Density cluster imaging

Abstract

A method of improving a generated image obtained from a gamma camera used in medical imaging, the image improved by excluding low energy and scattered gamma rays from contributing to the generated image, the method comprising the steps of: • the gamma camera detecting gamma ray photons; • converting the detected gamma ray photons into one or more optical photons; • detecting the resultant optical photons on a photoactive surface; • processing the detected optical photons into a primary data array comprising pixels, wherein the pixels map the detection of the optical photons on the photoactive surface, wherein populated pixels correspond to the detection of the optical photons on the photoactive surface; • identifying clusters of populated pixels in the primary data array; • determining the density of the populated pixels in the identified clusters; • categorizing the clusters into high density clusters (clusters having a high density of populated pixels) and into low density clusters (clusters having a low density of populated pixels); • excluding the low density clusters from contributing to the generated image, these low density clusters corresponding to low energy and scattered gamma rays.

Description

[0001] Density Cluster Imaging

[0002] Field of the invention

[0003] This invention relates to density cluster imaging. In particular, though not exclusively, the invention relates to density cluster imaging using gamma rays in medical imaging, as well as methods of using the same.

[0004] Background of the invention

[0005] A gamma camera (also called y-camera or scintillation camera) is a device used in medical imaging to image gamma radiation being emitted from radioisotopes. As such, the device must be suitable for the detection of low levels of gamma radiation given off by radioisotopes at doses administered to a subject, while also providing images with suitable resolution to allow a medical practitioner to perform medical analysis, diagnosis or treatment etc. Therefore, gamma ray medical imaging devices have a sensitivity and resolution that are tailored to their role. Typically, the device is required to form medical images of radioisotopes with a gamma emission energy between about 35 keV and about 511 keV, and where these are administered to a subject (e.g. typical body weight of about 70 kg) at levels between about 10 MBq and 1000 MBq. Ideally the images produced will have a clinically meaningful resolution acquired within about 30 minutes or less. The isotope, activity etc. will be adjusted to take account of the conditions e.g. the mass of the subject and the tissue being targeted.

[0006] Gamma cameras are commonly used to create images in a technique known as scintigraphy. In scintigraphy radioisotopes are commonly attached to tracer agents or drugs (radiopharmaceuticals) which travel to specific organs or tissues, and in that way those target tissues can be imaged. So, this technique can create visual representations of the physiological processes of the target tissues for clinical analysis and medical intervention, as well as visual representation of the function of organs or tissues. Therefore, this technique goes beyond revealing internal structures hidden by the exterior of the subject, it can target certain organs or disease states and / or provide more information about their anatomy and function.

[0007] Gamma cameras are typically large, expensive and fixed installations, and so are typically housed in a dedicated room in a hospital. Patients are typically required to go to the location and are inserted into the body of the device to be scanned. Smaller gamma cameras with a more limited field of view are more mobile but remain relatively large and unwieldy, and further may be lacking in sensitivity and / or resolution and / or suitable field of view. Examples of gamma cameras used for medical scintigraphy include Siemens Healthineers' Symbia Intevo Excel, Oncovision's Sentinella and Digirad's Ergo Imaging System.

[0008] A significant problem encountered using gamma ray cameras in medical imaging is making sure that the gamma ray camera substantially only records gamma ray photons that have come directly from the radioisotope that has been given to the subject. Gamma ray photons that have been deflected from their original path (scattered) should not contribute to the image. An ideal scenario is shown schematically in Figure 1. Figure 1 shows a gamma ray photon emanating from a subject (10) that has been dosed with a gamma ray (y) source (11) (shown as the dashed region in the subject). The gamma ray photon (12) passes through a collimator (e.g. a 'pinhole'), hits the detector and is detected by the detector. In this case the detector converts the gamma ray photon into a voltage (23), where the voltage corresponds to the energy of the gamma ray photon. In Figure 1 the output voltage from the device is shown as a voltage peak. However, scattered gamma ray photons (e.g. gamma ray photons emanating from the subject, but where the original trajectory of the gamma ray photon has been changed in some way) and background gamma ray photons can contribute unwanted noise, making the desired image worse.

[0009] An example of this is shown schematically in Figure 2 where a gamma ray photon is scattered (see the inflection point shown in the dashed line (13)) in an effect known as Compton scattering. In this process the gamma photon encounters and interacts with a charged particle, normally an electron. The gamma photon transfers some of its energy to the charged particle and changes its direction of travel. The gamma ray photon may be from a background source, or indeed originally scattered from the radiation source. Again, the detector converts the scattered gamma ray photon into a voltage. However, in this case, because the scattered gamma ray photon has a lower energy, the detector produces a lower resultant output voltage (24). So, in Figure 2 the scattered gamma ray photon gives a voltage corresponding to the lower (striped) voltage peak (24), whereas a gamma ray photon coming directly from the radiation source corresponds to the bigger (unstriped), voltage peak (23). A gamma ray photon may undergo multiple scattering events, losing energy and changing trajectory each time, such that scattered gamma ray photons have a broad range of photon energies and the original 'true' trajectory becomes difficult to determine.

[0010] A class of gamma ray imaging cameras, known as Compton cameras, rely on detecting scattered photons to determine the source distribution. Compton cameras must measure the energy lost in the scattering event, the total photon energy, and the angle through which the gamma ray photon was scattered to determine the gamma ray photons originally trajectory prior to scattering. These types of cameras are highly specialised and generally not suitable for medical imaging applications because of the diffuse distribution of the radioisotope to be imaged and the large and variable scattering effect of biological tissue.

[0011] A gamma camera for medical imaging is therefore designed to capture only gamma rays that originate directly from the radioisotope source, where these gamma ray photons will have a defined energy signature.

[0012] So, ideally, a gamma ray camera for medical imaging needs to be able to discriminate between gamma ray photons originating directly from the radioisotope source and gamma ray photons that have been scattered or are from a background source. It is advantageous to omit these lower energy photons which provide no direct information about the source distribution, and serve only to contribute to image noise.

[0013] In the prior art, an 'energy window' is used to filter out lower energy scattered gamma ray photons. Gamma ray photons that originate from the true line of sight (not scattered) of the radioisotope source will have a defined energy signature (e.g. having an energy falling within a fairly narrow energy range). This is because they will have lost none, or little, of their original energy when emitted from the radiation source. However, gamma ray photons that have been scattered will have lost some energy, and so will fall outside the energy window. Not only can an energy window exclude scattered gamma ray photons, it can also be used to exclude background gamma ray photons which also do not fit within the fairly narrow energy window.

[0014] With reference to Figure 2 and 3, this is done in gamma ray cameras by converting the gamma ray photon energy into a voltage, where the magnitude of the voltage is directly proportional to the gamma ray photon energy. The detector will also record the position of the detected event as an X and Y coordinate. The energy window comprising in effect an upper and lower voltage level, typically corresponding to around ± 10% of the gamma ray photon photopeak value (as a voltage, or converted into energy units of keV). Any generated signals which are outside this energy window are rejected, and so not included in the final image. The gamma camera then builds a pixelated image using only the accepted pulses with known X,Y coordinates. With reference to Figure 3, the energy window is shown between the dashed vertical lines which are either side of the photopeak value (33). In this case, for example, the commonly used medical isotope Tc-99m has an energy peak (photopeak) of 140 keV. So gamma ray photons coming directly from the radiation source with energies between about 120 and 160 keV will be accepted within the energy window. However, photons of lower energy such as 'Compton scatter' (e.g. shown as the low broad peak in Figure 3; (32)) falling outside the window are not counted and will not be recorded by the gamma camera. In this way the energy window can be used to filter out scattered and background gamma ray photons. The skilled person can adjust the size of the energy window to suit need.

