WO2008025771A1 - Optimized detection and processing of image data of an object in a device for the generation of computer tomographic images - Google Patents

Abstract

The invention relates to a method for optimizing the throughput when processing images from X-ray computer tomographies (CT). The proposed method for handling or processing image data from individual images or consecutive images is based on a substantially punctiform radiation source (70), by which an object (1) to be captured or reconstructed is depicted on a flat detector (80) readable as a screen. A plurality of processing strips (T1, T2) of the image are read at the detector (80), wherein each strip (T1, T2) takes up an image portion. The image portion is associated by a respectively associated volume layer (V1, V2) of the object (10) with respect to the radiation path of the radiation source (70). A respective image data portion in a respective processing strip (T2) is fed to a processing unit (E2). Each volume layer (V1,V2 , V6) of the object can hereby be reconstructed; a respective volume layer is reconstructed by one - of a plurality of - processing units (E1,E2 ,E6).

Description

 Optimized acquisition and processing of image data of an object in a device for generating computed tomography of images

The invention relates to a method for optimizing the throughput in the processing of images from computed tomography (CT). In particular, X-ray CT is addressed, in which a high-frequency radiation source in the X-ray region is used as the radiation source.

"Radiation" as such essentially represents any type of image-emitting radiation that is able to generate images on a detector in the transmission method, which can be embodied as an area detector, for example. The starting point of the radiation is a radiation source as a point source or line-shaped source (in the sense of a multiplicity of individual point-source radiation sources arrayed in a row), the

based on the object provided between the radiation source and the detector which is the subject of the analysis or detection of the CT, does not produce parallel but divergent rays.

In CT with a surface detector, large amounts of data accumulate on the image data

Image data processing to be distributed to multiple processing units to be processed in parallel. With increasing size or resolution and also with a growing speed of the detectors detecting the image data, the amount of data to be processed increases so much that this distribution, which also has to be managed, is no longer sufficient. Also, quantization of the detector, that is, resolution of each pixel, causes an increase in the workload. The increased amount of data thus includes other factors, such as the number of pixels in the detector, the frequency of the recording or the recordings and the gray value resolution of each pixel.

The prior art is already using image computation techniques in CT, where typically radiographs of an object are made from different in-plane viewing angles. The object is located between the radiation source and the detector. As a detector, a line or area detector is used, the area detector offers the advantage of an entire

To be able to depict an object. The line detector can only record the object layer by layer. From the image data recorded by the line or area detector, a data volume is then reconstructed using a calculation method which Volume describes the recorded object, which in turn is represented as a volume to be imaged.

In order to bring the data stream for calculation to manageable orders of magnitude, the state of the art works with smaller arithmetic units, which are used to reduce the

Calculation time or to accelerate the calculation, the existing image data of the detector partially each arrive. Each smaller unit receives only a portion of the entire image data of the detector. This division is carried out by a main computer to which the image data of the detector are supplied. Each individual computer then calculates only a partial volume of the total object to be reconstructed in order to later reassemble the partial volumes into a total volume.

Due to the described increase of image data, the overall processing speed of the image data between capture and completion for the entire object is reduced. It is therefore to be regarded as a technical problem of the invention to restore this calculation speed. With increasing amounts of data, the speed should be maintained or, with the volume of data remaining constant, the speed of calculating the total volume should be accelerated.

The invention has the (technical) task of proposing a method in which the recorded image data of the area detector are processed better, ie faster, and the throughput of a CT data processing system can be increased, so that throughput optimization in the processing or processing of

Image data is obtained in the CT.

This object is achieved by a method (claim 1, 2), in which the image data of the area detector are divided. For each of the processing strips a separate computer is connected, which reconstructs a corresponding partial volume of the object. This avoids the bottleneck of a central recording and distribution computer (so-called rendering PC).

The invention makes use of the fact, not only to accelerate a speed only of the executing computers, but to make a division of the detection of the image data right in the beginning and before the start of such calculations and such distributions of data to be calculated (claim 14). The division of the

Data stream is advanced so far that it extends into the detector of the object to be processed imaging. Already here, the data stream splits through the multiple parallel scans.

Preferably, the image data of a third volume layer as the third image data portion or third processing strip of a third arithmetic unit supplied (claim 3). Further preferably, further image data in one or more further volume layer (s) are supplied as still further image data component (s) or even further processing stripes to further arithmetic unit (s) (claim 4).

