Method and apparatus for imaging an object

Methods and equipment for object imaging Förfaranden och anordning vid fotografering av ett obj

The invention relates to methods for imaging an object using electromagnetic radiation, in which radiation reflected and / or transmitted by the object is detected by a sensor system providing an electrical signal containing image information of the object, the sensor system comprising a semiconductor sensor consisting of a plurality of successive and parallel .

The invention further relates to devices for imaging an object, the device comprising a device for applying electromagnetic radiation to the object to be imaged, the device further comprising a sensor system comprising a semico nductor sensor composed of a plurality of successive and parallel pixels, and the device further comprising devices for obtaining from the sensor system the electrical signal is detected and / or stored.

The present invention generally relates to the imaging of an object by electromagnetic radiation. In particular, the invention relates to digital imaging using a CCD sensor system.

The methods and devices according to the invention are intended to be applied in particular in medical X-ray Imaging, in particular in dental Panoramic Imaging, mammography Imaging or X-ray Imaging in general.

It is already known per se to use digital Imaging also in medical X-ray technology, where it has certain clear advantages over the use of a radiation-sensitive film to be developed. These benefits include the benefits of viewing, processing, storing and remotely transferring digitally captured and stored images, which will become even more pronounced in the future a s the healthcare and hospital system increasingly moves to digital technology in general and also in the processing of X-rays and the like. Additional benefits of digital imaging also include a reduction in the need for retouching as well as a reduction in the radiation dose received by the patient, as the semiconductor sensors used in digital imaging are more sensitive than analog films. It is therefore a main object of the present invention to take said development an essential step forward.

Semiconductor sensors for digital imaging are larger radiation-sensitive surfaces formed of small pixels, i.e. pixels, and in the extreme case are a single-row segment sensor. The electromagnetic radiation absorbed in the region of the pixels, such as light, infrared or X-rays, forms an electric charge proportional to the radiation (quantum amount and energy) absorbed in the pixels. In this case, the generation of the electric charge occurs as a function of time, i.e. the pixel integrates into its area the electric charge generated during the exposure time, so by changing the integration time, the strength of the generated pixel signal can in principle be adjusted. It should be noted, however, that the variation of the integration time does not change the sensitivity of the sensor.

The most well-known sensor types most commonly used in digital imaging are the diode array and the CCD sensor. CCD sensors will first be described in more detail below, as the present invention is preferably not suitable as such for use with known diode arrays.

In known CCD sensors, the electric charge is held in place in the pixel region during integration by an electric field which can be generated either by a 'conductor' placed in the center of the element at a suitable potential, which provides an electro25 or, according to the latest practice, by permanent charge fields copied to the sensor edges. form a potential barrier between pixels. The use of doped charge fields reduces the capacity of the pixel, i.e. they are not able to maintain as much charge as the potential wells, but they ‘overflow’ faster to their own charge in neighboring elements. Em. however, the charge fields have their own advantages, as the voltage required to form the potential well causes the so-called dark current, i.e. a charge leaking into the pixel that does not come from the actual image signal97665 list. As practical numerical values, it should be mentioned that the capacity of a known 'ordinary' pixel well is about 700,000 electrons while a 'doped' pixel can only hold about 400,000 electrons. The dark current at room temperature for a ‘normal’ pixel is about 30,000 electrons per second, while for a ‘doped’ it is only about 100 electrons per second. Prior to the development of doping, an attempt was made to control the dark current by cooling the sensor to about -40 ° C, which is difficult and expensive to implement and causes freezing and other problems.

In previously known CCD sensors, the reading of the image takes place by transferring the electrical charge accumulated in the pixel region of the sensor after the integration time away from the pixel well. controlled by clock signals, at which point the pixel immediately begins to collect the next charge representing the ‘image’. In the known TDI (Time Delay Integration) method, the transferred charge is left in the neighboring pixel. In the known FT (Frame Transfer) method, the charge is further quickly transferred to a radiation-protected area for the actual readout.

In the output of the image information detected by the CCD sensor, the charges of the outermost pixel row of the image area of ​​the TDI sensor and the protected area of ​​the FT sensor are first transferred to the so-called radiation-protected area. to the receiver register, which is an entire sensor-long row of charge wells in which the charges transferred from each pixel column are kept separate and transferred by the receiver clock signals and controlled one at a time to the sensor output well. Output20 The well is always emptied first and then a charge of one pixel is transferred to it, the magnitude of which is shown at the output of the sensor as a voltage signal containing the sensor pixel image information. As a practical numerical value, the voltage level of the output signal of the CCD sensor is of the order of -3 μν / electron, ie a charge of 500,000 electrons generates a sensor output voltage of -1.5 V.

As is well known, TDI imaging technology is applied to the imaging of a moving subject with maximum sensitivity with CCD sensors. Instead of shooting by a method known in the photographic technique with a sufficiently short exposure time to eliminate motion blur, the subject in the TDI imaging technique is allowed to move past the sensor while the subject's charge image is transferred to the sensor at the same speed. In this way, the image information generated by the object can be collected for a longer period of time, i.e. the sensitivity of the sensor can be easily multiplied. In order for a sensor with, for example, 100 pixels in the direction of movement of the subject to capture an image at maximum resolution using the normal shooting method, the shooting time should be so short that the image of the subject on the sensor does not move more than one pixel. In the known TDI method, the object is monitored along the entire length of the sensor, i.e. in this case it forms an image with a distance of 100 pixels, whereby the measured signal is 100 times, assuming that the object's speed remains constant during this time. Indeed, one of the basic principles that enables TDI imaging is precisely that the instantaneous movement speed of an object must be known in order to be able to track it successfully.