[0015] This image formation process applies to analogue (for example commonly using scintillation detectors and photomultiplier tubes) or digital detector imaging systems that convert the gamma ray photon energy to a voltage pulse where the magnitude of the generated voltage pulse is directly proportional to the energy of the gamma ray photon. It cannot be easily used in gamma cameras with other designs, for example in very small gamma cameras that do not employ photomultiplier tubes, or those that employ highly pixellated detectors where a gamma ray interaction is detected across multiple detector pixels. These types of detectors produce multiple voltage pulses, one per detector pixel (and these often resemble short lines over the detector) that would necessarily need to be combined to create a single voltage pulse. This process often needs to segregate multiple gamma ray interaction events and identify which detector pixels belong to which event. Various algorithms could be used, however these are often computationally intensive and where computational intensity scales (linearly or non-linearly) with the number of interaction events to segregate. This can limit performance (for example at large photon fluxes) and make implementation difficult.

[0016] There remains a need in the art for solutions to the problems encountered in gamma ray imaging. In particular, there remains a need in the art for improved solutions to the problem of image formation and rejection of scattered gamma ray photons.

[0017] Summary of the invention

[0018] In a first aspect of the invention, there is provided a method of improving a generated image obtained from a gamma camera used in medical imaging, the image improved by excluding low energy and scattered gamma rays from contributing to the generated image, the method comprising the steps of:

[0019] • the gamma camera detecting gamma ray photons;

[0020] • converting the detected gamma ray photons into one or more optical photons;

[0021] • detecting the resultant optical photons on a photoactive surface;

[0022] • processing the detected optical photons into a primary data array comprising pixels, wherein the pixels map the detection of the optical photons on the photoactive surface, wherein populated pixels correspond to the detection of the optical photons on the photoactive surface; • identifying clusters of populated pixels in the primary data array;

[0023] • determining the density of the populated pixels in the identified clusters;

[0024] • categorizing the clusters into high density clusters (clusters having a high density of populated pixels) and into low density clusters (clusters having a low density of populated pixels);

[0025] • excluding the low density clusters from contributing to the generated image, these low density clusters corresponding to low energy and scattered gamma rays.

[0026] Advantageously, the method of the invention removes low energy gamma rays during image acquisition. Advantageously, the method of the invention can also be used with gamma cameras where these are not used to produce voltage pulse proportional to the energy of the gamma ray photon detected. These advanced systems can be relatively small and can offer high spatial resolution.

[0027] In particular, the invention provides for improved solutions to the problem of image formation and rejection of lower energy gamma ray photons and detector noise, where high resolution pixellated detectors are employed for gamma ray imaging, e.g. where a gamma ray interaction is detected across multiple detector pixels. Advantageously, the method of the invention is a more computationally efficient approach than attempting to sum a total voltage over multiple pixels (e.g. using a DBSCAN method), in particular when the energy resolution is low. In an embodiment, the method does not use DBSCAN. Advantageously, because the method of the invention is more computationally efficient, this means that the associated hardware specification / requirements can be lessened and / or reduced in size. For example, if less processing is needed by the processing unit, then a smaller less powerful, less expensive processing unit can be used. In turn, a smaller processer cooling unit can be used or even dispensed with. This lowers the assembly cost and any ongoing maintenance costs. Advantageously, smaller hardware components also allow smaller more portable systems to be manufactured. Also, beneficially, because the method is more efficient, it uses less energy so lowers the associated running costs.

[0028] Advantageously, the method of the invention provides an enhancement in the modulation transfer function (signal-to-noise ratio and / or resolution) in the resultant image. For example, when the method of the invention is used in medical imaging, this gives improved images that are more useful in medical diagnosis and / or treatment and may require less time to acquire the images. This improvement is obtained because the images acquired using the method of the invention are derived substantially from only the high energy photons from the radiation source, and by eliminating signals from lower energy photons, that may have resulted from gamma ray photon scattering.

[0029] By the way of some context, an embodiment of the method of the invention will be discussed with reference to Figure 4. Figure 4(a) shows the image that would be produced by the gamma ray camera if the embodiment of the invention was not used. That is, the gamma ray camera has detected several detection events that could be gamma rays. These detection events have been converted into optical photons. The more energetic the detection event the more optical photons are produced (so a more energetic gamma ray photon will produce a bigger 'splash' of optical photons). These optical photons are then detected on a photoactive surface (e.g. a CMOS sensor), and then may be converted into an output image. As such, Figure 4(a) shows detection events detected by the gamma ray camera. These detection events could be from the desired radioisotope source, or could be background noise or could be from scattered gamma ray photons. However, the method of the invention is designed to discriminate between detection events from the desired gamma ray source and those that are not from the gamma ray source (e.g. photons that are of lower energy). The embodiment shown in Figure 4 does this by processing the detected optical photons into a primary data array comprising pixels, wherein the pixels map the detection of the optical photons on the photoactive surface, wherein the populated pixels correspond to the detection of the optical photons on the photoactive surface. In Figure 4(b) clusters in the primary data array are shown in illustrative internal boxes. In Figure 4(c) the largest cluster is shown overlaid with a 9 x 9 kernel. The embodiment in Figure 4 then identifies clusters of populated pixels in the primary data array, and determines the density of the populated pixels in the clusters. The method then categorizes the clusters into high density clusters (clusters having a high density of populated pixels) and into low density clusters (clusters having a low density of populated pixels). With reference to Figure 4(d), the method then excludes the low density clusters from contributing to the generated image. These low density clusters corresponding to low energy and scattered gamma rays. So, a single high density cluster is shown in Figure 4(d).

[0030] In an embodiment the method comprises assigning each high density cluster to the detection of a (single) gamma ray photon, and using the gamma ray photon detection event to contribute to the generated image. With reference to Figure 4(e), it can be seen that the high density cluster has been converted into a single populated pixel. So, with reference to Figure 4, the original unprocessed image in Figure 4(a) is processed using the method of an embodiment of the invention to give a single populated pixel in Figure 4(e), with the single populated pixel corresponding to the detection of a single gamma ray photon at an X-Y position on the detector. Low density clusters, corresponding to low energy gamma photons, have been rejected and do not contribute to image formation. For example, it may be considered that the method of the invention is making use of the size / intensity of the two dimensional impact ('splash') made by the gamma ray photon when it hits the detector (where the size / intensity of the substantially circular splash increases with the energy of the gamma ray photon); and so 'big / intense splashes' correspond to gamma ray photons originating from the radiation source and where 'small / diffuse splashes' are rejected. In this embodiment, and in any other aspects and / or embodiment herein disclosed, the 'splash' is substantially circular. In an embodiment the 'splash' is symmetrical. In an embodiment the 'splash' is largely connected. In these embodiments, only substantially circular 'splashes of photons' are considered relevant by the method. In an embodiment, 'splashes of photons' that are not substantially circular are rejected. In an embodiment, 'splashes of photons' that are substantially linear are rejected. Circular in the context of this application may be understood to take into account the pixelated nature of the splash recorded, and so these 'splashes' may not necessarily have a smooth or tidy edge. For example, in an embodiment, the shapes picked out with a square outline in Figure 4(b) may be identified as substantially circular (though having somewhat ragged edges); albeit in the end, only one of these features was ultimately deemed to qualify as coming from a gamma ray, as herein discussed in relation to Figures 4(a) to (e). Without wishing to be bound by theory, a circular splash in the present context may be considered to be a shape which from its centre to any edge is substantially the same.

[0031] In an embodiment, the single populated pixel is positioned at the centre of the high density cluster (centroided). This single populated pixel is deemed to correspond to the detection of a single unscattered gamma photon. The location of the populated pixel corresponds to the point of origin of an unscattered gamma ray photon and so may then be used in making an image. This centroided step might be understood for example with reference to the transformation shown in Figure 4(d) to Figure 4(e), where the centre of the large cluster is determined. In an embodiment, unpopulated pixels in the primary data array correspond to the absence of detected photons on the photoactive surface. In an embodiment, unpopulated pixels in the secondary data array correspond to the absence of detected photons on the photoactive surface. In effect any pixels that do not contribute to a high density cluster may be emptied or 'zeroed'. In that way, these pixels will not contribute to the final image. This might be understood for example with reference to the transformation shown in Figure 4(a) to Figure 4(e), where only one pixel contributes to the generated image (the rest of the pixels are empty or zeroed).