This or the preceding, starting from the imaging detector, which has as a surface sensor (claim 7) a plurality of pixels in the x and y directions. Due to the division of the image data into processing stripes which optically separate each substantially (or mainly) a volume layer, the image data can be processed without a distributing host.

This results in a plurality of image data outputs (claim 7, 8) of this optimized area camera, which operate in parallel and unaffected by each other, each of these outputs provides data of a layer of the object.

The reading out of the electrical information from the "directly converting detector" or the readout of the electrical information from interposed, each a strip on the optical imaging screen or detector reading signal generators (claim 4) should be understood analogously so that in each case a signal is "read" that is either optically or identically electric. A read-out would be understood more electrically, a reading rather optically, so that a reading of a respective strip T1, T2, ... should be used as a common descriptive term. Thus, there are data at a respective output, which represent the associated volume layer, additionally supplemented by an image data transition section on both sides at the edge of the layer for the one rays which pass through both volume layers on their way to the detector, the associated volume layer and the one on this side adjacent bulk layer, and the others

Rays which pass through both volume layers on their way to the detector, the associated volume layer and the other hand adjacent volume layer (claim 9).

Following this several sub-volumes of the several involved independent computing units can be rejoined to a total volume, which should not be explained in depth here.

The detector data (the stripes as an image of the volume layers) are not recorded centrally on the "rendering PC", but directly from the processing PCs from the

Detector taken. A small portion of the image data is processed twice, and used by each of two computers for the calculation. However, the amount of data from all the stripe overlaps (corresponding to the overlapping areas in the imaged volume) is substantially less than the image data amount from the non-overlapping main volume (or the image data from a volume layer of the object is large compared to the proportion of the volume parts of adjacent volume layers imaged overlappingly originates).

The rays of the point source are not all completely parallel to the longitudinal extent of the volume layers, but also not strongly inclined at an angle (claim 13). It's supposed to be from

"essentially parallel" are spoken, the strip overlaps can be kept small or narrow. As a result, the image data of overlapping volumes is also small.

This results in the following graduation: Once the total object as a total volume.

Then the contrast, much smaller stripe overlaps. The result of this graduation is that each layer can be calculated independently and independently, without cross-coupling between the individual calculation units.

Only when merging the image data from the individual volume layers, a reciprocal influence can again take place in order to put the interfaces between the volume layers true and to each other-add. Exemplary embodiments explain and supplement the claimed invention.

Figure 1 is a schematic representation of the structure of a

Radiographic method for reconstructing the object 10 arranged between the source and the detector and the resulting image data in the detector 80 (detector device).

FIG. 2 is a schematic representation of the multi-output optical detector device 80.

FIG. 3 illustrates the electrical signals of the image data from the overlapping regions of the image of the detector 80.

FIG. 4 is a two-dimensionally subdivided method with

Image sections in both directions x, y of the detector 80.

Figure 5 is a vertical sectional view of the structure of Figure 1 with beam paths.

FIG. 6 is an illustration of the overlap (s) of FIG

Image data from two adjacent volume areas.

Figure 1 illustrates the technical field of application and a first example of the method in a schematic view. A radiation source 70 emits a beam cone q, which radiates through an object 10 and then impinges on a screen, also called detector or detector device 80.

This screen receives an image 10 'of the volume of the object 10, which in the example is shown as a cylinder but may have other shapes as well. It is a voluminous object whose volume can be symbolically or thoughtfully divided into separate layers, which are later relevant in capturing or reconstructing the contents of the volume. The radiation q can, for example

Be X-ray. The detector 80 may, for example, be designed as a planar area detector which emits electrical signals on its rear side 81, which corresponds to the transmission cone. Assuming a point-shaped radiation source 70, the emitted radiation q is formed as a cone. Likewise, symbolically several points can be strung together to form a line-shaped radiation source. Then the beam form q is fan-shaped. The result is a shadow on the screen 80, with which the radiation after the irradiation of the body 10 can also be described. The

Scanning takes place from the back 81, either directly electrically or optically with a camera or with individual images.

Here it should be assumed that on the back 81 of the screen 80 (as an example of a detector) electrical signals can be removed, which in

Figure 2 are symbolized with e1 to e6. The detector device 80 may have on its rear side 81 associated strips T1 to T2, which correspond to the electrical signals e1 to e6.

