DE102015110359B4 - Method for detecting radiation-emitting particles - Google Patents

Abstract

Method for detecting particles (3) moving along a trajectory, which generate or at least influence electromagnetic radiation, an electric field or a magnetic field, wherein the electromagnetic radiation, the electric field or the magnetic field is detected, a first signal having Using a first structuring device (5, 10) is detected, the first structuring device (5, 10) either ensures that the particles (3) along a trajectory section essentially only in first non-periodic spatial intervals electromagnetic radiation, an electrical Generating a field or a magnetic field or at least influencing or ensuring that the electromagnetic radiation, the electric or magnetic field is essentially only detected in first non-periodic spatial distances along the trajectory section, characterized in that a second signal with the aid of a second structuring device (11), the second structuring device (11) either ensuring that the particles (3) along a trajectory section essentially only generate electromagnetic radiation, an electric field or a magnetic field at second non-periodic spatial intervals or at least influence or ensure that the electromagnetic radiation, the electric or magnetic field is essentially only detected in second non-periodic spatial intervals along the trajectory section, that the first and the second signal are correlated with one another and conclusions can be drawn from the correlated signal on the number and/or the speed of the particles (3) can be drawn.

Description

The present invention relates to a method for detecting particles moving along a trajectory which generate or at least influence electromagnetic radiation, an electric field or a magnetic field. In this case, the electromagnetic radiation or the electric or the magnetic field is detected. Furthermore, a structuring device is used, which either ensures that the particles along the trajectory essentially only at non-periodic spatial intervals generate or at least influence electromagnetic radiation, an electric field or a magnetic field, or ensures that the electromagnetic Radiation, the electric or the magnetic field is essentially only detected at non-periodic spatial intervals along the trajectory.

Such a method is used, for example, in flow cytometers. In flow cytometry, a cell suspension is passed through a flow cell. The cell suspension is irradiated with a light source, usually consisting of several lasers, whereby the cells to be detected or a dye bound to the cells are excited and emit fluorescence signals. A non-periodic, structured mask, also known as a shadow mask, is often used as the structuring device, and the particles are guided past one side of the mask at a speed v and on the other side the electromagnetic radiation, the electrical or the magnetic radiation, passing through the mask Field detected with a suitable detector.

However, it goes without saying that the structuring device does not have to be a shadow mask. It would also be conceivable, for example, for the particles to be detected to have to be excited using structured excitation radiation in order to generate electromagnetic radiation. This excitation radiation could be made available, for example with the aid of diffractive optical elements, only at non-periodic spatial distances along the trajectory, so that the electromagnetic radiation to be detected is essentially only emitted at these locations. The structuring device then consists of the diffractive optical elements.

The present invention is described below using a method for detecting radiation-emitting particles. Nevertheless, the method can also be used to detect particles that generate or influence a magnetic field or an electric field, for example. It would also be conceivable to conduct electromagnetic radiation through the mask and to detect it, so that the particles guided past briefly weaken or at least influence the directed electromagnetic radiation when passing through the masks.

For example, the temporal modulation of the fluorescence signal can be generated by a shadow mask structured along the direction of movement of the particle. Windows in this shadow mask allow the transmission of the fluorescent light and correspond to the transmission state "1". Closed areas of the mask prevent transmission and correspond to the "0" state. The spatial structuring of the mask is described mathematically by so-called binary sequences, i. H. by sequences of zeros and ones. The structured mask is therefore also referred to as a binary mask. In principle, however, areas with reduced transmission, which could then be assigned an intermediate state of, for example, “1/2”, would also be conceivable. Even though the invention is explained below using binary masks as an example, structured masks with intermediate states would also be conceivable.

Basically, the measurement signal is encoded by the structuring device. The code, for example the binary sequence, is made available by the structuring device.

The fluorescence signals encoded in this way are detected and evaluated in order to draw conclusions about the number, type, intensity and speed of the individual particles.