[0032] In an embodiment, the detected gamma-ray photons have an energy in the range 20 keV to 600 keV, optionally 50 keV to 500 keV and further optionally 100 to 400 keV. The method of the invention may be tuned such that gamma ray photons having a certain energy (e.g. corresponding to emissions from the radioactive source being used in a subject) produce clusters that are deemed to be high density. That is, a gamma ray in this energy range will produce a characteristic 'splash of photons' and the method of the invention may be tuned to identify that a splash of a certain size / intensity is a high density cluster. Smaller clusters are rejected. Significantly larger clusters may also be rejected e.g. if these relate to cosmic rays.

[0033] In an embodiment, the gamma rays are not directly converted into electrons for use in the preparation of the generated image. In an embodiment, a CZT detector is not used. In an embodiment, the gamma rays are not directly digitized for use in the preparation of the generated image. It should be noted that the invention is therefore not making an image using in effect the result shown in Figure 4(a). This is because it will contain too much noise.

[0034] In an embodiment, the photoactive surface has an array of photoactive cells, the photoactive cells activated by the optical photons. In an embodiment, the photoactive surface comprises a CMOS or CCD or EMCCD sensor.

[0035] In an embodiment, the method of improving the generated image is by increasing the signal to noise ratio. In an embodiment the method of improving the generated image is by improving the resolution. In an embodiment the method of improving the generated image is by increasing the modulation transfer function. In an embodiment, the method of improving the generated image is by rejecting gamma rays that have been scattered or deflected from their original trajectory. In an embodiment, the method of improving the generated image is by rejecting gamma rays that have been inelastical ly scattered or deflected from their original trajectory.

[0036] In an embodiment only the high density clusters are used to contribute to the generated image. In an embodiment, the method of improving the generated image is by rejecting gamma rays that result in a small optical splash on the photoactive surface from contributing to the generated image. In an embodiment, the method of improving the generated image is by rejecting gamma rays that result in a large but diffuse (low density) optical splash on the photoactive surface from contributing to the generated image.

[0037] In an embodiment, the method of improving the generated image comprises rejecting cosmic rays from contributing to the generated image. This may be done using the method of the invention, by looking for very dense and / or very large clusters (i.e. much bigger than those that are produced by the radioisotope source being used in the subject) and in effect removing them from the image.

[0038] In an embodiment the generated image comprises a single frame image, which may be combined with one or more additional single frame images to form a frame summed image. With reference to Figure 5 it can be seen how frames 1 to 4 (single frame images) and indeed up to 'n' single frame images maybe combined to form the frame summed image on the right. For example, in the schematic illustration, the single frames (41 to 44) have been combined to show a representation of the buildup of radioactive isotope (e.g. Tc-99m) in the thyroid of a patient (51). The resultant frame summed image (50) might be used to identify abnormal organ function.

[0039] In an embodiment, the gamma ray photons are detected for a period of time (frame period), and only the gamma ray photons detected in this period of time are used to contribute to the generated image, wherein the generated image is a single frame image. In an embodiment, the number of frames acquired in the frame period is optimized. In an embodiment, the number of frames captured per second (frame rate) is optimized.

[0040] In an embodiment, the frame rate is optimized such that the chance of two gamma rays being detected at the same location, such that the clusters begin to overlap, is low, e.g. less than about 1 in 10,000, optionally less than about 1 in 1000, further optionally less than about 1 in 100. In an embodiment, the period of time (frame rate) is optimized for improving the generated image. In an embodiment, the period of time (frame rate) is optimized so that the resultant single frame image is not (overly) saturated. In an embodiment, the frame rate is adjusted such that fewer than 50 gamma rays are detected per frame image, optionally fewer than 10 gamma rays are detected per frame image, further optionally fewer than 5 gamma rays are detected per frame image (for example the first frame (41) in Figure 5 shows the detection of 3 gamma ray photons and their relative X-Y locations / positions).

[0041] In an embodiment the generated image is a frame summed image (e.g. see 50 in Figure 5), the frame summed image being formed by combining two or more single frame images (e.g. see 41 to 44 in Figure 5). The invention can be used to produce a single frame image. In effect the longer the 'exposure window' the more gamma rays will be detected and captured in the frame. However, if the exposure window is too long the frame will become too saturated, and clusters will begin to merge and ultimately the image will become saturated. The method of the invention relies on picking out low and high density clusters and so if these begin to merge the method becomes less efficient. For example, two or more low energy clusters could merge and falsely give the impression that it is a high density cluster, so that noise increases. Also, two or more high density clusters could overlap and falsely give the impression of a single high density cluster so the number of true gamma events are undercounted, and the event would be recorded in the wrong location on the detector, e.g. between the two clusters. However, if the 'exposure window' is too short there may be too many frames that are empty and are needlessly processed by the method of the invention. This may inefficiently use computer processor time. Ideally, the method is tuned so that there is at least one (and up to about 10) high density clusters per individual frame image. In that way the method of the invention can efficiently identify each true high density cluster (corresponding to detection of gamma ray photons that have not been scattered i.e. 'true' signal) in each frame image, while also minimising erroneous counting of low density clusters (corresponding to detection of gamma ray photons that have reduced energy as a result of scatter i.e. 'noise'). These single frame images can then be combined into a framesummed image.

[0042] In an embodiment, the pixels in the primary data array are averaged or combined with one or more adjacent pixels in the primary data array (binning), to form an updated primary data array to replace the previous primary data array. In the art this process is known as 'binning' and can be used to in effect reduce the amount of information that needs to be processed. For example, with reference to Figure 6 it can be seen how an 8 by 8 matrix of 64 pixels can be reduced to a 4 by 4 matrix of 16 pixels by using 2 by 2 binning. Advantageously, in this case the data processing is therefore reduced to a quarter.

[0043] In an embodiment, the updated primary data array is smaller than the previous primary data array, optionally the updated primary data array is reduced to 1 / 4 (quarter, e.g. 64 pixels reduced to 16 pixels), 1 / 9, 1 / 16, 1 / 25, 1 / 36, 1 / 49, 1 / 64, 1 / 81 or 1 / 100 of the pixels in the previous primary data array. In an embodiment, the aspect ratio is preserved in binning. It is also possible to bin where the aspect ratio is not preserved, but this requires additional action / processing. In an embodiment, the primary data array is 1024 x 1024, 512 x 512, 256 x 256, or 128 x 128 pixels. In an embodiment, the primary data array is 512 x 512 or 256x 256 pixels.

[0044] In an embodiment, the position of each pixel in the primary data array maps to a relative position in the array of photoactive cells. Because a frame summed image is made up of many combined single framed images, it is useful that each detected gamma ray event has a common reference orientation. For Example, in Figure 5, single frame image 44 and single frame image 'n' have gamma ray photons in the same X-Y locations / positions. So, in the frame-summed image these positions will overlap / superimpose, giving more intensity at these locations (in Figure 5, the darker locations indicate the detection of more gamma ray photons).

[0045] In an embodiment the step of excluding the low density clusters comprises the step of removing / zeroing the populated pixels from the primary data array that are not associated with the high density clusters, to form a secondary data array. Low density clusters are associated with noise, so it is beneficial to remove any contribution of these low density clusters to the image.

[0046] In an embodiment the pixels in the secondary data array are averaged or combined with one or more adjacent pixels in the secondary data array (binning), to form an updated secondary data array to replace the previous secondary data array. In an embodiment, the updated secondary data array is smaller than the previous secondary data array, optionally the updated secondary data array is reduced to 1 / 4 (quarter, e.g. 64 pixels reduced to 16 pixels), 1 / 9, 1 / 16, 1 / 25, 1 / 36, 1 / 49, 1 / 64, 1 / 81 or 1 / 100 of the pixels in the previous secondary data array. In an embodiment, the aspect ratio is preserved in binning. It is also possible to bin where the aspect ratio is not preserved, but this requires additional action / processing. In an embodiment, the secondary data array is 1024 x 1024, 512 x 512, 256 x 256, or 128 x 128 pixels. In an embodiment, the secondary data array is 512 x 512 or 256 x 256 pixels. In an embodiment, the position of each pixel in the secondary data array maps to a relative position in the primary data array. In an embodiment, the position of each pixel in the secondary data array maps to a relative position in the array of photoactive cells.