The optically imaged stripes T1 to T6, which reproduce the image 10 'of the body 10, are to be processed. They are mapped in the electrical signals e1 to e6 and supplied from there to arithmetic units E1 to E6, each of which processes or processes a strip. These are directly connected.

The optical data from the example. Single image or of successive

Recordings can thus take place without a division in the sense of an intermediate computer, which divides the area image on the back 81 into individual tiles (image sections) and then feeds them to independent computers for independent processing. Each computer of Figure 3, here, for example, six computers E1 to E6 for six strips T1 to T6 and their associated electrical signals e1 to e6 is able to its strip section to

Reconstruct reconstruction. The strip section can also be referred to as "TiIe", is analogously an image section or image section of the whole, and at least two strips T1, T2 are to be provided in order to obtain a division of the computing power on at least two computers E1, E2. Of course, preferably three, four or more such strips.

Large objects can be imaged on a large screen, for example, the screen can have an area of 1000 x 1000 pixels. The larger the area, the lower the area unit that is produced per pixel.

The accumulating large amounts of data must be able to be processed cost-effectively and time-effectively. The processing is, for example, an acquisition. The processing may be a reconstruction or an analysis. A doubling of the resolution leads to a four-dimensional detector to a quadrupling of the optical data and to a multiplication of the necessary computing power. Therefore, not all optical data is processed with a computer and not provided a computer with which the present optical data are first divided, but already in the detection of the optical data as electrical signals (directly or indirectly via an image sensor, the shadow cast on the Reverse 81 scans) image sections are processed. The throughput of the data processing system consists of the units E1 to E ₁ (shown are the units E1 to E6 in Figure 3).

The goal is the image B.

The object 10 to be analyzed and reconstructed by means of X-rays, for example, is reconstructed in individual layers called volume layers V1 to V6, each substantially depicted in tiles

T1 to T6. For reconstruction with a CT, shots from different angles are provided by each layer. However, the layers can basically be reconstructed independently of one another if all data (all image data) of a respective layer are available. This results - depending on the geometry of the receiving arrangement and the shape of the beam q - a different

Influence of adjacent volume layers. This is illustrated in FIG. 3 in the left section. The electrical signal e3 from Ti3 T3 essentially contains information from the third volume layer V3, as shown in FIG. For this purpose, it also contains a small proportion of the volume layer V4, which is adjacent to the volume layer V3 below.

If the plane between the volume layers V3 and V4 in the central plane z1 of the radiation source 70, no proportion, but for a proportion of the overlying volume layer V2 results here. These two adjacent parts are called e23 and e34.

Essentially, the signal from the volume layer V4, but also from both adjacent volume layers V3 and V5, results for the TiIe T4 and its electrical signal e4 a small proportion. He is also called e34 and e45. The volume layers V2 and V5, the figure 3 behaves accordingly and is not further elaborated here. However, the transmission of the electrical signals from the strips to the separate arithmetic units E1 to E6 is carried out. The overlapping areas in the volume and in the image of the object on the screen will be explained later with reference to FIGS. 5 and 6.

According to FIG. 4, it should be noted that the object 10 reconstructed in horizontal layers can not only be processed in horizontal image strips T1 to T6, but can also be divided into vertical stripes and decomposed, for which purpose further arithmetic units are used which are not shown separately. It is shown for the area of the processing strips T3, T4 that, compared to a vertical plane 200 and a horizontal plane 201, four image sections are possible, which are called T33, T34, T43 and T44, two of them from the processing strip T3 and two of them Processing strip T4. A vertical banding and consequently a further subdivision of the image data onto the detector into a plurality of arithmetic units is thus apparent. The symbolically represented formation of image sections can be continued further in the manner of an xy raster formation such that TiIe T35 in the processing strip T3, which corresponds to the volume layer V3 the signal. The TiIe is so generally definable as T _1J , where i runs from 1 to n and j from 1 to m.

From the substantially punctiform radiation source 70 flows a conical or fan-shaped beam q, which passes through the radiopaque object 10. This object 10 is shown in Figure 5, the beam is shown interrupted to the weak

To make conicity clear, which comes close to parallel rays. The transmittable object 10 forms a shadow on the detector 81, which is designed as a flat detector. The screen 80 may make the shadow of the object 10 either optically visible, for example with a scintillation layer, a conversion of X-ray light to visible light. An interposed optical detector reads a strip, another another strip and a still further a third strip on the back 81 of the detector 80 from. However, the detector 80 can also be designed such that it emits the same electrical signals that stand for a respective strip or for a further divided TiIe of Figure 4. A reference line 100 may be associated with the object as a rotation axis. This reference axis can also be assigned to the entire arrangement of radiation source 70, object 10 and detector 80, likewise a center plane of the object 10 can be used as the reference plane in which the reference line lies.