Such a method is described, for example, in Kiesel et al., Appl. physics Latvia 94, 041107-1 (2009), at DE 10 2013 105 953 A1 and in Sommer, Christian et al: "The equilibrium velocity of spherical particles in rectangular microfluidic channels for size measurement." In Lab chip 2014, 14, pp. 2319-2326. Here, the measured signal is filtered using an optimal filter in order to achieve pulse compression and then differentiated according to time in order to eliminate a possible signal offset. The resulting signal uses a fixed event detection threshold.

It is known from Kiesel et al., Cytometry Part A, Vol. 79A, 2011, pp. 317-324, to carry out two measurements simultaneously with the structure described, with different particles being detected with each measurement.

The detected signal is naturally unipolar since there can be no negative amplitudes in the fluorescence emission.

The use of non-periodic masks or signal curves serves to improve the signal-to-noise ratio and temporal pulse compression. However, with this method, signal sidelobes occur with correlation signals, which makes it much more difficult to detect weakly fluorescent objects in the presence of larger, more luminous objects.

For clarification, reference is made to the 1 until 4 referred. 1 shows a schematic representation of a measurement signal as a function of time without taking into account the noise and offset component. In 2 shows the amplitude-modulated square-wave signals for two particles. It can be seen that the measurement signals have different radiation intensities. The first measurement signal has twice the amplitude of the second measurement signal. Superimposition results in a third level which is 1.5 times the value of the first measurement signal. However, the square-wave signals of the individual particles not only have different amplitudes, they are also shifted in time (in the x-direction), i.e. the case is considered in which two particles of different light intensity and different dimensions move past the binary mask in quick succession, so that the overall signal, which is detected by a detector, is a superimposition of the individual signals originating from the individual particles.

This signal is usually processed with an optimal filter, whereby the in 3 shown filtered signal results. An optimal filter is understood to mean a filter which supplies a filtered signal with the greatest signal-to-noise ratio. In the case of a unipolar signal, such a filter always has exclusively unipolar filter coefficients.

The signal from the first particle (P1) is clearly visible, while the weaker signal from the second particle (P2) is only weak. For automated particle detection, a limit is usually set and signals stronger than the set limit are interpreted as particle detection. However, this is not possible, particularly when detecting several particles with different signal strengths. Examples are in 3 two threshold values have been drawn as dashed lines. If the larger value is used as the threshold value, only the first particle is detected since the signal strength of the second particle is below the larger threshold value. If, instead, a threshold value is chosen that is so small that the second particle is correctly identified, the background signal associated with the first particle is also above the threshold value, so that the first particle may be incorrectly counted more than once.

Therefore, usually the filtered signal is derived with respect to time, whereby the in 4 signal shown is obtained. The signal no longer has an offset. Again, the signal from the first particle (P1) is clearly visible, while the signal from the second particle (P2) is only slightly larger than the signals in the side lobes. In the 4 a threshold value (dashed line) has been drawn in as an example. If the threshold is too high, the signal from the second particle will not be detected. On the other hand, if the threshold value is too low, not only is the second particle detected, but additional peaks of the side lobes, which are highlighted in the figure with a dotted line, are incorrectly recorded as an event.