[0047] In an embodiment each high density cluster in the secondary data array is assigned to a single populated pixel in a tertiary data array. This is because ideally each high density cluster is associated with a single gamma ray event.

[0048] In an embodiment values of the pixels in the tertiary data array are averaged or combined with one or more adjacent pixels in the tertiary data array (binning), to form an updated tertiary data array to replace the previous tertiary data array. In an embodiment, the updated tertiary data array is smaller than the previous tertiary data array, optionally the updated tertiary data array is reduced to 1 / 4 (quarter, e.g. 64 pixels reduced to 16 pixels), 1 / 9, 1 / 16, 1 / 25, 1 / 36, 1 / 49, 1 / 64, 1 / 81 or 1 / 100 of the pixels in the previous tertiary data array. In an embodiment, the aspect ratio is preserved in binning. It is also possible to bin where the aspect ratio is not preserved, but this requires additional action / processing. In an embodiment, the tertiary data array is 1024 x 1024, 512 x 512, 256 x 256, or 128 x 128 pixels. In an embodiment, the tertiary data array is 512 x 512 or 256 x 256 pixels. In an embodiment, the tertiary data array is 256 x 256 pixels. In an embodiment, the primary, secondary and / or tertiary data array is symmetrical (e.g. 256 x 256 pixels). In an embodiment, the primary, secondary and / or tertiary data array is asymmetrical (e.g. 512 x 256 pixels). In an embodiment, the primary and secondary data arrays are 512 x 512 pixels, and the tertiary data array is 256 x 256 pixels. In an embodiment the position of each pixel in the tertiary data array maps to a relative position in the secondary data array, and where the position of the single populated pixel substantially corresponds to the centre of the assigning high density cluster. In an embodiment, the position of each pixel in the tertiary data array maps to a relative position in the primary data array. In an embodiment, the position of each pixel in the tertiary data array maps to a relative position in the array of photoactive cells.

[0049] In an embodiment a populated pixel in the tertiary data array corresponds to the detection of gamma ray photons by the gamma camera. In an embodiment, the pixel value of a populated pixel in the tertiary data array corresponds to the number of detected gamma ray photons. In an embodiment, the pixel value of a populated pixel in the tertiary data array corresponds to the number of detected gamma ray photons at that (relative) position. For example, the tertiary data array may record a single event at a unique X-Y location (a binary image); and there may be multiple hits in the frame, but where these are not at the same place.

[0050] In an embodiment, the detected gamma ray photons in the generated image are counted to give a total number of gamma ray photons in the generated image. This can be useful for quantifying the amount of radioisotope in the region of the subject that is within the generated image, and to smaller regions-of-interest within the generated image.

[0051] In an embodiment the populated pixels in the tertiary data array are used in forming the generated image. Since the tertiary data array relates to detected gamma ray events, these can be used to make an image, including the framed summed image. In an embodiment unpopulated pixels in the tertiary data array correspond to the absence of the detection of gamma ray photons by the gamma camera. In an embodiment, the unpopulated pixels in the tertiary data array correspond to the absence of the detection of gamma ray photons by the gamma camera at that (relative) position.

[0052] In an embodiment a kernel is used to identify the clusters in the primary data array.

[0053] In an embodiment, two kernels are used to identify the clusters in the primary data array, wherein the two kernels comprise a first and last kernel, wherein the first kernel operates on the primary data array to produce a refined primary data array, which is operated on by the last kernel to identify the clusters present in the refined primary data array. In an embodiment, two or more kernels are used to identify the clusters in the primary data array, wherein the two or more kernels comprise a first kernel and one or more subsequent kernels forming a sequence of kernels, wherein the subsequent kernels are arranged to operate on the output array from the kernel preceding it in the sequence of kernels, such that the final output array is a refined primary data array and the last kernel in the sequence of kernels is used to identify the clusters present in the refined primary data array. In an embodiment, two or more kernels are used to identify the clusters in the primary data array, wherein the two or more kernels comprise a first kernel and one or more subsequent kernels forming a sequence of kernels, wherein the subsequent kernels are arranged to operate on the output from the kernel preceding it in the sequence of kernels, wherein the first kernel operates on the primary data array to refine the primary data array, wherein the refined primary data array is operated by the sequence of kernels to further refine the primary data array, wherein the last kernel (in the sequence of kernels to operate on the refined primary data array) is used to identify the clusters present in the refined primary data array. In an embodiment, one or more kernels operate on the revised primary data array.

[0054] In an embodiment the clusters identified by the kernel are categorized as having a high density of populated pixels, or as having a low density of populated pixels. With reference to Figure 4(c), an illustrative kernel (the grid like structure) can be seen operating on a group of pixels (the pixels that fall under the kernel). Essentially the kernel makes an assessment of the group of pixels under the kernel, and decides by a rule system if the cluster meets a certain criteria or not, in this case high or low density. For example, if the kernel is using a median rule filter, it will decide if the median value of the pixels (the value of the middle pixel when the pixels are arranged in size order) under the kernel is 1 (populated) or less (unpopulated). If less than 1, the cluster is 'low density' and if 1 or more the cluster is 'high density'.

[0055] In an embodiment a low density cluster is a cluster wherein fewer than about 90%, 80%, 70%, 60%, 50%, 40%, 30%, 20% or 10% of the pixels under the kernel are populated pixels. In an embodiment a low density cluster is a cluster where fewer than about 50% of the pixels under the kernel are populated pixels. In an embodiment a high density cluster is a cluster where about at least 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80% or 90% (e.g. chose the value at the 90th percentile) of the pixels under the kernel are populated pixels. In an embodiment a high density cluster is a cluster where at least about 50% of the pixels under the kernel are populated pixels. In an embodiment a low density cluster is a cluster where the median value of the pixels under the kernel is 0. In an embodiment a high density cluster is a cluster where the median value of the pixels under the kernel is greater than 0.

[0056] In an embodiment the identifying of the clusters and the categorizing them into having a high or low density of populated pixels is done in a single step.

[0057] In an embodiment, the identifying of the clusters, categorizing them into having a high or low density of populated pixels, and excluding the low density clusters from contributing to the generated image is done in a single step. Advantageously, this is most efficient and can save on computer processing.

[0058] In an embodiment, the method bypasses the use of the secondary data array. In an embodiment, the method moves directly to the tertiary data array from the primary data array in a single step, whereby the clusters are identified and classified as high density clusters, and these high density clusters are centroided, e.g. to a single pixel corresponding to a single gamma ray photon detection event in a single step.

[0059] In an embodiment the kernel is an edge preserving filter. This might be understood for example with reference to the transformation shown in Figure 4(c) to Figure 4(d), where the edge of the large cluster is preserved. Advantageously, this efficiently segregates detection events.

[0060] In an embodiment the size and / or shape of the kernel is adjusted to improve the generated image. In an embodiment the size and / or shape of the kernel is optimised to maximise rejection of scattered gamma photon detection events, while minimising rejection of unscattered gamma photon detection events. A kernel can be adjusted to suit need, and may be adjusted to suit the radioisotope being used. This is because different isotopes have different characteristic energies and so will produce different sized 'splashes', and the sensitivity of the detector and any signal amplification in a particular device may result in smaller or larger splashes for a particular gamma photon energy. Therefore, a more energetic isotope might need a bigger kernel, or the rules used to discriminate may need to be changed depending on the details of the device design (e.g. the size of the kernel may be increased or decreased depending on the number of pixels in the frame, or the number of populated pixels under the kernel would need to be increased for a cluster to meet the criteria of being a high density cluster or decreased to avoid eliminating genuine clusters). In an embodiment, the kernel is adapted to detect substantially circular clusters corresponding to optical splashes.