In the past, CCD image sensors and their control electronics have always been optimally designed for a specific application and purpose, and current sensor systems as such are not suitable for other, not always similar, uses. The development of new CCD sensors is very expensive and time consuming. Typically, the development of one new type of CCD sensor takes at least another year and the cost is 5 million. In the order of magnitude of the FIM. The development of control electronics for a new type of CCD sensor takes the same amount of time and increases costs by at least 1 million. FIM: 11a. Against this background, it is understandable that because the development of a new type of CCD sensor requires an application that pays for the development costs in a reasonable time, applications requiring smaller amounts of CCD sensors cannot be implemented at any realistic cost. On the other hand, new CCD sensors are designed for applications where volumes are sufficient, but also

The cost share of CCD sensors becomes relatively high when using the prior art.

Since with the prior art it has not been possible to modify the CCD sensors and their control electronics to be optimal for the respective subject, the imaging has had to be performed under the conditions of the available devices and their CCD sensors. In known imaging systems, the CCD sensor and its control electronics are either dimensioned for the highest required resolution or sensitivity, or compromise solutions have been made that are satisfied with an average satisfactory result by compromising both resolution and sensitivity. Adjusting the resolution and sensitivity of the CCD sensor to optimal for each imaging target is of particular importance in medical X-ray imaging, which must seek to provide adequate imaging resolution for the particular purpose while keeping the patient's radiation dose to a minimum.

As is known per se, it is known in CCD sensors to change the imaging resolution by combining pixels, i.e. 'binning', but this procedure has only been used standard in each application, and the freely controllable configuration of CCD sensors to change the 'binning' ratio to set the imaging resolution and sensitivity is not optimized

Imaging methods are already known in which the amount of data produced by the sensor is processed in digital form in the computer's memory, for example by summing adjacent memory locations together, whereby the end result is in principle similar to the above-mentioned 'binning'.

It is already known to connect a dental panoramic X-ray imaging device to a skull imaging stand, and the digital implementation of such a device combination is shown e.g. EP patent application 0 634 671 A1.

With regard to the state of the art incidental to the present invention, reference is made, by way of example, to the following patent and application publications: U.S. Pat. No. 5,426,685, U.S. Pat. No. 5,289,520,

US 4,188,537, US 5,365,562 and US 5,216,250.

It is an object of the present invention to further develop an imaging technique, and in particular a medical X-ray imaging technique, in which CCD sensors and digital imaging are applied so that the problems discussed above are substantially solved and the drawbacks eliminated.

It is a particular object of the invention to further develop imaging with CCD sensors so as to provide an even wider range of CCD sensor dynamics. This is important in applications where increasing the sensitivity of the CCD sensor at the expense of imaging resolution provides benefits. Such applications are, in particular, medical X-ray techniques, in which the radiation dose must be kept to a minimum and at the same time the exposure times must be as short as possible in order to minimize the blur caused by the patient's movement.

In order to achieve these and later objects, the first method of the invention is mainly characterized in that the sensor system is configured with control by combining parallel and / or successive sensor elements sensitive to the radiation of the sensor, i.e. 'binning' them into sensor element combinations comprising several pixels, that the charges containing the image information from the different pixels of said sensor element combinations are immediately combined in the charge form and the electrical signal thus obtained is further passed and that said combining of the pixels is performed by external control to mutually optimize the resolution and sensor sensitivity.

The second method of the invention is mainly characterized in that the method uses TDI imaging technique and arranges mutual motion between the object to be imaged and the sensor, that the image in the charged form formed by the object to be transferred to the CCD sensor is transferred at the same speed and said image information is collected for a longer time and sensor sensitivity. grown, that the physical pixel size of the sensor is selected to be smaller than the required maximum sensor resolution and that the blur caused by the relative motion of the subject and the sensor is reduced by shifting charge elements according to the physical pixel size of the CCD sensor so that the subject and CCD sensor are monitored with optimal accuracy.

The second method of the invention can be applied independently, but most preferably in connection with the first method of the invention, which is the main invention in this application.

The first device according to the invention, in turn, is mainly characterized in that the device comprises a sensor system configuration system by means of which parallel and / or successive image elements sensitive to the sensor radiation of the sensor system can be combined or "binnable" that the charges containing image information from the various elements of said combinations are immediately combinable in the form of charge.

The second device according to the invention is mainly characterized in that the device comprises a CCD sensor and its control system and devices for tracing between the object to be imaged and said sensor system, said control system being adapted to transfer to the CCD sensor a charge image of the object keskinäisliike, that the physical pixel size of the sensor is selected to be smaller than the required maximum sensor resolution and that the blur caused by the subject and the sensor relative motion is reduced by shifting charge elements according to the CCD pixel's physical pixel size, optimally tracking the subject and CCD sensor.

The second method and apparatus of the invention may be applied independently, but most preferably it is applied in synergistic combination with the first method and apparatus of the invention.

In applying the first method and apparatus of the invention, the pixel charges of the CCD sensor are summed by the sensor itself in still charge form before they are converted into a sensor output signal, amplified and converted to digital form. Thus, thanks to the invention, no noise or similar error due to summation is generated in the image signal, as is the case in the solutions according to the prior art.

