Method and apparatus for processing detector signals

The invention relates to a method for processing Dietektorsignalen and an apparatus for carrying out the method.

German Patent DE 42 26 175 C2 and US Patent US 5,307,299 describe a method of digitally processing Dektektorsignalen a semiconductor detector.

In a semiconductor detector each Strahlungoereignis unleashes the abeorbierten energy corresponding amount of charge that is collected in a pre-amplifier. The output signal of the preamplifier can be viewed as a convolution between the charge distribution function, and the preamplifier.

German Patent DE 42 26 175 C2 and US Patent 5,307,299 describe a particular method that makes the fold in the output reversed and as sliding deconvolution ( "Moving Window deconvolution") is referred to.

In this method, the successive samples in an over the sliding window with a certain number in the distance of a sample interval following samples is recursively determines the total accumulated charge from the digitized output of the preamplifier, from which the total energy of the incident radiation can be reconstructed.

German Patent DE 42 26 175 C2 and US Patent US 5,307,299 further describes a circuit arrangement for evaluating detector signals that generate at the detectors detector signals amplified by preamplifier and subsequently for example, about 2 meters to 200 meters long analog Sighalübertragungskabel to scanning analog-to-digital converter to be transmitted. Behind the scanning analog-to-digital converters (ADU) are connected in digital signal processing units. The digital signal processing units to process, depending on the performance of the continuous data stream of a single or multiple scanning analog-to-digital converter (ADU). Data of various digital signal processing units are then collected in these known methods by a central data acquisition unit DAQ

The European patent application EP 0166169 A describes a process method and an electronic device for the evaluation of successive detector events. A typical application is the measurement of detector events, such as a gamma camera, said detector events in the form of temporally successive pulses are applied. This series of pulses results in a pulse accumulation, which makes the provision of a single pulse in the midst of the series problematic. For this reason, the particular type of signal detection in the present case is characterized in that is used for detecting more than a sampling frequency which is a harmonic of a first sampling frequency. In addition, the time of commencement Detention a function of the additional sampling frequencies is known. From this information, and utilizing a processing of a number of measured data points per signal detection, the individual successive signals are processed further in relation to the respective preceding signal. Thereby, the accuracy of detection of individual Detektorereigniesen is increased.

The presented in the US patent US 5,608,222 A method allows the use of non-linear detectors, for example, for gamma rays spectroscopy, while reducing the caused by the non-linear detectors distortions of the spectral resolution. In particular cause non-linear detectors for changing the measured spectral release as a function of the irradiated energy of the detector event, and this undesirable effect in the process according to US 5,608,222 is A reduced by an appropriate transfer function. For this, the detector event is mathematically set with a predefined transfer function in relation so that any non-linearities in the receiver system does not influence the subsequent determination of the spectral events.

The US patent US 5,347,129 discloses a method by which a particular detector event can be extracted by Korrelationaanalyse with a predetermined reference signal from a sequence of detector events. The procedure is in this case characterized by the following process steps substantially. First occurs immediately following the detection of a pre-amplification of the detector events. A subsequent analog-to-digital converter digitizes the data, the digitized data over an interface and a DSP (<i>Digital Signal Processing</i>) Unit will be made available to a PC. The PC takes on the actual expiry of the correlation analysis, by having recourse to an earlier document stored in memory reference signal.

The US patent US 2,550,488 discloses an apparatus for registering events, especially events caused by alpha particles, beta particles, gamma radiation or neutron radiation. Galvanic isolation between an input signal, the event and an output signal is obtained by means of a gas-tetrode.

The European patent application EP 0981997 A discloses an imaging, X-ray generating device, in particular a computed tomography apparatus for generating in vivo or <i>in vitro</i> Slice images of an object, the apparatus operates in a Mehrechichtakquisitionsmodus. A special feature of this arrangement is the use of a multiplexer for combining the signals of individual radiation detectors, so that only a data bus for transmitting the signals is necessary.

The device described in US Patent US 5,535,033 A, enables a contact-free data transmission between a transmitter and a receiver, in particular between a mobile and a stationary part of a computer tomograph. For this purpose, at least one lateral emitting optical fiber cable is used, coupled with an opto-electronic element.

Further, it is known to detect by detector signals handsets. The handsets each have a single detector. A data transmission path is not provided here, because the detection and detecting means are combined in a housing.

The invention is based on the object, a generic process further so that an influence of interference is avoided.

This object is achieved by a method having the features of claim 1 and by an apparatus having the features of claim. 5 Expedient developments of the method and apparatus are the subject matter of the dependent claims.

It is particularly provided that analogue measurement signals immediately after a pre-amplification through a scanning analog-to-digital converter (ADU) are converted into digital signals and transmit the digital signals subsequently via a digital data transmission link to a digital signal processing unit and a data acquisition unit (DAQ) will.