The plurality of strips into which the object 10 is subdivided for reconstruction are the volume layers V1 to V6 of FIG. 5. Each of these layers essentially forms a processing strip, here as strips T1 to T6 on the detector. At least two of these stripes are used to increase the amount of data to reduce. Each of these strips then feeds its own processing device E1, E2, without an interposed division of the electrical signals or image data by an intermediate computer.

The strip, in the example strip T1 (or T2), is formed to correspond to the image area formed by all the rays passing through the bulk layer V1 (or V2). These are all rays between the boundary rays 50 and 52 for V1. The rays come from the radiation source 70. The second strip T2 has the image surface which is defined by all the rays which pass through the second volume layer V2 of the object 10. These are all the rays that are between the boundary rays

51 and 54 of the radiation source 70 are located, as shown in FIG.

It is to be understood that the volume layers V1 and V2 do not overlap for this imaginary subdivision of the object 10, but it is apparent that overlaps occur in the processing strips T1 and 12, which is illustrated in FIG.

The image information of the first strip T1 and at least one further processing strip T2 are each supplied to a separate or independent arithmetic unit, which is symbolized in FIG. 3 by E1 and E2. The further strips are formed accordingly, which can be easily understood in Figure 5. A third strip T3 arises from the volume layer V3 and adjacent overlap area, as well as the still further strip T4. In the example of FIG. 5, the center plane z1, which runs through the radiation source 70, is drawn between the strips T3 and T4. Therefore, at this point, no overlap area of the volume layers V3 and V4 is present at the boundary between the processing stripes T3 and T4. Rather, only above and below the border to the next volume layer V2 or V5.

The reconstruction of each of the strips mentioned in each case takes place independently from the respectively assigned volume layers of the irradiated object 10.

Several of these images can be made for the individual layers V1 to V6, from different directions. Also, several shots successively from the same direction can be used. It also remains here, however, that the detector data is not centrally recorded on a computer and from there to

Instead, the detector is provided with a plurality of outputs for a respective recording or for several successive recordings, which are either provided electrically at the detector 80 itself, or from intermediate switched image sensors are provided which detect the visible light on the back 81 of the detector and convert it into electrical signals, and this either strip-shaped or type of tiles according to Figure 4. The delivery takes place directly to a respectively assigned arithmetic unit E _k , where k = 1 to n ^* m at n ^* m tiles and k = 1 to n at n processing strips T ₁ .

The reading out of the electrical information from the directly converting detector or the reading of the electrical information from the interposed, each a strip on the optical imaging detector 80 reading - not separately shown - signal generators should be mutatis mutandis understood that in each case a signal is read, either is optically or the same electric. The reading would be understood rather electrically, the reading rather optically, so that a reading of a respective strip T1, 12, ... should be used as a common descriptive term.

It is understood that the major part of each volume layer is the

Main portion of a processing strip, or it corresponds. By contrast, the proportions of the imaged overlap can be regarded as very small. The more parallel the radiation is to the interfaces between the bulk layers, the lower the signal or image portion from the overlap region. This will now be explained in detail with reference to FIG.

For this purpose, a first strip is recorded as a processing strip T1 and a second strip as a processing strip T2 on the detector 80. The detector is only partially reproduced here, as well as the irradiated object 10. The lightly or slightly diverging rays of the radiation source 70 are clear from the

Horizontal inclined shown to make the overlapping irradiated areas at the interfaces between the volume layers more clearly and consequently also the overlap portions of the image signals and the electrical signals e1, e2.