By means of the first method and device of the invention, the properties of the CCD sensor and its control electronics can be easily configured, i.e. freely changed within predeterminable limits by software control to the optimum configuration required for each application, the same CCD sensor and control electronics can be used in several substantially different applications. Thus, the usage rates of that CCD sensor are increased and the development cost per application remains more reasonable. On the other hand, when the same CCD sensor and its control electronics can be used in one application for different purposes and in different modes of operation, the need for several different CCD sensors is eliminated, making the total cost of each application substantially more reasonable than the prior art.

In addition to saving sensor costs, the invention achieves the essential advantage that, in addition to the resolution of the imaging, the sensitivity / resolution control can affect the radiation dose received by the patient, in which respect compromises have previously had to be made.

In a preferred embodiment of the first method and device of the invention, the pixels can be combined, i.e. 'binnata', into larger entities both horizontally and vertically. For the sake of clarity and simplicity, the horizontal and vertical orientation of the sensor element field has been discussed above and below, which by no means literally refers to the ground attraction, but horizontally refers to the direction of the sensor element rows in the imaging plane where the charges are shifted. a direction perpendicular to the horizontal, which is, for example, in the case of segment sensors, their longitudinal direction. The allocations are combined horizontally by transferring the assignment of several pixel rows to the receiver register before the receiver register reservations are read out of the sensor. In this case, the allocations of several rows of pixels are 'poured into the same cup', i.e. they are summed. The vertical 'binning' of the sensor, i.e. the summing of the charges, is performed by emptying the output well and then transferring the contents of the charge wells of several receiver registers there before measuring the output signal of the sensor.

The invention can be applied to all kinds of imaging by means of a CCD sensor, where the required dynamic range is wide and where increasing the sensitivity at the expense of the imaging resolution or vice versa, a benefit is achieved. The most typical of these applications is X-ray imaging, where exposure times must be kept as short as possible to minimize blur caused by patient movement, while the beam dose must be kept to a minimum. In X-ray imaging, resolution requirements vary widely depending on what is observed and it is generally accustomed in the art to increase the beam dose as the required resolution increases. The implementation according to the invention can therefore cover most of the different aspects of X-ray imaging without substantially changing the construction of the sensor system without utilizing the free software configuration of the CCD sensor system according to the invention. Particularly preferred applications of the invention are mammography imaging for both magnification and contact imaging, as well as combined dental panorama and skull imaging devices. The scope of the invention can be extended to various ward x-ray devices and lung imaging devices.

In the first method and apparatus of the invention, the control electronics of the CCD sensor are constructed to operate according to the input signals tracked therein, thereby determining the 'binning' ratio of the pixels of the sensor vertically and horizontally at any time during imaging, but preferably at the beginning of imaging.

The invention is preferably applied in a dental panorama-skull imaging device combination so that the image resolution is user-selectable, whereby overviews can be taken at a lower resolution and beam dose and, if necessary, images from critical points can be taken at a higher resolution. For skull images, a resolution of 2 lp / mm is completely sufficient and the image is obtained with a beam dose of 15% compared to the resolution required in panoramic imaging, when the resolution is, according to the invention, user-adjustable. In practice, the focal length used in cranial imaging is more than three times (1700/480 mm) compared to dental panoramic imaging, so that the beam dose received by the sensor, decreasing in the square of the distance, is only about 8%. Therefore, with sensor resolution for dental panoramic imaging, it is almost impossible to take skull images because it would require an exposure time of tens of seconds during which the patient to be imaged must be completely immobile. In practice, panoramic images are taken with a TDI scan at a speed of approx. 20 mm / s. When the skull imaging scan is performed at a length of 240 mm, with the same sensor resolution and the same imaging parameters as the dental panorama imaging, it would take 12 s * 12.5 = 150 s, i.e. two and a half minutes, at this speed.

According to a preferred embodiment of the invention, the magnification of the mammography is performed as a contact image with a smaller pixel size than the normal mammography image. In this case, a significantly better resolution is achieved than in the previously known magnification description when using the pixel size of the fixed sensor, as will be explained in more detail later. In addition, in this embodiment of the invention, a simpler and less expensive X-ray source is required, where only one focal size is required. Also advantageous are shorter exposure times and a reduced risk of blurring the image due to the patient's movements, as well as a reduction in the radiation dose received by the patient. When, in this embodiment of the invention, magnification images are taken with a smaller pixel size and using contact shooting, magnification images can be taken at the size of the entire image area, if necessary. In the prior art, in which the object is brought closer to the X-ray tube, for example, with a magnification factor of 2, it forms an image on the imaging device so large that only a quarter of it can be imaged. These effects and advantages of the invention will be described in more detail later.

In the following, the invention will be described in detail with reference to some application examples of the invention shown in the figures of the accompanying drawing, to the details of which the invention is in no way narrowly limited.

Figure 1 schematically shows a CCD sensor system according to the invention and its control electronics applied to TDI-type imaging.

Figure 2 shows a CCD sensor system according to the invention and its control electronics in FT-type imaging according to Figure 1.

Figure 3 shows a typical output signal of a CCD sensor system according to the invention.

Figure 4 shows a circuit for measuring the zero level per pixel of a CCD sensor.

Figure 5 shows an example of the output signal of a CCD sensor system according to the invention in the vertical connection of element charges.