1. 1. Aufgrund der Abschattungseigenschaft der binären Maske, wird nicht das vollständige, zur Verfügung stehende Fluoreszenzlicht genutzt. Wird beispielsweise eine „balancierte Maske“, d.h. eine Maske mit der gleichen Anzahl von Nullen und Einsen, genutzt, so werden circa 50 % des zur Verfügung stehenden Fluoreszenzlichtes nicht genutzt. Könnte dieses Licht genutzt werden, würde sich das Signal-Rausch-Verhältnis um einen Faktor √2 verbessern.
2. 2. Das bekannte Verfahren mit der binären Maske liefert lediglich ein unipolares Signal, so dass aus der Radartechnik bekannte Detektionsverfahren nicht angewendet werden können.
3. 3. Durch die Verwendung eines einzelnen Detektors fließt dessen Rauschhintergrund immer direkt in das Signal mit ein und verfälscht somit das eigentliche Messsignal und erschwert die Detektion.
4. 4. In der Praxis ist die Partikelgeschwindigkeit nicht bekannt. Zudem wird bei der Messung mehrerer Partikel eine Geschwindigkeitsverteilung beobachtet, so dass deren Verweildauer im Detektionsbereich, das heißt die zeitliche Ausdehnung des gemessenen Signals nicht nur unbekannt ist, sondern sich auch noch von Partikel zu Partikel unterscheiden kann. Daher muss ein breiter Geschwindigkeitsbereich abgefragt werden, in dem das gemessene Signal mit verschiedenen, zeitlich unterschiedlich ausgedehnten Filtern korreliert wird. Dieser Vorgang ist mit erheblichem Rechenaufwand verbunden.

However, the known method has a number of disadvantages:

1. 1. Due to the shading properties of the binary mask, not all of the available fluorescent light is used. If, for example, a "balanced mask", ie a mask with the same number of zeros and ones, is used, around 50% of the available fluorescent light is not used. If this light could be used, the signal-to-noise ratio would improve by a factor of √2.
2. 2. The known method with the binary mask only supplies a unipolar signal, so that detection methods known from radar technology cannot be used.
3. 3. By using a single detector, its noise background always flows directly into the signal and thus falsifies the actual measurement signal and makes detection more difficult.
4. 4. In practice, the particle velocity is not known. In addition, when measuring several particles, a velocity distribution is observed, so that their dwell time in the detection area, i.e. the temporal extension of the measured signal, is not only unknown, but can also differ from particle to particle. Therefore, a wide speed range must be queried, in which the measured signal is correlated with various filters with different temporal extensions. This process is associated with considerable computational effort.

The DE 10 2013 015 016 A1 describes a method for particle detection in a particle-containing fluid. A first time-dependent measurement signal and a second time-dependent measurement signal are provided and these are correlated with one another.

Proceeding from the prior art described, it is therefore the object of the present invention to provide a method that reduces at least some of the disadvantages described.

According to the invention, this object is achieved by a method according to claim 1, in which a second signal is detected using a second structuring device, the second structuring device either ensuring that the particles along a trajectory section essentially only in second non-periodic spatial distances electromagnetic Radiation, an electric field or a magnetic field generate or at least influence or ensures that the electromagnetic radiation, the electric or magnetic field is detected essentially only in second, non-periodic spatial distances along the trajectory section, the first and the second Signal are correlated with each other and then conclusions about the number and / or the speed of the particles are drawn from the correlated signal.

According to the invention, a second signal is thus detected and the two detected signals are correlated with one another.

For example, the first and the second structuring device can use a non-periodic, structured mask and the particles can be guided past one side of the mask and the electromagnetic radiation passing through the mask, the electric field or the magnetic field can be detected on the other side. A beam splitter could be used for this purpose, which feeds part of the fluorescent light to the first structuring device, while the other part is fed to the second structuring device. The result is then two coded signals, which - if necessary after prior filtering - are correlated with one another.

By using two separate detectors, the noise floor of a single detector is reduced, improving the resolution of the method.

However, some of the signal is still lost when shadow masks are used. The disadvantage of the masked signal intensity can be overcome, for example, by using a spatially structured detector. A spatially structured detector could consist of two intermeshing surface sequences, each of which is assigned to a detector output. In this case, one detector output always supplies inverse signals to the other detector output, since those surface sections that are not assigned to one detector output and therefore represent a "0" for this are assigned to the other detector output and therefore represent a "1" for that one.

Such a structure generates an analog measurement signal. The disadvantage, however, is that the modulation sequence and therefore the coding are fixed for this detector due to the spatial structuring of the detector.