[0061] In an embodiment, the kernel has a square, rectangular, diamond, hexagonal, irregular or circular shape. In an embodiment, the kernel has a plane (line) of symmetry. In an embodiment, the kernel has rotational symmetry. In an embodiment the size of the kernel is about the size of a high density cluster. Typically, this produces good results. In an embodiment the size of the kernel is adjusted to suit the peak energy of the gamma rays being used in the medical imaging. In an embodiment the size of the kernel is adjusted to suit the peak energy of the gamma rays being used in the medical imaging, and wherein the peak energy of the gamma rays results from gamma rays emitted from Tc-99m, 1-123, 1-131, TI-201LLU-177, In-111, Y-90, Sc-47, Ga-67, Cr-51, Sn-177m, Cu-67, Tm-167, Ru-97, Re-188, Au- 199, Pb-203, Ce-141, Co-57, F-18, Ga-68, C-ll, 0-15, N-13, Zr-89, Rb-82, Cu-64, Cd-109, Cs-131, 1-125, Am-241, Cd-109, Ag-109m, U-235, Pb-212, Re-186 and Ho-166.

[0062] In an embodiment, the kernel is adjusted to suit the peak energy (see first column below) of the gamma rays being used in the medical imaging, and wherein the peak energy of the gamma rays results from gamma rays emitted from the isotope used (see second column below):

[0063] Nominal Gamma energy (keV) Isotope

[0064] 22 Cd-109

[0065] 30.4 Cs-131

[0066] 35 1-125

[0067] 60 Am-241

[0068] 88 Cd-109 / Ag-109m

[0069] 68-82 TI-201

[0070] 113 Lu-177

[0071] 122 Co-57

[0072] 136 Re-186

[0073] 140 Tc-99m

[0074] 143 U-235

[0075] 155 Re-188

[0076] 159 1-123

[0077] 171 In-111

[0078] 208 Lu-177

[0079] 239 Pb-212

[0080] 245 In-111

[0081] 279 Pb-203

[0082] 284 1-131

[0083] 364 1-131

[0084] 511 F-18, Ga-68, C-ll, Cu-67

[0085] In an embodiment the size of the kernel is adjusted to suit the peak energy (see first column below) of the gamma rays being used in the medical imaging, and wherein the peak energy of the gamma rays results from gamma rays emitted from the isotope used (see second column below):

[0086] Nominal Gamma energy (keV) Isotope 22 Cd-109

[0087] 30.4 Cs-131

[0088] 35 1-125

[0089] 113 Lu-177

[0090] 122 Co-57

[0091] 140 Tc-99m

[0092] 155 Re-188

[0093] 159 1-123

[0094] 171 In-111

[0095] 208 Lu-177

[0096] 239 Pb-212

[0097] 245 In-111

[0098] 279 Pb-203

[0099] 284 1-131

[0100] 364 1-131

[0101] 511 F-18, Ga-68, C-ll, Cu-67

[0102] In an embodiment, the kernel is adjusted to suit the peak energy (see first column below) of the gamma rays being used in the medical imaging, and wherein the peak energy of the gamma rays results from gamma rays emitted from the isotope used (see second column below):

[0103] Nominal Gamma energy (keV) Isotope

[0104] 60 Am-241

[0105] 88 Cd-109 / Ag-109m 143 U-235

[0106] In an embodiment, the kernel is adjusted to suit the peak energy of between about 20 keV and 520 keV.

[0107] In an embodiment the kernel is sized and / or shaped to operate on between about 5 to 1000, optionally about 25 to 169, further optionally about 36 to 121 pixels in the data array that it is operating on. In an embodiment, the kernel is sized and / or shaped to operate on between about 2x2 to 15x15, optionally about 3x3 to 11x11, further optionally about 5x5 to 10x10 pixels in the data array that it is operating on. In an embodiment, the kernel is sized and / or shaped to operate on 0.001% to 0.2%, optionally 0.003% to 0.1%, further optionally 0.005% to 0.075% of the total number of pixels in the data array that it is operating on.

[0108] In an embodiment, the data array that is being operated on by the kernel is the primary data array. This is a good time to identify high density clusters. In an embodiment, the data array that is being operated on by the kernel is the secondary data array. In an embodiment, the data array that is being operated on by the kernel is the tertiary data array. In an embodiment, the kernel is sized and / or shaped to exclude gamma ray photons from contributing to the generated image which have an energy of less than about 20 keV, optionally less than about 50 keV, and further optionally less than about 100 keV. In an embodiment the kernel is sized and / or shaped to exclude gamma rays from contributing to the generated image which have a photon energy greater than about 550 keV, optionally greater than about 400 keV, and further optionally greater than about 250 keV. In an embodiment one kernel is used to exclude lower energy gamma ray photons, and different kernel is used to exclude higher energy gamma ray photons from contributing to the generated image. In an embodiment the kernel is sized and / or shaped not to exclude gamma rays from contributing to the generated image which have a photon energy equal to or greater than about 100 keV, optionally equal to or greater than about 150 keV, and further optionally equal to or greater than about 200 keV. In an embodiment the kernel is sized and / or shaped to include gamma rays to contribute to the generated image which have a peak photon energy equal to or greater than about 100 KeV, optionally equal to or greater than about 130 keV, and further optionally equal to or greater than about 140 keV. In an embodiment the kernel is sized and / or shaped not to exclude gamma rays having a peak energy in the range of about 50 to 550 keV, optionally about 120 to 165 keV, further optionally about 140 keV.

[0109] In an embodiment the gamma camera comprises a CCD, EMCCD or sCMOS chip.

[0110] In an embodiment, the gamma camera is calibrated prior to use (or periodically calibrated) with a uniform reference source. In an embodiment, the gamma camera is calibrated prior to use (or periodically calibrated) with a uniform reference source to exclude any artefacts present in the gamma camera, or that may arise from the method used to detect the gamma ray photons. It can be useful to remove background noise or perhaps correct for errors in the sensors by means of calibration before use.

[0111] In an embodiment the generated image is periodically refreshed. In an embodiment a snapshot of the generated image is periodically obtained. In an embodiment a time lapsed image of the generated image is obtained. The method of the invention can be used to make still images or live images, or time-weighted / averaged images.

[0112] There are many ways in which the method of the invention could be used with a gamma ray camera. Ideally, it should be used in such a way to reduce processing load, and hence lag or excess energy use. This might mean that some (or all) of the processing is done within the camera, for example using firmware, or some or all of the processing is fed into a secondary system (which could be a dedicated system) where it is processed - for example on a local computer, on a pen drive / dongle, or in the cloud. In an embodiment the method is carried out partly or wholly within the medical imaging device. In an embodiment the method is encoded to software which is partly or wholly held within the medical imaging device. In an embodiment the method is encoded to software which is stored on memory which is partly or wholly within the medical imaging device. In an embodiment the method is encoded to software which is stored on memory, and which can be partly or wholly installed on the medical imaging device. In an embodiment the method is encoded as a program / application ('app') that can be run partly or wholly on the medical imaging device. In an embodiment the method is encoded to software which is stored on portable memory (e.g. a pen drive / dongle). In an embodiment the method is encoded to software which is stored in the cloud. In an embodiment, the software will need to be installed. In an embodiment the software is self-executing and does not need to be installed. In an embodiment, the medical imaging device comprises a computer processor to carry out all, or part of, the method of any aspect or embodiment of the invention. In an embodiment, the gamma camera and computer processor are contained in the same housing. In an embodiment, the computer processor is firmware. In an embodiment, the medical imaging device is portable. In an embodiment, the medical imaging device is handheld. In an embodiment, the gamma camera and computer processor are not contained in the same housing but are connected by wired or wireless connection. In an embodiment the computer processor is in the cloud. In an embodiment the computer processor is in a desktop type computer or a handheld device such as a smart phone or tablet. In an embodiment, the gamma camera and second processor are not in the same housing. In an embodiment, the first processor and second processor are not in the same housing. In an embodiment, the first and second processors are not in the same housing and are connected by wired or wireless connection. In an embodiment, the second processor is in a desktop type computer or a handheld device such as a smart phone or tablet. In an embodiment, the gamma camera and first processor are in the same housing. In an embodiment, the gamma camera and first processor are in the same housing and the first processor is firmware. In an embodiment, the first processor and second processor are in the same housing. In an embodiment, the first processor and second processor are in the same housing and the processors are firmware in the gamma camera. In an embodiment, the gamma camera, first processor and second processor are in the same housing. In an embodiment, the first processor and second processor are in the same housing and are firmware in the gamma camera. In an embodiment, the gamma camera and second processor are not in the same housing. In an embodiment, the first processor and second processor are not in the same housing. In an embodiment, the first and second processors are not in the same housing and are connected by wired or wireless connection. In an embodiment, the second processor is in a desktop type computer or a handheld device such as a smart phone or tablet. In an embodiment a computer processor produces the primary, secondary and tertiary array. In an embodiment a first computer processor produces the primary array and a second computer processor produces the secondary and tertiary array. In an embodiment a first computer processor produces the primary and secondary array and a second computer processor produces the tertiary array. In an embodiment a first computer processor produces the primary, secondary and tertiary array and a second computer processor carries out any further processing as necessary. In an embodiment one computer processor produces the single frame images and another processor produces the framed summed images.