Figure 6 shows a preferred embodiment of the invention in which full resolution image tracking is applied in TDI imaging.

Fig. 7 shows a signal diagram of the TDI imaging mode according to Fig. 6.

Figure 8 shows a signal diagram in a preferred embodiment of the invention using a receiver register 'ditheration' method to eliminate dark current.

Figure 9 shows the control of some 'ditheration' clock signals typical of a three-phase sensor to the charge transfer mode.

Figure 10 is a schematic axonometric view of an application of the invention in dental panoramic imaging.

Figure 11 is a schematic axonometric view of an application of the invention to a mammography imaging device.

Figure 12 is a schematic axonometric view of an embodiment of the invention in connection with a lung transillumination device.

Figure 13 shows mainly a block diagram of a control system of a sensor system according to the invention.

Fig. 14 is a schematic axonometric view of a dental panoramic device with a skull imaging stand in a skull imaging mode as an application environment of the invention.

Figure 15A schematically illustrates a mammography imaging of the so-called a description of the contact.

Fig. 15B shows in a similar manner to Fig. 15A a mammography description of the so-called suurennuskuvauksena.

Figure 16 shows plots of modulation transfer function (MTF) caused by the finite size of the focal point in mammography imaging in various applications.

Fig. 17 shows the resolution of different imaging means in a manner similar to Fig. 16.

Fig. 18 shows, in a manner similar to Figs. 16 and 17, the resolution curves of the entire imaging system in different applications.

Figure 1 schematically shows the structure and operation of a TDI-type CCD sensor system and its control electronics. In the sensor system, the control electronics 17 are connected to the pixel field 10 of the CCD sensor with clock signals IOx, typically two to four (2-4), and to the receiver register 12 with clock signals ROx, which are also generally two to four (2-4). The control electronics 17 have a signal OR with which it clears the output well 13 of the CCD sensor. The charge of the output well 13 is converted to the output signal of the sensor by an output amplifier 14 and then passed to a zero level scrambling stage 15 which sets the zero reference level of each pixel field 10 by the control electronics 17 control signal CDS, as described in more detail below. After zero level equalization, the signal is passed to an analog-to-digital converter 16, which converts the image signal to digital form in sync with the SH signal provided by the control electronics 17. The digital data DI is further transmitted for storage and processing by means of the control signal LD ​​produced by the control electronics 17.

Fig. 2 schematically shows the structure and operation of the FT-type CCD sensor system according to the invention and its control electronics, which is similar to the TDI-type solution according to Fig. 1 except that the pixel field 10 to which the control electronics 17 are associated with clock signals AOx and receiver register Between the 12, a storage area 11 the s ize of a pixel field is traced, to which the control electronics 17 are connected by clock signals IOx.

The operation of the sensor system described below is mainly related to the operation of a TDI-type sensor and the FT-type sensor differs in principle only in that the transfer of the pixel field 11 storage to the receiver register 12 is always preceded by the pixel field 10 charge transfer through. In addition, the entire image area of ​​an FT-type sensor is always read completely after the integration time, and not line by line, as is usually the case with a TDI-type sensor.

The CCD sensor and its control known per se operate in accordance with the following description, with the exception of the new functional and structural features described in connection with pixel merging, which are essential for the present invention.

The CCD sensor of Figures 1 or 2, when Imaging, has a sensor for the elementary field 10; 10.11 Any accumulated charges must be removed and its pixel wells must be emptied. This is done by moving the pixel field 10 by the start signal START of the control electronics 17; 11 allocations by means of clock signals IOx column by column to the receiver register 12. The charges transferred to the receiver register 12 are in turn transferred one by one under the control of the clock signals ROx to the output well 13, which is emptied by means of the signal OR always before the transfer of each charge. The operation should be repeated a few times if there is reason to suspect that the sensor will not be completely discharged in one go, depending on how many charges have been accumulated in the pixel wells. The recording begins when image information, i.e. the charge produced by the radiation impinging on the sensor, begins to accumulate in the pixel wells of the emptied sensor. When the integration time set by the control electronics 17 has elapsed or alternatively synchronized with the external SYNC signal, the pixel field 10 is transferred; 11 single column allocations to the receiver register 12 and from there on to the outlet well 13 as described above. Now, unlike the previous one, each charge of the pixel well 13 is measured, i.e. by means of the CDS signal given by the control electronics 17, first the zero level of each pixel charge is corrected and then

Controlled by the SH signal, the analog signal is converted to digital by the converter 16, and when the conversion is completed, it is further recorded by the signal LD.

Figure 3 shows a typical CCD sensor output signal. When the output well 13 is emptied by means of the © R signal, a large pulse © RO is displayed at the sensor output, after which the sensor output signal settles to a certain level at time Z0, which now represents an empty output well. This zero level per pixel is measured by means of a CDS signal, for example by means of the connection according to Figure 4. Here, the CDS signal closes the switch 15k shortly before the moment Z0, when the output end of the capacitor 15c is grounded and the zero-level bottom signal present at the output of the sensor output amplifier 14 is charged to the capacitor 15c. At time Z0, the switch 15k is opened and the pixel charge now transferred to the sensor output well 13 is displayed as a signal level, which is measured at time SO synchronized by the signal SH. Although the zero level at time Z0 and Z1 varies, the circuit shown in Fig. 4 always measures the difference in signal levels at times S0-Z0, S1-Z1, ... i.e. the actual sensor signal. The operation is already known in these respects and is referred to as

Correlated Double Sampling.