In a particularly preferred embodiment, a detector array is used as the first and/or second structuring device. A detector array consists of a number of detectors (pixel detectors) that can be read out selectively. A spatially structured code can be implemented by analogously or digitally combining individual pixel detectors (“binning”), as a result of which the structuring device can be easily modified and used in an application-specific manner. In addition to the advantage of simple software implementation, this detection scheme also has the advantage that the noise caused by the number of detectors is further reduced.

In addition, since the fluorescence signal is not masked out locally, but the entire information is retained, the combining of cells can be carried out in parallel in many different variations using software. In other words, the same signal can be processed in parallel with almost any number of codes, e.g. binary sequences are encoded, so that a plurality of encoded signals are available for correlation and subsequent evaluation.

The same measurement signal could thus be assigned to different masks. The multiple evaluation of the signal also allows various information to be extracted from the signal. For example, a particle detection with an autocorrelation of two spatially offset LABS could be carried out first and then the particle speed could be determined with a periodic binning pattern. LABS is understood to mean so-called “low autocorrelation binary sequences”, as described, for example, in Mertens S, Exhaustive search for low-autocorrelation binary sequences, J.Phys. A - Math. Gen. 29, L473 (1996).

Another preferred embodiment provides that the first and second structuring device are each designed such that when the trajectory section is divided into segments of equal length, the function of the first and second structuring device is written by a binary code, ie a sequence of zeros and ones can be, with each digit designating a segment and segments in which the respective structuring device either ensures that the particles generate an electric field or a magnetic field or at least influence or ensures that the electromagnetic radiation, the electric or magnetic field is detected will be denoted by the number "1" and all other segments will be denoted by the number "0".

The configuration of the structuring device is thus characterized by a sequence or a code consisting of zeros and ones.

In a further preferred embodiment, either the sequence of the second structuring device is the same as the sequence of the first structuring device read from back to front, or the sequence of the second structuring device is the same as the sequence of the first structuring device, in which case both sequences are then fed simultaneously to a respective FIFO buffer memory, the two FIFO buffer memories being arranged parallel to one another, the two sequences being supplied in opposite directions and respectively adjacent memory locations of the two FIFO buffer memories being correlated for the correlation.

In other words, the first memory location of the first FIFO buffer memory is correlated with the last memory location of the second FIFO buffer memory, with the second memory location being filled “from behind”. As a result of this measure, the time-inverse signal is correlated with the actual signal. This improves the signal-to-noise ratio and thus the detection of particles.

The result is irrelevant whether the time-inverse signal is generated using the second structuring device, or whether the second structuring device generates the identical signal as the first structuring device, but the signal determined by the second structuring device is sent in the opposite direction to a parallel FIFO buffer memory is entered.

In a further preferred embodiment, the sequence of the second structuring device is the inverse sequence of the sequence of the first structuring device. In this case, preferably either the signal detected with the first structuring device or the signal detected with the second structuring device is multiplied by a factor “−1” and then the two detected signals are combined with one another by addition to form a bipolar signal.

This bipolar signal can then be evaluated using conventional methods known from radar technology in order to detect a particle and determine its speed. For example, LABS can then be used and evaluated using known methods for narrow velocity distributions

In a further preferred embodiment, the sequence of the second structuring device is designed to be complementary to the sequence of the first structuring device. This means that the binary codes on which the structuring of the two structuring devices is based are complementary to one another. Binary codes that are complementary to one another have the essential property that the addition of their autocorrelation functions causes all signal sidelobes to disappear. Only the main peak adds up constructively.

A further improvement can be achieved if the bipolar signal is converted to a Manchester code before further processing. Manchester codes are defined, for example, in Foster R, Manchester encoding - opposing definitions resolved, Engineering Science & Education Journal 9 (6), 278 (2000).

For example, every "1" could be replaced with the sequence "1,-1" and every "-1" with the sequence "-1,1".