It must be pointed out that the choice of T1, T2 and V1, V2 here is the

Simplification took place and they do not correspond to the volume layers V1, V2 of the previous figures. It must also be pointed out that the beam S1 lies in the center plane z1 and therefore no overlap fraction at the lower edge g5 of the volume fraction V1 in the signal of the processing strip T1 at the right edge of FIG. 6 arises. The overlying bulk layer V2 may, for example, be cylindrical when the body of FIG. 1 is used. In section, it is shown in Figure 6 so that a vertical cut was made, which lies in the axis 100 and passes through z1. It should also be assumed that the beam shape is fan-shaped, ie does not form a cone, but an inclined surface S4 is formed at the upper edge and a horizontal plane S1 at the bottom of the section of Figure 6 is formed. The four boundary surfaces of the second volume layer V2 are the top and bottom surfaces g2, gl and front and rear g3, g4. For the volume layer V1, there are the boundary surfaces g1, g5 at the top and bottom, and the boundary surfaces g6, g7 at the front and the back.

A beam S2 and a beam S4 from the radiation source 70 are located so as to detect the entire volume V2. This is between the left upper corner and the lower right corner of the volume V2 in the figure, and thus also the volume components V3 'and VV, which come from the adjacent volume layers V3 and V1 yet. These

Main layer (and adjoining edge layer portions) are irradiated by the rays which are above the boundary ray S2 and below the boundary ray S4, and this in only a small solid angle.

Would you between the rays S4 '(dashed lines) and S3 the

Radiating volume layer V2, the shadow cast on the screen would be shorter than the illustrated height h2, which results between the boundary rays S4 and S2. For this smaller section, only shadowing and thus only image signals from the pure volume V2 would be imaged on the detector 80, but that is not sufficient, since above and below S2 / S4 a small portion of the volume V2 remains that does not have an image (shadow cast). generated in the processing strip T2 of height h2.

Thus, the electrical signal e2 would be incomplete. Therefore, the two hatched volume sections V3 'and VV are still added, which allow the remaining strip V2' (below) and V2 "(above) to still participate in the measurement signal, but with an overlapping area of another volume section in the image (FIG. However, this can be calculated out again in the processing strip T2 and from its electrical signal e2, which reaches the independent or respectively independent arithmetic unit E2 for the processing strip T2.

In the same way, the area at the boundary between the layers V1 and V2 between the beams S3, S2 can be seen. The volume fraction V2 'is added as an overlapping area to the TiIe T1 and its electrical signal e1. This TiIe has the height h1 and is supplied to the arithmetic unit E1. By using the

Boundary beam S3 and the boundary beam S1, the entire volume layer V1 can be imaged in the Ti1 T1 of the height h1. The explanation of FIG. 6 can be transferred to FIG. 3 and to FIG. A respective image data component T ₁ , where i in the example runs from 1 to 6, is supplied to a respective independent arithmetic unit E ₁ , i also running from 1 to 6. The arithmetic units are thus responsible for the volume layers V ₁ , wherein each arithmetic unit is responsible for a (substantially) own volume layer, and obtained via each adjacent reduced volume fractions a small impact, which is excluded, but involved the total volume of the respective volume layer on the measurement signal ,

An independent computing unit is not necessarily to be understood that it is physically a self-sufficient computing unit. It is also possible to provide independent arithmetic units in a larger computer, so that the independent arithmetic unit represents a better explanation of the respective independent calculation for an independent volume layer.

The optical overlap portion T12 on the detector corresponds to the volume proportions V2 'and VV. Depending on the measured volume layer, the proportion V2 'is parasitic, for the processing strip T1, or the volume fraction V1' for the processing strip T2. Correspondingly, this also applies to the interface g2 at the upper edge of the volume layer V2 and that above T1 (in the upper

Border region of T2 overlapping) signal, which would be called here T23 (between S4, S4 ').

If the receiving geometry or the shape of the radiation source 70 also changes, the overlapping sections change both on the boundary beams S2, S3 and S4, S4 'explained here in FIG. 6 and the associated optical signals on the detector 80 Calculation of the autonomous computing units E1 to E6 is adjusted.

Due to the fanning of the beam path q it can be seen that the

Overlap portion propagates in the detector signal, the more conical the rays run. The overlap T12 shown as a height in FIG. 6 corresponds to a greater height in the strip, the farther away it is from the central plane z1. The overlap portion in the signal is also circumscribed to be defined by all the rays of the radiation source 70 which pass through both the one and the adjacent bulk layers (eg, at the boundary gl) along a respective beam, the two volume layers V1, V2 are adjacent and do not overlap. This symbolizes the interface gl.