When the pixel field 10; 10.11 pixels are combined horizontally according to the invention, in its simplest form, so that the pixel field 10; 10.11 column allocations are transferred to the recipient register 12 by more than one column before initiating the transfer of the serial register 12. In this case, the charges of several columns are combined, representing the number of charges accumulated in the area of ​​as many pixels as they have been summed in the saq register 12. Since the transmission efficiency of the CCD sensor is typically 99.9999%, this summation of charges can be performed with near-electron accuracy, which would be impossible when performed outside the CCD sensor, taking into account the required zero level per pixel, gain and other possible disturbances.

The control electronics 17 according to the invention are tracked by input signals HBx, by means of which it transfers the number of column charges indicated by them to the log register 12 before starting its charge transfer. The invention is characterized in that its control electronics 17 perform each pixel column transfer according to the number of columns indicated by the input signals HBx in each case, i.e. according to the sensor configuration.

Pixel field 10; The charges of the 11 pixels are combined vertically according to the invention in their simplest way, so that the charges of the log register 12 are transferred to the emptied output well 13 more than the field 10; 11 one item before measuring the output signal. In this case, the charges of several elements are combined, i.e. summed up, representing the amount of charge accumulated in the area of ​​as many pixels as has been summed in the output well 13. Since the transmission efficiency of the CCD sensor is typically 99.9999%, the summation in this case can be performed with almost electron accuracy, which would be impossible to perform outside the sensor, taking into account the necessary zero-level measurements per pixel, gain noise and possible other disturbances.

Figure 5 shows an example of the output signal and control signals of a CCD sensor system according to the invention in the vertical connection of element charges. After the discharge pulse ORO of the output well 13, the zero level of the output signal at time Z0 is measured by means of the CDS signal. The source wells 13 are then summed with the element charges at times Sn-2, Sn-1 and Sn, whereby the measurement of the image signal is performed only by means of the SH signal and the output well 13 is emptied for the next measurement.

The control electronics 17 according to the invention are tracked by the input signals VBx, as determined by it, according to the sensor configuration, it transfers the number of element charges indicated by the input signals VBx to the output well 13 before detecting the image signal. The invention is characterized in that its control electronics 17 perform the charge transfer of each individual receiver register and its signal measurements according to the number of 'binning' elements indicated by the VBx signals at the beginning of the charge transfer of said receiver register and the sensor configuration. If the length of the receiver register is not evenly divisible by the given number of 'binning' items, the control electronics 17 directs the last image signal to sum up the number of empty item charges indicated by the remainder so that row synchronization is not lost.

The implementation of the present invention allows the pixels of the CCD sensor to be freely combined by software control into entities set separately in the horizontal and vertical directions according to the configuration of the sensor. When a sensor with a sufficiently small pixel size is selected as the CCD sensor and both the control and measurement electronics are implemented so that they can utilize the full resolution in the shortest time required, all selected sensor pixel sizes and coarser resolutions pixel size in multiples according to the sensor configuration.

The following table shows some practical examples of the main imaging features of X-ray equipment implemented with a 30 gm x 30 gm pixel CCD sensor: In the following table, BIN stands for sensor 'binning ratio', Pix eff 'bundled' square pixel dimension in gm per millimeter, and lp / mm .

<td>bin</td><td>Pix eff</td><td>lp / mm</td><td>Rel. the radiation dose</td><td>Device</td>

<td>LXL</td><td> 30</td><td> 17</td><td> 100</td><td>Suurennusmammografia</td>

<td>2x2</td><td> 60</td><td> 8,5</td><td> 25</td><td>mammography</td>

<td>3x3</td><td> 90</td><td> 5,6</td><td> 11</td><td>Hammaspanoraama</td>

<td>4x4</td><td> 120</td><td> 4,2</td><td> 6,3</td><td>Hammaspanoraama</td>

<td>5x5</td><td> 150</td><td> 3,3</td><td> 4,0</td><td>Hammaspanoraama</td>

<td>6x6</td><td> 180</td><td> 2,7</td><td> 2,8</td><td></td>

<td>7x7</td><td> 210</td><td> 2,4</td><td> 2,0</td><td>skull Description</td>

<td>8x8</td><td> 240</td><td> 2,0</td><td> 1,6</td><td>skull Description</td>

It is clear from the above that at least three types of imaging systems can advantageously be implemented with one and the same CCD sensor system, using one and the same CCD sensor system and its control electronics 17. With the software configuration of the sensor according to the invention, both ordinary and magnifying images can be taken with a mammography device according to the invention without having to resort to separate magnification requiring mechanical devices, which tends to substantially reduce the manufacturing costs of the device and facilitate its use.

Applying the TDI imaging technique described in the preamble of the application in connection with the invention, the object to be imaged moves past the image-generating CCD sensor at a known speed at the control electronics 17 represented by the SYNC signal in Fig. 1 and the object direction of travel. When the charges of the pixel field 10 are transferred at substantially the same speed as the image formed by the moving object on the sensor, a sharp image is created of the moving object.