The invention provides for converting analog signals from detectors in the detectors into digital signals and then transmit the digital signals over a data transmission path.

Here, it's possible to transmit unchanged or processes the digital signals.

The resulting from the conversion of the detector signals into digital signals digital signal stream can be further processed using various methods.

The further processing of the digital signals can be carried out both before the data transmission path in the detectors as well as to the data transmission path in the field of data acquisition unit (DAQ).

A first advantageous embodiment of the invention is characterized in that the digital signals are substantially unchanged to the digital signal processing unit and data acquisition unit (DAQ) are transmitted.

A further advantageous embodiment of the invention is characterized in that the digital signals are subjected to prior to transmission over the data transmission link at least one digital signal processing step.

The digital signal processing ride search can be done in many ways, for example by a programmable logic device (PLD).

The implementation of the digital signal processing step before the data transmission over the data transmission path makes it possible to reduce the number of over the data transmission path to be transmitted.

A particularly effective reduction of the data stream can be achieved in that the signal processing step comprises a triggering of the digital signals. Triggering means that the areas scanned by the ADU and digitized waveform are marked which contain information about the detection event.

This makes it possible, first, the respectively to be examined event - filtering out from the data stream and then to transmit the information about the respective event - in particular, a detection event.

Further, it is for reducing the data stream appropriate to that in one or more signal processing steps, a power information is extracted about the detection event from the digital samples.

It is particularly advantageous due to the characteristics of the detector signals to use a moving window deconvolution technique.

Furthermore, it can reduce the data stream is advantageously achieved in that information on interaction places of detection events in the detector are determined by further processing steps.

Furthermore, it is provided in particular to embody a device for processing detector signals with pre-amplifiers for amplifying detector signals, with a scanning analog-to-digital converters (ADU) for converting the detector signals into digital signals, with a signal processing unit and a data transmission line so that the scanning analog-to-digital converter (ADC) are located between the pre-amplifiers and the data transmission path.

A particularly compact and convenient design of the device can advantageously be achieved if there are between the scanning analog-to-digital converters (ADC) and the data transmission path signal processing units.

Here, a particularly expedient because of the flexibility of programmability and its compactness embodiment, is formed in at least a part of the digital signal processing units using digital signal processors (DSP), programmable logic devices (PLD), or field programmable gate arrays (FPGA).

A reduction in the number of required data transmission links and digital signal processing units is achieved in a particularly advantageous embodiment of the invention, that a plurality of scanning analog-to-digital converter (ADU) are connected to inputs of a multiplexer, which combines the signals and to a common digital signal processing unit übertr ägt.

Further, it is preferable that outputs of several digital signal processing units are connected to a multiplexer, wherein the multiplexer transmits the data output from the digital signal processing units signals in a data transmission channel. This makes it possible to transmit yet complete information on the detection signals in data transmission lines with a smaller number of data transmission channels.

A reduction of interference can be achieved in that the transmission path includes at least one optical fiber in a particularly advantageous and expedient manner. This is additional galvanic isolation between the signals acquisition and data collection.

Further advantages, special features and practical refinements of the invention result from the dependent claims and the description of preferred embodiments.

The invention will be explained using the example of detectors for γ-radiation.

However, the invention is not limited to the detectors shown, since the signal processing illustrated can be carried out regardless of the type of the detectors.

In the following illustration, the detection events are detected and processed together with energy values, times and places.

It is expedient that digital signal processing steps are used and combined.

A signal processing step is the trigger. By triggering a first significant reduction of the data flow is preferably achieved in particular by the suppression of the areas of the sampled and digitized by the ADC signal path which do not contain relevant information about the detection event.

In one implementation of the trigger in a programmable device, the triggering can be varied by reprogramming of the block.

So for example it is possible to adapt the device for processing the detection signals and the associated method to each experimentally desired parameters.

Another advantageously acquired information is the energy of the detection event.

In order to derive energy information to various process steps which include a reduction of the full data stream from one sample per sampling on only one energy value per detection event are.

In order to reduce the required data transfer and processing power, it is possible to perform determination of the energy before a sub-sampling (decimation) or averaging over several samples of the digitized detector signals.

For the detection of time information, it is expedient to take into account process steps that enable a determination of time information.

In this case, preferably, the knowledge is utilized, that it is sufficient, only selected samples, especially those from the region of the rising edges of the digitized detector signals to analyze to determine the timing of the detection event.

For determining the interaction sites of detection events a variety of processes, or process steps is also suitable.

Preferably also in this case the shape of the rising edges of the digitized signals of detection events is evaluated. Moreover, it is advantageous to use wavelet transforms for the extraction of the relevant properties of the signal.

In the determination of the interaction sites of detection events the data of multiple signal processing channels are combined so that there is a further data reduction.