An important aspect when measuring intensities is the offset that results from detector noise, electronic noise and amplification and is contained in every signal. Bipolar codes offer the advantage that a constant offset voltage at the detector is largely eliminated if the binary sequence used has as many ones as inverse ones. Due to the autocorrelation (on the resonant velocity channel), a constant offset therefore adds up to zero, which is an advantage. In the case of complementary sequences, both codes usually have different signal energies, so that a DC offset does not disappear after autocorrelation and addition of the correlation signals. The Manchester coding makes the sequences twice as long, but this way it is secured ensures that both codes are always balanced, i.e. the sum of their coefficients is always zero and thus a time-constant detector offset is cancelled.

• 1 eine vereinfachte Form eines Messsignals mit der Amplitude 1 ohne Rauschen und Offset,
• 2 eine vereinfachte Darstellung eines überlagerten Messsignals,
• 3 das mit einem Optimalfilter gefilterte Messsignal,
• 4 das differenzierte Signal von 3,
• 5 eine schematische Querschnittsdarstellung eines Durchflusszytometeraufbaus,
• 6 eine schematische Darstellung einer ersten Ausführungsform,
• 7 eine schematische Darstellung eines Detektors einer zweiten Ausführungsform,
• 8 eine schematische Darstellung einer dritten Ausführungsform und
• 9 eine schematische Darstellung einer vierten Ausführungsform.

Further advantages, features and application possibilities become clear based on the following description of preferred embodiments and the associated figures. Show it:

• 1 a simplified form of a measurement signal with amplitude 1 without noise and offset,
• 2 a simplified representation of a superimposed measurement signal,
• 3 the measurement signal filtered with an optimal filter,
• 4 the differentiated signal from 3 ,
• 5 a schematic cross-sectional representation of a flow cytometer assembly,
• 6 a schematic representation of a first embodiment,
• 7 a schematic representation of a detector of a second embodiment,
• 8th a schematic representation of a third embodiment and
• 9 a schematic representation of a fourth embodiment.

The method according to the invention is illustrated using the example of a flow cytometer, as is shown schematically in 5 is shown, explained. The 5 shows a cross section. The flow cytometer has a microchannel 2 applied to a substrate 1, through which fluorescent objects 3 are moved. In doing so, they reach an excitation zone 4, which is illuminated by means of a laser beam, for example, whereby the objects are excited and emit fluorescent radiation. An optical amplitude grating 5, ie a non-periodic, binary filter 5, is provided in this area, through which fluorescence radiation can exit, which reaches the detector 7 via the optical filter 6. FIG. There essentially the in 2 receive the signal shown above. The signal thus consists of different square-wave pulses, the width of which corresponds to the width of the openings in the binary filter 5 .

In 6 a schematic representation of a first embodiment of the invention is shown. A fluorescent light beam 8 is split at a beam splitter 9 . A first portion of the fluorescent light beam is imaged through the mask 10 onto the detector 12 . A signal which can be represented by a sequence of “0” and “1” is thus received at the detector 12 . The second part is detected by the second detector 13, which is associated with a second mask 11, which is inversely configured to the first mask 10. FIG. The signal detected by the second detector 13 can also be represented by a sequence of “0” and “1”. The signal from the second detector 13 is multiplied by "-1" using a multiplier 14 . The two signals, ie the signal from the first detector 12 and the signal from the second detector 13 multiplied by "-1", are then correlated with one another, ie added in the present case using the adder 15 .

through the in 6 shown embodiment, a bipolar signal is generated, which is used in the further evaluation. Bipolar signals are much easier to evaluate than the previously used unipolar signals. In particular, the signal-to-sidelobe ratio can be significantly improved since the low-frequency components are suppressed by the bipolar character of the signal.

In 7 a schematic representation of a detector of a second embodiment is shown. In this embodiment, a spatially structured detector 16 with two separate detector outputs 17, 18 is used. It can be seen that the part of the detector surface 19 that is not hatched is assigned to the first detector output 17 and the other part that is hatched is assigned to the second detector output 18 . The interconnection of the area associated with the second detector output 18 thus provides the complementary signal related to the binary sequence.