It should now be noted that due to some physical factors, such as inter-pixel doping, the pixel charges of the sensor can only be shifted by a pixel dimension at a time, i.e., by distances of a pixel or its multiple. The transfer of charges from one pixel to another by the sensor must be performed at a speed substantially higher than the motion speed of the object, so that the set of charge elements transferred to the receiver register is recorded to be transferred from the sensor, converted and stored before the next transfer. However, when the object being photographed usually moves at a constant speed rather than jerking as the charges move in the sensor, this asynchrony of the motion speeds results in a motion blur that is at worst half the pixel's direction of motion. Fig. 7 shows the sensor integration time t1, after which, synchronized with the signal SYNC, the pixel charges are transferred by means of the clock signals IOx in one sensor column, after which the receiver register 12 is read at Ro by means of the clock signals Röx.

In the case shown in Figure 6, the object moves smoothly with respect to the sensor, with the sensor 'binned' over five (5) pixels. In Figures A1-A6, the sensor charges are always shifted by an entire 'bound' pixel distance at a time, while in Figures B1-B6, the transfer takes place in parts of a single physical pixel of the sensor. It can be seen from Figures 6 that the object O to be imaged in method A is imaged not only by the five 'bound' pixels p1 of the sensor 18 but also by its neighboring pixel p2, while in method B the imaging is imaged by the fully 'bound' pixel p1 of the sensor 18 and no motion blur.

When operating with an implementation according to the basic idea of ​​the second method and device of the invention, where the physical pixel size of the sensor is smaller than the required maximum resolution, this above-mentioned motion blur can be reduced by the method and device of the second embodiment of the invention without transferring charge elements if the first method and / or device of the invention is applied, but according to the physical pixel size of the sensor. In this case, the movement of the object is monitored as accurately as possible. For example, when a sensor 18 'with a pixel size of 30 Mm is used, but with an image resolution of 150 Mm, if the pixels are' binned ', it is done vertically as previously described, but the horizontal' binning 'is now performed slightly differently. Pixel field 10; 11 booking items will not be transferred

The amount indicated by the HBx input signals to the receiver register 12 at a time, after which the receiver register 12 would be read out, but to the pixel field 10; 11 bookings will be transferred to mj mu:

to the serial register 12 always one column at a time until the input signals are transferred there

The number of columns indicated by HBx and only then the serial register is normally read out.

This mode of operation can be selected by the input signal BM of the control electronics 17.

In the example above, the blur in the direction of motion is only 30 μτη instead of 150 μτη otherwise present, i.e. significantly better. By using the free 'binning' according to the invention and a sensor with a higher resolution than the highest resolution required by the procedure according to the above embodiment, an accuracy which would not otherwise be achievable at all is achieved in the TDI description. Practical experiments have shown that this achieves a significant improvement in image quality.

In the above embodiments of the invention, the charges are stored in the receiver register 12 until all the 'binnable' columns have been summed there and the receiver register 12 can be read out. Unfortunately, this causes a dark current, since the elements of the receiver register 12 cannot be separated from each other by potential walls, as pixels can be, but in the elements of the receiver register 12 charges must be kept in place by energizing one of the clock phases. under the stage.

Prior to the development of potential ramparts, the pixel field charges of the CCD sensor were held in place by the same technique, and various methods were developed to minimize dark current without cooling. One of these known techniques is called 'ditheration', in which the potential well of one pixel is not permanently formed under one clock phase, but is continuously moved alternately under different clock phases of one pixel, while remaining within the range of one pixel. The technology is based on the fact that the dark current does not start to flow as soon as the clock phase is energized, but only starts to increase after a small moment. When the adjacent phase is energized and the current is de-energized, the dark current does not have time to increase significantly, as long as the operation is performed frequently enough and each clock phase is given sufficient recovery time.

When the clock phase is energized, it causes a leakage current to a pixel called a dark current (a signal without radiation to that pixel) due to the thermal movement of the electrons. The dark current is strongly temperature dependent and can be reduced by cooling the sensor. When the polarity of the clock phase is changed to reverse, i.e. negative to the sensor, it prevents the formation of the dark current described above. Indeed, this technique is used today, as previously described, by holding the pixel charges in place by means of fields doped on the sensor and by switching all clock phases to negative. The receiver register 12 does not have this possibility, but the charges must be held in place by means of a potential10 formed by the voltage clock phase, which unfortunately results in a dark current. The aim is to reduce this by means of 'ditherification' according to a preferred embodiment of the invention.

When the clock phase is switched to negative, the dark current immediately ceases to flow, because the potential is formed opposite to it, i.e. a diode is formed at the interface as if in the blocking direction. When the clock phase is switched to positive again, for example only a few so-called after it was switched to negative, the dark current does not immediately start to flow at the original level, but starts to grow exponentially as a function of time with a time constant inversely proportional to the number of natural charge carriers. The principle of 'dithering', i.e. vibrating clocking, is to keep the clock phases below which there is no need to form a potential well negative enough so that the dark current does not pass under them and alternate the clock phase considered positive between different clock phases of one pixel as shown in Figure 8. When one phase is considered positive for only a very short time compared to the above-mentioned time constant, a very significant reduction in dark current generation is achieved. In practice, this time constant for CCD sensors is strongly temperature dependent and is typically 10 s at -40 ° C, 15 ms at 20 ° C, and 180 gs at 80 ° C. When, according to Fig. 8, one clock phase is kept positive for less than 10 ms, a reduction of the dark current to less than one thousandth is achieved at room temperature.