[0113] In an embodiment the generated image is used together with an output from another imaging modality such as CT, MRI, US, X-ray, fluorescence, and / or an optical camera. In an embodiment the medical imaging device defined in any aspect of the invention (and any embodiments thereof), also comprises an optical camera. In an embodiment the gamma camera defined in any aspect of the invention (and any embodiments thereof) is used together with an optical camera (capturing optical photons directly emanating from the subject) in the medical imaging device, and wherein the optical images and gamma images are superimposable. In an embodiment the gamma camera defined in any aspect of the invention (and any embodiments thereof) is used together with an optical camera in the medical imaging device, and wherein the optical images and gamma images are superimposable, and wherein the images are substantially free of parallax when the images are superimposed.

[0114] In an embodiment the generated image is used in medical imaging. In an embodiment the generated image is used in a method of medical diagnosis, or to aid in a medical diagnosis. In an embodiment the generated image is used in a method of medical treatment and / or monitoring.

[0115] In a further aspect there is provided a method of improving a generated image obtained from a gamma camera used in gamma ray medical imaging, the image improved by excluding low energy and scattered gamma rays from contributing to the generated image, the method comprising the steps of:

[0116] • the gamma camera detecting gamma rays;

[0117] • converting the detected gamma rays into one or more optical photons;

[0118] • detecting the resultant optical photons on a photoactive surface;

[0119] • rejecting gamma rays that produce a small optical splash on the photoactive surface from contributing to the generated image, these small splashes corresponding to low energy and scattered gamma rays. Gamma ray medical imaging devices are not suitable for imaging all forms of electromagnetic radiation such as infrared, ultraviolet and X-ray radiation types. In particular, although both gamma and X-ray radiation are types of ionising radiation, it should be understood that gamma rays are distinguished from X-rays in that gamma rays, like alpha and beta particles, originate from the radioactive decay of an unstable nucleus, whereas X-rays originate from electrons outside the nucleus.

[0120] So, sources of gamma rays originate from radioactive materials that decay over time. X-rays on the other hand are actively produced by X-ray tubes, in which electrons are accelerated in a vacuum and impacted on a metal plate. Gamma rays generally have higher energies than X-rays. X-rays typically have energies in the range lOOeV to 100,000eV (or 100 keV) whereas gamma rays generally have energies greater than about 100 keV. Furthermore, X-ray imaging is typically completed rapidly with image acquisition completed within seconds or, at most, 1 to 2 minutes. Scintigraphic images of gamma emitting radio isotopes typically take from about 5 to 40 minutes to acquire, presenting additional challenges in terms of imaging device design. This means that in practical terms devices tailored to gamma ray detection are not suitable for X-ray imaging, and vice versa. This is because X-ray detection / imaging systems have low efficiency for gamma ray detection, as the higher energy gamma rays tend to pass through the system without being detected. Conversely, detection / imaging systems used for gamma ray imaging are substantively unresponsive to the lower energy X-rays. In an embodiment the method is optimised for gamma ray detection. In an embodiment the method is not used to detect X-rays. In an embodiment the method detects ionising radiation with energies greater than about 30 keV. In an embodiment the method detects ionising radiation with energies greater than about 100 keV. In an embodiment the method is adapted to detect gamma rays emitted from a radio isotope source. In an embodiment the method is adapted to detect gamma rays emitted from a radio isotope source administered to a subject.

[0121] Herein disclosed is a device for use in imaging a subject using gamma rays emanating from the subject, the device comprising the means to run the method of the invention, or embodiments thereof.

[0122] Herein disclosed is a device for use in imaging a subject using gamma rays emanating from the subject, the device equipped with software and / or hardware to run the method of the invention, or embodiments thereof.

[0123] In a further aspect of the invention, there is provided a system comprising one or more devices arranged to run the method according to any aspect of the invention, and any embodiments thereof.

[0124] In an embodiment, the system comprises one or more of a: display; display monitor, support stand / frame, movable arm, power supply, battery, memory, Wi-Fi capability, Bluetooth capability, communication interface, and communication cables.

[0125] In an embodiment the system comprises one or more devices arranged to generate a 3D image.

[0126] In a further aspect of the invention, there is provided use of the method according to any aspect of the invention, and any embodiments thereof, to image a subject using gamma rays, wherein an imaging agent is administered to the subject. In an embodiment the imaging agent is a gamma ray emitting agent. In an embodiment the gamma ray emitting agent comprises one or more radioisotopes. In an embodiment the gamma ray emitting agent is selected from one or more of Tc-99m, 1-123, 1-131, Lu-177, In-111, Y-90, TI-201, Sc-47, Ga-67, Cr-51, Sn-177m, Cu-67, Tm-167, Ru-97, Re-188, Au-199, Pb-203, Ce-141, Co-57, F-18, Ga-68, C-ll, O- 15, N-13, Zr-89, Rb-82, Cu-64, Cd-109, Cs-131, 1-125, Am-241, Cd-109, Ag-109m, U-235, Pb-212, Re-186 and Ho-166. In an embodiment the gamma ray emitting agent is selected from one or more of Tc-99m, 1-123, 1-131, Lu-177, In-111, Y-90, Sc-47, Ga-67, Cr-51, Sn-177m, Cu-67, Tm-167, Ru-97, Re-188, Au- 199, Pb-203, Ce-141, Co-57. In an embodiment the gamma ray emitting agent accumulates in cells, tissue and / or an organ to be imaged. In an embodiment a fluorescent agent accumulates in a cell type, cells, tissue and / or an organ to be imaged. In an embodiment the cells, tissue and / or organ to be imaged is selected from one or more of: bladder, bone, blood, blood vessel, brain, colon, eye, gall bladder, heart, intestine, kidney, liver, lung, pancreas, skin, stomach, thyroid or parathyroid. In an embodiment the cells, tissue and / or an organ to be imaged comprises abnormal cell growth. In an embodiment the cells, tissue and / or an organ to be imaged comprise cancer. In an embodiment the cells, tissue and / or an organ is imaged performing one or more biological functions. In an embodiment the biological function comprises a physical function. In an embodiment the physical function comprises filling, emptying, contracting or relaxing. In an embodiment the physical function is imaged in real-time.

[0127] Herein disclosed is a method of analysis or diagnosis comprising the step of imaging a subject using the method according to the invention, or any embodiments thereof.

[0128] In an embodiment the method is used in a device or system that is portable and is brought to the subject to be imaged. In an embodiment the device or system is used together with an optical camera wherein the optical images and gamma images are superimposable, and wherein the images are substantially free of parallax when the images are superimposed.

[0129] Herein disclosed is a method of treatment or surgery comprising the step of imaging a subject during the treatment or surgery using the method of the invention, or any embodiments thereof.