In the illustrated examples, a combination of pre-amplifier and scanning analog-to-digital converter (ADU) is referred to as a digital pre-amplifier, since this digital signals are obtained from the original analog measuring signals.

When transferring the digitized signals, it is usually possible to be limited to areas of the signal, containing relevant information, or to pre-process the signals in digital signal processing units, resulting in a reduction of over the data transmission path to transfer data when detector arrays with the one plurality of detectors makes it particularly noticeable.

To illustrate examples are mentioned below, in which concrete numbers (ND: = Number of Detectors) are called by detectors, the invention is not limited to certain numbers.

A detector array for example 180 segmented detectors each example, 32 segments (NS: = Number of segment) and a central contact widen, results of example the number of 5,940 data processing channels (NC: = Number of Channels). The other values ​​shown here are also meant to be exemplary for explaining the calculation examples.

In the calculation examples, the scanning analog-to-digital converter (ADC) have a data width of 14 bit vo (CR: = Conversion Range) and a sampling rate (SF: = Sampling Frequency) of 50 MSPS to.

Further, in the examples are a maximum detection rate of events (EN: = Detector Event rate) of 20,000 Hz Zählereiguissen, an average number of 4 interactions per detection event (NI: = Number of Interactions), a number of 10 active segments per detection event (NA: = Number of Active segments) per detector and an external trigger rate (TR: = trigger rate), for example, 100 kHz or 1 MHz adopted.

It is also felt in the examples of a relevant for the pulse shape analysis charge collection time of the detector of about 400 ns, a maximum number of 25 samples for the leading edge portion of the waveform of the detection event corresponds (NL: = Numher od Leading Edge samples).

The data width of the output data is in Examples 48 bit for the timestamp (WTS: = Word width of Time Stamp), 16 bit for the energy (WEN: = Word width of Energy), 8 bits for specifying the exact point in time within a sampling (WTI: = Word width of timing) and 32 bit for the position information (WPO: = Word width of position).

Here, the 32 distributed bits for the position data to 8 bit for the detector number and each 6 bit for the segment number, the radius, the corresponding azimuthal angle and the axial z-coordinate of a detector.

Example 1:

In a preferred embodiment of the invention of the digital preamplifier accepts only pre-amplification and digitization of the detector signal, while the other signal processing steps are carried out after transmission of the data via the data transmission link in a signal processing unit.

The data transfer rate amounts for each data transmission channel CR * SF = 87.5 MByte / s.

The data of the individual data processing channels can be output via individual data links. Alternatively, a plurality of data transmission paths can be combined with a smaller bandwidth to a data transmission path with a high bandwidth through the use of one or more multiplexers.

In the illustrated case where the data channels of a 32-fold segmented detector are combined, the total required bandwidth of the transmission path of a detector is 2.89 GB / s plus transmission protocol data. For higher sampling rates and / or a higher number of detector segments, the required bandwidth can be considerably higher. Appropriately, the bandwidth is matched to the characteristics of each detector array.

By combining data links the scope of self-employed for the actual transfer digital components is reduced, particularly for the data format matching, protocol adaptation and for the line driver.

In contrast to the known methods in which analog signals on the data transmission path to be transmitted, this system is much less sensitive to interference signals.

To further reduce interference, it is advisable to separate the analog processing of the signals in the sensitive analog part of the preamplifier of their digital processing. This is particularly advantageous in that the digital data transmission path includes at least one light guide, or otherwise electrically isolated.

In a further preferred embodiment of the invention, additional signal processing steps in the digital pre-amplifier are integrated.

The following examples include a broad application of signal processing steps prior to the transmission of the data over the data transmission path, so that the data stream, and thus the required bandwidth of the data transmission paths as well as the load on the subsequent signal processing units can be reduced.

To most effectively and flexibly perform these signal processing steps, it is expedient to use the digital preamp Programmable Logic Devices (PLD), Field Programmable Gate Arrays (FPGA), or digital signal processors (DSP) in the region.

Example 2:

For signal processing steps such as the moving window deconvolution (MWD) for extracting the energy of detection events, which do not require the full bandwidth of the detector signal, a sub-sampling (decimation) can be effected by an appropriate factor. In the illustrated case, the factor is 3, but this is only an example to understand. This reduction factor is CR '* SF / 8 = 13.3 MByte / s as required bandwidth per signal processing channel for the transmission of energy information.

To further reduce the data stream triggering the digitized detector signals can be used in this method.

Triggering of the signals can take place either through a self-triggering on channel or detector plane, as well as by a trigger on a global scale with an external trigger signal, wherein the different trigger methods can be combined.

The Selbsttriggerung is particularly suitable to reduce the data stream in the sense that only a few samples from the region of the rising edge of the signals of detection events are taken into account, while the vast remainder is not transmitted.