In the preferred embodiment of the invention, the 'ditheration' or vibration clocking method known per se in the pixel field is applicable to the receiver register 12 under the operating conditions described above. In this case, the dark current can be virtually eliminated. In Fig. 7, according to this embodiment of the invention, either the reading of the saq register 12 is performed when it is its turn and otherwise the register is subjected to 'ditheration' during the integration period following. Figure 8 shows an example of the clock signals of a three-phase log register 12 when performing 'ditheration', where the timings are typically so short that no dark current is generated.

One inconvenience is that this self-paced 'ditheration' cannot 10 be synchronized to the image area trans mission intervals, which are random for the control electronics 17, and the shift clocks of the saq register 12 must be in a certain state whenever the pixel field 10 or storage field 11 charges are transferred. saqar register 12 in order to maximize the transfer transfer efficiency. This problem can be solved by driving the clock signals of the saq register 12 in a controlled manner to their correct state at the time when the charges of the pixel field 10 or the storage field 11 begin to move towards the receiver register 12. It can be seen from Figure 7 that when the SYNC signal arrives, the clock signals IOx of the image area perform the sequence required for charge transfer, during which the clocks of the register 12 must be put in their correct state before the charges move to the register 12, which in fact only occurs at step V14. Figure 9 shows the control of some 'ditheration' clock signals typical of a three-phase sensor from any state to a charge transfer state in a sufficiently short time. In the charge transfer state, the clock clocks 1 and 2 are kept energized in order to maximize the potential well below them, while the clock phase 3 is kept de-energized so that the charges in the different columns are not mixed.

With regard to the 'ditheration' method, which is preferably used in connection with the invention, reference is made to the following publications:

- DEVELOPMENT OF X-RAY CCDs, DJ.Burt, GEC-Marconi Limited, Hirst Research Center, and EEV Ltd, Chelmsford; Proceedings of an ESA Symposium on Photon Detectors for Space Instrumentation, held at ESA / ES30

TEC Noordwyk, The Netherlands, 10-12 November 1992. (ESA SP-356 December 1992).

Burke B and Gqjar SA 1991, Dynamic Suppression of interface state dark current in buried channel CCDs, IEEE Trans. Electron Devices, Vol. ED38-2.

Fig. 10 shows an X-ray panoramic device mainly for dental imaging, consisting of a vertical column 20 suspended by a motor 25 vertically movable and motor 24 rotating stem 21 having a radiation source 22 at one end from which a narrow X-ray beam 26 passes the patient P and hits the sensor 23. which is the above-described freely configurable CCD sensor according to the present invention within the set limits.

Figure 11 shows a mammography device standing on a stand 30 as an application environment of the invention. The base 30 is supported by a fixed vertical frame 32, inside which a motor 31 moves a vertically movable vertical frame 38. In this part there is a motor 33 which rotates the C-arm 44 about the horizontal axis HH. At one end of the C-arm 44 there is a radiation source 42 and at the other end a shelf 36 against which the breast 37 to be imaged is pressed by a pusher 43 moved by the motor 34. During the imaging of the breast M, the narrow beam 41 emanating from the X-ray tube 39 passes through the laterally movable primary blind by the motor 40, passes the imaged breast M and strikes the sensor 45 movable by the motor 35 simultaneously with the primary blind.

Fig. 12 shows a lung imaging device in which a narrow X-ray beam 73 emanates from a radiation source 74 and passes through the thorax of a patient P and strikes a sensor 72. During imaging, the radiation source 74 and sensor 72 are moved vertically by a motor 70. The sensor 72 is CCD sensor.

All the devices described above operate on the principle of slit imaging, to which, however, the invention is not limited. For example, the device shown in Fig. 11 can operate on the normal X-ray principle, i.e. by irradiating the whole breast M at once, the CCD sensor according to the invention being the size of the entire shelf or at least rectangular CCD area sensor within the set limits.

Fig. 13 schematically shows a control system 80 according to the invention of the imaging device according to the first embodiment, to which the user enters the control panel

81 using the keypad to determine the required imaging configuration parameters. Based on these input parameters and the pre-programmed presets in the parameter memory 82, the control system 80 sets the most suitable sensor settings for the sensor control electronics 17 (Figs. 1 and 2) for each shooting event by means of signals VBx, HBx and BM. During the imaging operation, the control system 80 synchronizes the operation of the sensor system with the START and SYNC signals to other operation of the application environment device, for which motors 83 and 84 are schematically shown in Figure 13, such as motors 35 and 40 in Figure 11).

Fig. 14 shows a combination of a panoramic and skull imaging device, otherwise similar in construction to that shown in Fig. 10 except that a horizontal arm 27 is connected to one end of which is connected to a cranial imaging camera 28 for cranial imaging. CCD-sensor system.

As is known, mammography images have been taken with the arrangements of Figures 15A and 15B. Figure 15A shows the so-called a contact description in which the breast M to be imaged is pressed substantially against the imaging means 100 and in which an object C is imaged by means of an X-ray beam X onto the imaging means 100 at a magnification calculated from the notation of Figs. 15A and 15B;

M (SID - OID) (1)

In practice, the average thickness of the compressed breast M is 45 mm, so in the contact description according to Fig. 15A, the object C in the middle of the breast is imaged at a magnification of slightly more than one. In an arrangement with X-ray focus

The distance F from the SID imaging means 100 is 65 cm, then the magnification ratio is then 1.036.