[0130] In an embodiment the subject is treated with a gamma ray emitting agent which accumulates in tissue to be removed in the surgery, optionally the tissue comprises cancer.

[0131] Herein disclosed is a method of evaluating a treatment or surgery conducted on a subject, comprising the step of imaging the subject after the treatment or surgery using the method according to the invention, or any embodiments thereof.

[0132] In an embodiment the subject is human or animal or part or tissue thereof; or the subject may be nonhuman or non-animal and contain or be contaminated with a gamma ray emitting substance. In an embodiment the method is run in a device or system that is portable and is brought to the subject to be imaged.

[0133] In an embodiment the subject is human or animal or part or tissue thereof; or the subject may be nonhuman or non-animal and contain or be contaminated with a gamma ray emitting substance. In an embodiment the method is run in a device or system that is portable and is brought to the object / subject to be imaged. In a further aspect of the invention, there is provided the method according to any aspect of the invention, and any embodiments thereof, encoded to software or hardware.

[0134] Herein disclosed, the disclosure herein above, inclusive of the aspects and / or embodiments of the invention may be adapted / used for non-medical imaging, e.g. for imaging inanimate subject objects using both suitable gamma rays emanating from the subject, optionally having utility in radioactive waste management, identifying small localised leaks involving radioactivity, or in radiation protection. For example, suitable applications may include: finding / detecting / monitoring accidental spills of gamma ray emitting components (e.g. a spilled component used in a medical treatment, or a spilled component used in a laboratory); finding / detecting / monitoring / localising the possible inadvertent removal of gamma-ray emitting components from a store; finding / detecting / monitoring / localising spent / waste low grade gamma-ray emitting components used in connection with industry; finding / detecting / monitoring / localising possible inadvertent radioactive contamination; or monitoring the integrity of containment vessels holding / storing gamma emitting components used in connection with industry.

[0135] Embodiments or disclosures disclosed herein may be independently combined with any other embodiment, embodiments, aspect or aspects of the invention.

[0136] The present invention will now be further described with reference to the following non-limiting examples and the accompanying illustrative drawings, of which:

[0137] Brief description of the drawings

[0138] Figure 1 shows a schematic representation of a gamma ray medical imaging device with gamma ray photons emanating from a subject.

[0139] Figure 2 is the same as Figure 1, but wherein a scattered gamma ray photon is also shown.

[0140] Figure 3 shows a schematic representation which shows mapping of output voltages from Figure 2 as an energy distribution (with Compton scattering).

[0141] Figure 4 shows a schematic representation of an embodiment of the method of the invention.

[0142] Figure 5 shows a schematic representation of a framed summed image built up of single frame images.

[0143] Figure 6 shows a schematic representation of 'binning'.

[0144] Detailed description of the invention

[0145] Figure 1 shows a schematic representation of a gamma ray medical imaging device of the prior art with gamma ray photons emanating from a subject. A subject (10) has been treated with a radiation source (11) which is emitting gamma ray photons (y). The trajectory of the gamma ray is indicated by the dashed arrow (12). These gamma ray photons travel from the subject and are detected by a gamma ray detection system (20). First the emitted gamma ray photons pass through a collimator (21) in the detection system and it hits a gamma ray detector (22). The detector outputs a voltage (23), the voltage is proportional to the energy of the gamma ray photon detected by the detection system. Figure 2 is the same as Figure 1, but wherein a scattered (second) gamma ray photon is also shown. In this case the second unwanted gamma ray photon (y*) also enters through the collimator (21). In this case this happens when the second gamma ray photon is scattered. Scattering is depicted as the inflection point in the path (13) of the second gamma ray photon. The second gamma ray photon may be from background radiation. The gamma ray may also be from the radiation source (11) but where in this case the original trajectory (not shown) of the gamma ray photon has already been changed e.g. by a first scattering event. It should be noted that background radiation could enter the collimator without necessarily being scattered.

[0146] In this scenario, both the first and second gamma ray photons passes through the collimator (21). Both gamma ray photons hit the gamma ray detector (22). The detector outputs a first voltage (23) corresponding to the first gamma ray photon, and a second smaller voltage (24; striped peak) corresponding to the second (scattered) gamma ray photon. The voltage peaks are proportional to the energy of the gamma ray photons detected by the detection system. In this case the second gamma ray photon is less energetic than the first gamma ray photon, and so gives a lower output voltage.

[0147] Figure 3 shows a schematic representation which depicts the mapping of output voltages from Figure 2, as an energy distribution. It can be imagined that if the gamma ray detection system (20) was allowed to amass a number of counts over time, and the voltages were converted to energy, an energy distribution (30) could be obtained. This energy distribution would contain a high energy region (31) which would contain a contribution from the first voltage peak (23) and a low energy region (32; Compton scattering) which would contain a contribution from the second low voltage peak (24). An energy window is shown between the dashed vertical lines which are either side of the photopeak value (33).

[0148] Figure 4 shows a schematic representation of an embodiment of the method of the invention:

[0149] Figure 4(a) shows the image that would be produced by the gamma ray camera if the embodiment of the invention were not used. That is, the gamma ray camera has detected several detection events that could be associated with gamma ray photons. These detection events have been converted into optical photons in the gamma ray camera. The more energetic the detection event the more optical photons are produced (so a more energetic gamma ray photon produces a bigger 'splash' of optical photons). These optical photons have been detected on a photoactive surface (e.g. a CMOS sensor), and are converted into an output image / pixel array (e.g. the primary data array).

[0150] Figure 4(b) shows clusters in the primary data array as indicated by illustrative boxes.

[0151] Figure 4(c) shows the largest cluster sitting in a 9 x 9 pixel array. In this case the grid can be considered to be a 9 x 9 kernel operating on the primary data array.

[0152] Figure 4(d) shows that only a single large density cluster has been identified by the method using this kernel, with the rest of the pixels being set to zero / unoccupied (e.g. the secondary data array). In this case the kernel is an edge preserving filter.

[0153] Figure 4(e) shows the high density cluster being centroided (e.g. to give a tertiary data array). This single populated pixel in Figure 4(e) may be deemed to correspond to the detection of an unscattered gamma ray photon, having an X and Y coordinate in the array, and wherein the location of the populated pixel corresponds to the true point of origin of an unscattered gamma ray photon, and this may then be used in making an image. So, with reference to the overall transformation shown in Figure 4(a) to Figure 4(e), only one pixel contributes to the generated image (the rest of the pixels are empty / zeroed). For example, this could be a single frame image used to build up a frame summed image e.g. as shown in Figure 5.

[0154] Figure 5 shows a schematic representation of a framed summed image (50) built up of single frame images (41 to 44). That is, Figure 5 shows how frames 1 to 4 (single frame images) and indeed up to 'n' single frame images (where the rest of the single fame images are not shown in the figure) maybe combined to form the frame summed image (50) on the right. For example, in the schematic illustration, the single frames (41 to 'n') have been combined to show a representation of the buildup of radioactive isotope (e.g. Tc-99m) in the thyroid (51) of a patient. The darker areas indicate the most activity in the thyroid. The resultant frame summed image might be used to identify abnormal organ function.

[0155] Figure 6 shows a schematic representation of 'binning'. That is, in Figure 6 it can be seen how an 8 by 8 matrix of 64 pixels can be reduced to a 4 by 4 matrix of 16 pixels by using 2 by 2 binning:

[0156] Figure 6(a) shows an empty 64 pixel array.

[0157] Figure 6(b) shows how the pixels could be filled, for example the top left pixel has a value of 1.

[0158] Figure 6(c) shows the pixels may be subject to 2 by 2 binning, the grid showing which pixels are being binned.

[0159] Figure 6(d) shows the result of binning, for example the top left pixel now has a value of 3, corresponding to the sum of values of the pixels binned as indicated in Figure 6(c).