Since a Selbsttriggerung it is controlled by the data itself, in this case, the resulting data stream depends on the number of detection events DE.

In a Selbsttriggerung on detector plane, it is necessary, all data channels NC 'of a detector, that is, the segments 32 plus the contact of the central read out. In the case illustrated applies NC '= NC / ND = 33rd

This produces for each 25 samples NL from the leading edge region of the signals of the detection events DE a data rate of NC '* DE * NL * CR = 28.9 MByte / s per detector.

For the samples (decimated) for power provision NC '* 13.3 = 438.9 MB / s.

DE For the stream of time marks * WTS = 0:12 MByte / s.

The total data stream each detector is in the illustrated example 468 MB / s, which is a reduction by a factor of 6.2 compared with the example i.

By selectively trigger on active individual channels and / or by zero suppression, that is, suppression of channels that do not contain signals of detection events, it is possible to prevent reading out of channels that do not contain any information relevant for the measurement. For the reading of samples from the leading edge area of ​​the signals of detection events, this results in a further reduction of the data stream by a factor of NA / NC '= 0.3.

With such a reduction but then identification information must be mitübermittelt on the active segments, which again * 6 bit = increases the data stream slightly by DE * MA 00:15 MByte / s.

The generation of a global trigger signal is preferably performed by an external device.

Triggering from an external device requires an addition of corresponding inputs and outputs to the digital preamplifiers, and a galvanic decoupling of the lines via which the trigger signals are received.

External triggering has the advantage, that the counting rate of the read out detection events to trigger the effective rate divided by the total number of detectors, that is, for example, to DE_H = (TR_H * 30) / ND = 16.7 KHz for a 100 KHz trigger signal TR_HWhich acts on detectors 30 (high multiplicity trigger), respectively on DE_L = (TR_L * 1) / ND = 5.6 KHz for a 1 MHz trigger signal TR_LWhich acts on only one detector (low multiplicity trigger) is reduced.

This results in a further reduction of the data transmission rates for the samples from the leading edge region of the waveforms of detection events, in the case illustrated results to 24.1 MB / s for the high multiplicity trigger, or at 8.0 MB / s for the low multiplicity trigger.

Example 3:

In Example 3 explains how the data transfer rate can be further reduced through the implementation of additional functions into the digital signal processing units in the field of digital preamp.

This requires that the digital signal processing units complex functions, like moving window deconvolution technique used. The resulting analysis carried out of the energy of the detection events leads to a particularly drastic reduction in the data transmission rate. In the illustrated case, the reduction NC '* CR' * SF / 8 = 438.9 MB / s to NC '* DE * WEN = 1:32 MByte / s, because of several hundred samples only a respective energy value WEN is extracted that must be transferred ,

It is also possible by the energy analysis, data channels, the energy information included those which do not contain any energy information to distinguish.

By selectively triggers and reading of energy values ​​from only the data channels containing energy information, in the case illustrated in ((NS / 3) +1) * DE * WEN = lets the data transfer rate further reduce 0:17 MByte / s.

Together with the appropriate stream for the leading edge area of ​​the signals of detection events of NA * DE * NL * CR = 8.7 MByte and the information to be transmitted over the timings of detection events, and to identify the active detector segments results in a total data stream of about 9 MByte / s per Detector.

This represents a very significant reduction compared to the initial situation, in which the entire data stream was 2.89 Gbyte / s per detector.

As described above can reduce the Detektionsereigniszählrate EN further by external triggering, for example, by a factor of DE_H / DE = 0.84 or DE_L/ DE = 12:28.

The examples illustrated show preferred embodiments of the invention. In particular, the examples show advantages arising from the fact that already are converted in the immediate vicinity of the detectors into digital signals, the detector signals. This shows that both the number of required communication links as well as an influence of noise is reduced.

To further reduce the number of required data transmission links and their transfer bandwidth by reducing the information to be transmitted data is carried out in the range of the detectors and located there scanning analog-to-digital converter (ADC) in each of the Examples of the invention, a pre-processing of the digitized signals.

The pre-processing of the signals shown is only possible after their digitization, but meant to be exemplary, as other Vorverarheitungsschritte which enable elimination of unimportant for signal processing and a reduction of the interest to the signal processing data, can be used.

Particularly preferred embodiments of the invention include the concept to make this a first evaluation of the data in the field of scanning analog-to-digital converter (ADC) and associated with them digital signal processing units.

By selection of suitable electronic components, it is possible to adjust the respective procedures to experimental needs. Preferably to programmable digital components are used.

It is particularly advantageous that individual elements of the invention can be combined as required.

In this way further adaptation of the method and suitable for carrying out the method to different detector arrangements and expected characteristics of detection events is possible.