Two identical cards 102, 108 complete the dual energy detector configuration.

Another practical embodiment of this invention is shown in figure 11. Here, a single circuit card 110 is used to provide the mechanical and electrical interface to two detector arrays 112, 114, i.e. both front and rear. A single cable connection 116 takes data from both detector arrays off to a single data acquisition system. Typically, the front and rear detector data is interleaved and is re-formatted by digital components of the data acquisition system down-stream of the front end detector modules.

In some embodiments, the detector module will provide mechanical, environmental and thermal protection for the sensitive front end radiation detectors. To facilitate quantitative imaging, it is beneficial for the front and rear detector elements to be maintained at equal temperatures.

Further, in some embodiments the detectors point substantially back towards the source of radiation in order to limit parallax detection of the radiation beam. This can be achieved quite simply by using the "saw-tooth" arrangement that is shown schematically in figure 13. Here, the normal to each detector array 138 points back towards the source 139 so that parallax error can by typically contained to within a few degrees.

Referring to figure 12, in another embodiment of the invention, there is provided a system 120 for scanning a moving object 122 in a similar manner to that shown in the embodiment of figure 3. A system 120 includes an X-ray source 126 for irradiating the object with high energy radiation (of peak value 900 keV or more). The system 120 also includes a detector arrangement 128 including a first detector 130 and a second detector 132. This embodiment differs from the embodiment shown in figure 3 in that the object 122 is provided on a conveyor 134. The object 122 is stationary on the conveyor 134. The system 120 also includes a controller 136 which drives the conveyor and controls its movement. The controller 136 controls the movement of the object 122, and the timing of the X-ray bursts emitted by the source 126 such that each set of detection data is captured at a different time and co-ordinated to provide information on the same region of the object as previously described.

In other embodiments it will be apparent to the skilled person that there are other arrangements which can be used to monitor the movement of the object through a scanning zone and thus to co-ordinate scanning as required by claim 1. A combination of a conveyor and movement sensor may be provided in some useful embodiments.

The image data is be displayed in some examples to the operator as shown in figures 14a and 14b. Figure 14a shows how data should be displayed in low dose mode; r-a means data from a trailing detector element with pulse a, f-b means data from a leading detector element with pulse b. In this case the displayed pixel is rectangular. Figure 14b shows how data should be displayed in full Nyquist sampling mode. Here ra means data from a trailing detector pulse a while ra' means data from a trailing detector pulse a' . In this case the displayed pixel is square.

It is also possible to display a materials discrimination image in which case the data should be represented as shown in figure 14b but this time the pixel should display the ratio value F/R. For example F(ra)/R(ra), F(fb)/R(fb) and so on. The value should be scaled to a suitable set of values such that F/R = 1 is scaled to 256 and F/R = 0 is scaled to 0, for example, so that the average brightness of the screen that is presented to the operator maintains a comfortable viewing level.

When the operator screen does not have sufficient pixels to display all of the image data, an interpolation can be adopted in both the vertical and horizontal directions as required until the image dimension is matched appropriately to the screen resolution. The high intrinsic spatial resolution of the image data can be used to provide enhanced display zoom functionality over that which can be used with a conventional linear detector array due to the use of full Nyquist sampling, giving a factor of two increase in zoom level over conventional sub-sampled systems.

Various modifications may be made to this invention without departing from its scope (as defined by the claims). The disclosed scanning system wherein the detector arrangement comprises a first detector region arranged to detect radiation and a second detector region arranged to detect radiation wherein the second detector region is arranged to receive radiation that has passed through the first detector region uses an offset detector arrangement to additionally provide Nyquist sampling rate scanning. In some embodiments the detectors are offset but not sufficiently to achieve the Nyquist sampling rate. In such embodiments the image that is produced still has a greater resolution than a non-offset detector.