For a more detailed analysis, mammography takes the so-called magnification images that show the details of the subject in more detail. The arrangement is shown in Figure 15B. The breast M to be examined now is not pressed against the imaging means 100, but against the magnification imaging 101 above it. According to the previous example, when the distance OIDm is set to 32.5 cm, the magnification ratio is now 2.0, whereby the image of the object C on the imaging means 100 is twice as large.

Unfortunately, the procedure shown in Figure 15B also has its drawbacks. When the object M is brought closer to the focus F of the X-ray tube, the beam dose received by it increases in the square of the distance, i.e. in the case of the example, fourfold. In addition, the size F of the X-ray tube begins to have an increasing effect on degrading the imaging accuracy. The mapping can be viewed using the modulation transfer function (MTF), calculating it to be the square of the assumed focal point by the formula:

MTF (Ug, v) sin (π · v · Ug) π · v · Ug

¢), where Ug = &amp; (3)

M

The term M is the magnification ratio used and F is the focal point size, which is typically 0.3 mm for mammography devices for contact imaging and 0.1 mm for magnification imaging25 sa.

Figure 16 shows the MTF plot caused by the finite size of the focal point in different cases. Curve C31 shows the resolution in contact imaging with a focus point of 0.3 mm and curve C32 in magnification imaging with a factor of 2, using the same focus size. It can be seen from Figure 16 that the resolution of the magnified image ends at about 6 lp / mm, which is far from sufficient. Because of the above, a smaller focus size is generally used for magnification shots, typically 0.1 mm in size. The resolution of the magnification image taken with such focus and magnification factor 2 is shown in Fig. 16 by curve C12, which shows a considerable increase in resolution compared to curve C32. However, according to Fig. 16, it is not as good as curve C31 showing the resolution of contact imaging with a focus of 0.3 mm, and the usefulness of magnification imaging is explained by the fact that the worst link in the imaging chain is the film / gain plate combination shown in Fig. 17 as curve FS. Enlarging the image improves the resolution of the film / gain plate combination so much that even if the image blur caused by the focus point increases, the final resolution in the magnification images is better than the contact images.

Figure 17 shows the resolutions of the different imaging means. Curve D30 shows the resolution of a digital CCD image sensor with a pixel size of 30 gm and curve D60 the corresponding with a pixel size of 60 gm. Curve FS shows the resolution of a modern film / reinforcement plate20 combination. The MTF of the digital sensor is calculated from the formula:

MTFc (i) sin (tc · i · p) (4) where p is the pixel size in micrometers. The superior performance of the digital sensor can be immediately seen in Figure 17.

The resolution of the whole imaging system is obtained by multiplying the modulation transfer functions of the different subsystems by each other, i.e. in this case the focus point transfer function is multiplied by the imaging means transfer function:

(5)

MTFtot (i) = sin (π · i · p) sin (n 'i' Ug}

Fig. 18 shows the curves calculated according to the previous formula in the following cases, and for comparison, the resolution of the film has also been added to the figure.

<td>Curve</td><td>Focus</td><td>Enlargement</td><td>pixel</td>

<td>D60c</td><td> 0,3</td><td> 1,04</td><td>60 gm</td>

<td>D60m</td><td> 0,1</td><td> 2,00</td><td>60 μτη</td>

<td>D30C</td><td> 0,3</td><td> 1,04</td><td>30 μτη</td>

<td>FSM</td><td> 0,1</td><td> 2,00</td><td>Film / reinforcing plate</td>

From the examination of Fig. 18, it can be immediately seen that by using a sensor according to the invention, the pixel size of which can be varied if necessary and performing 'magnification imaging' as a contact imaging with a smaller pixel size, a significantly better resolution is obtained than with conventional magnification imaging and known fixed pixel size. The difference between curves D30c and D60m is only due to the difference in the transfer functions of the 0.3 mm focus point used in contact imaging and the 0.1 mm focus point used in magnification imaging, with the sensor transfer function remaining essentially the same in both cases. After all, the imaging resolution of the sensor is the same with a pixel size of 30 gm in contact imaging as with a pixel size of 60 μτη when the object is doubled.

In addition to achieving better resolution by changing the sensor resolution by the method of the present invention than current magnification methods using a fixed pixel size, a simpler and less expensive X-ray source requiring only one focal size is also provided.

It should also be noted that the 0.1 mm focal point of the X-ray tube anode plate is considerably smaller in physical area than the 0.3 mm focus and can withstand only about 20% of the power at which a larger 0.3 mm focus can be loaded. This inevitably results in longer, approximately five-fold imaging times when shooting a similar subject and is more likely to cause motion blur as the patient moves during imaging.

Since the area of ​​30 μιη pixels is only a quarter of the area of ​​60 μτη pixels, the amount of radiation required to achieve the same sensor signal when imaging at 30 gm pixels is basically four times, which is exactly the same as in the magnification conventional method. However, since the radiation passing through and attenuated by the object has to travel only a very short distance before hitting the imaging device compared to conventional magnification imaging, it is not significantly attenuated in air and is practically achieved with a lower radiation dose than the conventional method.

The mammography application of the invention presented above also has the advantage that by taking magnification images with a smaller pixel size and using contact imaging, it is possible to take magnification images at the size of the entire image area, if necessary. In this case, the disadvantage in the prior art is eliminated, in which the object is brought closer to the X-ray tube that, e.g.

In the following, the claims are defined, within the scope of the inventive idea defined by which the various details of the invention may vary and differ from the preferred embodiments set forth above by way of example only.