By using a structured detector, the fluorescence light collected from the object is fully utilized, since no shading is necessary. Basically, the spatially structured detector differs from the shadow masks in that the shadowed components are now also detected on the detector and fed to the inverse signal. Here, too, it is possible to multiply the inverse signal by "-1" and add it to the original signal, so that a bipolar coded signal is generated.

Although the fluorescent light is fully utilized in this embodiment, the structured detector 16 has a predetermined, unchangeable structure.

In a third embodiment, which is shown schematically in 8th is shown, a detector 20 consisting of a large number of pixels 21 is therefore used, in which each pixel 21 can be read out selectively. By analogously or digitally combining individual pixel areas (also called "binning") it is therefore possible to implement a spatially structured code. The binning is done using the connected lines in 8th sketched. Thus, the individual pixels 21 of the detector 20 in this figure are combined into a first pixel area 22 and a second pixel area 23. With the help of the appropriate control software, almost any code can be displayed by appropriately combining pixel areas. In addition to the advantage of the simple software implementation, this embodiment also has the advantage that the noise caused by the large number of pixels of the detector 20 is further reduced.

In addition, since the signal is not masked out locally, but the entire information is retained, the combining of pixels can also be carried out in parallel in many different variations using software technology. The same measurement signal could thus be assigned to different masks. This is particularly advantageous for the class of so-called complementary binary codes, as described below, although any other type of code can be implemented as well. Due to the multiple parallel evaluation of the signal, it is also possible to extract different information from the signal. For example, in the first step, particle detection is possible by autocorrelation of two spatially offset LABS, and in the second step, the particle speed can then be determined using a periodic binning pattern. In particular, if the bipolar LABS is selected to be correspondingly short, the same LABS can be read out three times by means of the detector, for example, and the particle velocity can be determined from the difference in time between the correlation peaks.

By converting the initially unipolar signal into a bipolar signal, methods known from radar technology for particle detection and for determining the speed of the particles can be used. The measurement signal is advantageously converted into complementary binary codes. These have the essential property that the addition of their autocorrelation functions makes all signal sidelobes disappear. Only the main peak adds up constructively. Examples and definitions of complementary binary codes can be found, for example, in "Radar signals", Levanon, Nadav, and Eli Mozeson, John Wiley & Sons, 2004. and in M.J.I. Golay, "Complementary series", IRE Trans. Inform. Theory, IT-7 , pp. 82-87 (1961).

Another aspect that should not be neglected when measuring intensities is the offset that results from detector noise, electronic noise and amplification and is therefore contained in every signal.

The bipolar codes described here offer the advantage that a constant offset voltage at the detector is largely eliminated. Due to autocorrelation (on the resonant velocity channel), a constant offset adds to zero, which is beneficial. In the case of complementary sequences, however, both codes usually have different signal energies, so that a DC offset does not disappear in the correlation signal after autocorrelation and addition. One way to get around this is to Manchester encode the complementary sequences. This can be done, for example, by replacing each one "1" with the sequence "1, -1", while an inverse one "-1" is replaced with the digit sequence "-1, 1". Although this makes the sequences twice as long, it ensures that both codes are always balanced, i.e. the Manchester-encoded binary sequence has as many ones "1" as inverse ones "-1" and the sum of its coefficients therefore always results in zero and thus a time-constant detector offset is eliminated.

In 9 the structure of a corresponding detection device is shown schematically. The objects 3 to be detected move through a microchannel 2. In this microchannel 2, illumination optics 24, consisting of a laser (not shown), from which a laser beam 25 is emitted, an optical excitation light filter 26, a Powell lens 27 and an aspherical lens 28, an excitation zone uniformly illuminated. The objects 3 fluoresce as they pass through the excitation zone. The corresponding fluorescent light 29, after passing through an optical filter 30, strikes a gradient index lens (GRIN) array 31, which in turn directs the light onto corresponding macropixels 32 used for binning are responsible.