Claims

Claims1. A method of improving a generated image obtained from a gamma camera used in medical imaging, the image improved by excluding low energy and scattered gamma rays from contributing to the generated image, the method comprising the steps of:• the gamma camera detecting gamma ray photons;• converting the detected gamma ray photons into one or more optical photons;• detecting the resultant optical photons on a photoactive surface;• processing the detected optical photons into a primary data array comprising pixels, wherein the pixels map the detection of the optical photons on the photoactive surface, wherein populated pixels correspond to the detection of the optical photons on the photoactive surface;• identifying clusters of populated pixels in the primary data array;• determining the density of the populated pixels in the identified clusters;• categorizing the clusters into high density clusters (clusters having a high density of populated pixels) and into low density clusters (clusters having a low density of populated pixels);• excluding the low density clusters from contributing to the generated image, these low density clusters corresponding to low energy and scattered gamma rays.

2. The method of claim 1, wherein the method comprises assigning each high density cluster to the detection of a gamma ray photon, and using the gamma ray photon detection event to contribute to the generated image.

3. The method of claim 1 or 2, wherein the generated image comprises a single frame image, which may be combined with one or more additional single frame images to form a frame summed image.

4. The method of any one of the preceding claims, wherein the generated image is a frame summed image, the frame summed image being formed by combining two or more single frame images.

5. The method of any one of the preceding claims, wherein the step of excluding the low density clusters comprises the step of removing / zeroing the populated pixels from the primary data array that are not associated with the high density clusters, to form a secondary data array.

6. The method of claim 5, wherein the pixels in the secondary data array are averaged or combined with one or more adjacent pixels in the secondary data array (binning), to form an updated secondary data array to replace the previous secondary data array.

7. The method of claim 5 or 6, wherein each high density cluster in the secondary data array is assigned to a single populated pixel in a tertiary data array.

8. The method of claim 7, wherein pixels in the tertiary data array are averaged or combined with one or more adjacent pixels in the tertiary data array (binning), to form an updated tertiary data array to replace the previous tertiary data array.

9. The method of claim 7 or 8, wherein the position of each pixel in the tertiary data array maps to a relative position in the secondary data array, and where the position of the single populated pixel substantially corresponds to the centre of the assigning high density cluster.

10. The method of any one of claims 7 to 9, wherein a populated pixel in the tertiary data array corresponds to the detection of gamma ray photons by the gamma camera.

11. The method of any one of claims 7 to 10, wherein the populated pixels in the tertiary data array are used in forming the generated image.

12. The method of any one of claims 7 to 11, wherein unpopulated pixels in the tertiary data array correspond to the absence of the detection of gamma ray photons by the gamma camera.

13. The method of any one of the preceding claims, wherein a kernel is used to identify the clusters in the primary data array.

14. The method of claim 13, wherein the clusters identified by the kernel are categorized as having a high density of populated pixels, or as having a low density of populated pixels.

15. The method of claim 14, wherein the identifying of the clusters and the categorizing them into having a high or low density of populated pixels is done in a single step.

16. The method of any one of claims 13 to 15, wherein the kernel is an edge preserving filter.

17. The method of any one of claims 13 to 16, wherein the size and / or shape of the kernel is adjusted to improve the generated image.

18. The method of any one of claims 13 to 17, wherein the size and / or shape of the kernel is about the size of a high density cluster.

19. The method of any one of claims 13 to 18, wherein the size and / or shape of the kernel is adjusted to suit the peak energy of the gamma rays being used in the medical imaging.

20. The method of any one of claims 13 to 19, wherein the size and / or shape of the kernel is adjusted to suit the peak energy of the gamma rays being used in the medical imaging, and wherein the peak energy of the gamma rays results from gamma rays emitted from Tc-99m, 1-123, 1-131, Lu-177, In-111, 201-TI, Y-90, Sc-47, Ga-67, Cr-51, Sn-177m, Cu-67, Tm-167, Ru-97,Re-188, Au-199, Pb-203, Ce-141, Co-57, F-18, Ga-68, C-ll, 0-15, N-13, Zr-89, Rb-82, Cu-64, Cd-109, Cs-131, 1-125, Am-241, Cd-109, Ag-109m, U-235, Pb-212, Re-186 and Ho-166.

21. The method of any one of claims 13 to 20, wherein the kernel is sized and / or shape to operate on between about 5 to 1000, optionally about 25 to 169, further optionally about 36 to 81 pixels in the data array that it is operating on.

22. The method of any one of claims 13 to 21, wherein the kernel is sized and / or shape to operate on between about 2x2 to 15x15, optionally about 3x3 to 10x10, further optionally about 5x5 to 9x9 pixels in the data array that it is operating on.

23. The method of any one of claims 13 to 22, wherein the kernel is sized and / or shape to operate on 0.001% to 0.2%, optionally 0.003% to 0.1%, further optionally 0.005% to 0.075% of the total number of pixels in the data array that it is operating on.

24. The method of any one of claims 13 to 23, wherein the kernel is sized and / or shape to exclude gamma ray photons from contributing to the generated image which have an energy of less than about 20 keV, optionally less than about 50 keV, and further optionally less than about 100 keV.

25. A method of improving a generated image obtained from a gamma camera used in gamma ray medical imaging, the image improved by excluding low energy and scattered gamma rays from contributing to the generated image, the method comprising the steps of:• the gamma camera detecting gamma rays;• converting the detected gamma rays into one or more optical photons;• detecting the resultant optical photons on a photoactive surface;• rejecting gamma rays that produce a small optical splash on the photoactive surface from contributing to the generated image, these small splashes corresponding to low energy and scattered gamma rays.

26. The method of claim 25, wherein the splashes are substantially circular.

27. The method according to any one of claims 1 to 26, wherein the method is encoded to software or hardware.

28. A system or device arranged to run the method defined in any one of claims 1 to 27.

29. The system or device of claim 28 which is equipped with software and / or hardware to run the method defined in any one of claims 1 to 27.

30. The system or device according to claim 28 or 29 which comprises one or more of a: display; display monitor, support stand / frame, movable arm, power supply, battery, memory, Wi-Fi capability, Bluetooth capability, communication interface, and communication cables.

31. The system or device according to any one of claims 28 to 30, which is portable and optionally is handheld.

32. The system or device according to any one of claims 28 to 31, which is arranged to generate a 3D image.

33. The system or device according to any one of claims 28 to 32, which comprises an optical camera, wherein the optical images and gamma images generated are superimposable, and wherein the images are substantially free of parallax when the images are superimposed.

34. Use of the method, system or device according to any one of claims 1 to 33 to image a subject using gamma rays, wherein an imaging agent is administered to the subject.

35. Use of claim 34, wherein the imaging agent is selected from one or more of Tc-99m, 1-123, 1-131, Lu-177, In-111, Y-90, TI-201, Sc-47, Ga-67, Cr-51, Sn-177m, Cu-67, Tm-167, Ru-97, Re- 188, Au-199, Pb-203, Ce-141, Co-57, F-18, Ga-68, C-ll, 0-15, N-13, Zr-89, Rb-82, Cu-64, Cd- 109, Cs-131, 1-125, Am-241, Cd-109, Ag-109m, U-235, Pb-212, Re-186 and Ho-166.

36. The use according to claim 34 or 35, wherein the imaging agent accumulates in cells, tissue and / or an organ to be imaged, optionally the cells, tissue and / or organ to be imaged is selected from one or more of: bladder, bone, blood, blood vessel, brain, colon, eye, gall bladder, heart, intestine, kidney, liver, lung, pancreas, skin, stomach, thyroid or parathyroid.

37. A method of analysis or diagnosis comprising the step of imaging a subject using the method, system or device according to any one of claims 1 to 36.

38. A method of treatment or surgery comprising the step of imaging a subject during the treatment or surgery using the method, system or device according to any one of claims 1 to 37.

39. A method of evaluating a treatment or surgery conducted on a subject, comprising the step of imaging the subject after the treatment or surgery using the method, system or device according to any one of claims 1 to 38.

40. A method of evaluating a treatment or surgery conducted on a subject of claim 39, wherein the subject is human or animal, or part or tissue removed therefrom.

Images