* *

Claims

Claims: 1. A method for processing or processing of optical data from individual images or 5 consecutive images, preferably from the CT, with the following steps (a) on a - in particular substantially punctiform - Radiation source (70) is a durchstrahlfähiges object (10) on a two-dimensional detector (80) as a screen optically detectable lo mapped, wherein the object a reference line as the main axis (imagined), in particular as a rotation axis (100) of the object (10) or the arrangement of the radiation source (70) and the screen (80) or a center plane of the irradiated object (10); i5 (b) a plurality of stripes (T1, T2) of the imaged image are optically detected from the overall image of the detector (80); wherein at least two of the plurality of processing stripes (T1, T2) are formed such that; (d) a first of the strips (T1) has the image area which is 20 of all - a first volume layer (V1) of the object (10) passing rays of the radiation source (70) is defined; (c2) a second of the strips (T2) has the image area defined by all the rays passing through a second volume layer (V2) of the object (10), the two volume layers not overlapping, in particular adjacent; (d) the image information of the first and further Processing strip (T 1, T2) of the detector is in each case supplied to a separate or independent computing unit (E1, E2); for each autonomous reconstruction of the at least two volume layers (V1, V2) of the irradiated object in the at least two separate arithmetic units (E1, E2). 35 2. A method for processing or processing of image data from individual images or successive images, preferably from a CT, wherein (a) a substantially point source (70) of radiation to be detected or analyzed or reconstructed 5 object (10) on a flat detector (80) as a screen optically readable images or provides electrically readable provides; wherein the object is a main axis, in particular as a rotation axis (100) is intended; (B) a plurality of processing strips (T1, T2) of the image on the detector lo (80) are read or read, each of the strips (T1, T2) occupying an image portion, which of a respectively associated volume layer (V1, V2) of the object (10) is assigned from the beam path of the radiation source (70) and the respective volume layer in the respective image portion is represented as a respective image data component i5, and wherein each of the volume layers is substantially smaller than the volume of the object (10), in particular half as large; (c) a respective image data portion or a respective processing strip (T1, T2) of each one of them processing 20 arithmetic unit (E1, E2) is supplied to the reconstruction of the individual volume layers (V1, V2) of the object in the individual arithmetic units (E1, E2). 3. The method of claim 1 or 2, wherein the object is detected in the context of an X-ray CT and the rays are X-rays. 4. The method of claim 1 or 2, wherein the processing strips (T1, T2) of the image from the detector independently and / or each independently 30 are read, in particular by an appropriate number of Image acquisition devices are read out. 5. The method of claim 1, wherein the image data of a third volume layer (h3) as the third image data portion or third processing strip of a third 35 arithmetic unit (E3) are supplied. 6. The method of claim 5, wherein the image data of yet another Volume layer (h3) as still another image data portion or still further processing strips of yet another arithmetic unit (E3) are supplied. 5. The method of claim 1 or 2, wherein the image of an optical Detector (80), which is a back surface image generator (81), for outputting electrical signals upon the arrival of X-rays (q). 8. The method of claim 1 or 2, wherein the detector is provided as an imager with multiple lo outputs, each for the delivery of all image data portions of an associated Volume layer of the object, corresponding to a processing strip (T1, T2) at the detector, in particular the detector receives the image or the recording in a transmission method. 9. The method of claim 1 or 2, wherein each processing strip having an adjacent processing strip has a common overlap portion (T12) except for the top / bottom or right / left edge processing strip at the detector. 10. The method of claim 9, wherein the common overlap portion is changed, this with a change in the receiving geometry. A method according to claim 1, 2 or 10, wherein the overlapping portion of two adjacent processing stripes varies in width as it gets farther from 25 of the horizontal (z1) is removed by the radiation source (70). A method as claimed in claim 9 or 10, wherein the overlapping portion (T12) is defined by all of the rays of the radiation source (70) which along each beam comprise both the one and the adjacent bulk layers 30 (V1, V2) of the object (10), wherein the two adjacent Volume layers (V1, V2) do not overlap. 13. The method of claim 1, wherein the radiation is an X-ray emanating from a substantially point-shaped radiation source (70). 35 14. The method according to claim 1 or 2 or 4, wherein all image data portions (T ₁ ) of all non-identical volume layers (V ₁ ) of the object (10) independently or each independently a separate arithmetic unit (E ₁ ) are supplied. 15. The method of claim 2, wherein a plurality of processing strips (T1, T2) of the image on the detector (80) optically detected reading or reading, each of the strips (T1, T2) occupies a respective image portion. * * ¥