Reference List

1

substrate

2

microchannel

3

objects

4

excitation zone

5

optical amplitude grating (binary filter)

6

optical filter

7

detector

8th

fluorescent light beam

9

beam splitter

10

mask

11

mask

12

detector

13

detector

14

multiplier

15

adder

16

detector

17

first detector output

18

second detector output

19

detector surface

20

detector

21

pixel

22

first pixel area

23

second pixel area

24

lighting optics

25

laser beam

26

excitation light filter

27

Powell lens

28

aspheric lens

29

fluorescent light

30

optical filter

31

GRIN array

32

macro pixels

Claims (8)

Method for detecting particles (3) moving along a trajectory, which generate or at least influence electromagnetic radiation, an electric field or a magnetic field, wherein the electromagnetic radiation, the electric field or the magnetic field is detected, a first signal having Using a first structuring device (5, 10) is detected, the first structuring device (5, 10) either ensures that the particles (3) along a trajectory section essentially only in first non-periodic spatial intervals electromagnetic radiation, an electrical Generate a field or a magnetic field or at least influence or ensure that the electromagnetic radiation, the electric or magnetic field is essentially only detected in first non-periodic spatial distances along the trajectory section, characterized in that a second signal with the aid of a second structuring device (11), the second structuring device (11) either ensuring that the particles (3) along a trajectory section essentially only generate electromagnetic radiation, an electric field or a magnetic field at second non-periodic spatial intervals or at least influence or ensure that the electromagnetic radiation, the electric or magnetic field is essentially only detected in second non-periodic spatial intervals along the trajectory section, that the first and the second signal are correlated with one another and conclusions can be drawn from the correlated signal on the number and/or the speed of the particles (3) can be drawn. procedure after claim 1 , characterized in that a non-periodic, structured mask (10, 11) is used as the first and/or second structuring device (10, 11) and the particles (3) on one side of the mask (10, 11) at a speed v are passed and on the other side through the mask (10, 11) passing electromagnetic radiation, the electric field or the magnetic field is detected. procedure after claim 1 or 2 , characterized in that a detector array (31) is used as the first and/or second structuring device. Procedure according to one of Claims 1 until 3 , characterized in that the first and the second structuring device (10, 11) are each designed in such a way that when the trajectory section is divided into segments of equal length, the function of the first and second structuring device (10, 11) is in each case defined by a sequence of zeros ( "0") and ones ("1") can be described, with each digit identifying a segment and segments in which the respective structuring device (10, 11) either ensures that the particles (3) have an electric field or a generate or at least influence a magnetic field or ensure that the electromagnetic radiation, the electric or magnetic field is detected are identified by the number "1" and all other segments are identified by the number "0". procedure after claim 4 , characterized in that either the sequence of the second structuring device (11) is the same as the sequence of the first structuring device (10), read from back to front, or the sequence of the second structuring device (11) is the same as the sequence of the first structuring device (10), and both sequences are simultaneously fed to a respective FIFO buffer, the two FIFO buffers being arranged parallel to one another, the two sequences being fed in opposite directions and for the correlation in each case adjacent memory locations of the two FIFO buffer memories are correlated. procedure after claim 4 , characterized in that the sequence of the second structuring device (11) is the inverse sequence of the sequence of the first structuring device (10), with preferably either the signal detected with the first structuring device (10) or the signal detected with the second structuring device (11). signal is multiplied by a factor of "-1" and then the two acquired signals are correlated by addition to form a bipolar signal. procedure after claim 4 , characterized in that the sequence of the second structuring device (11) is designed to be complementary to the sequence of the first structuring device (10). procedure after claim 6 , characterized in that the bipolar signal is converted into a Manchester code before further processing.

Images