Method for processing an input data stream and mass spectrometer

Abstract

Method for processing an input data stream including at least one data spike, comprising: Detecting at least one peak in the input data stream having a peak with an amplitude above a predetermined threshold, wherein the input data stream includes a signal output from an ion detector; Extrapolating the segment of the tip that has an amplitude above the predetermined threshold, based on a shape characteristic of the tip, to estimate the amplitude of the segments of the tip that have a lower amplitude than said threshold; Generating a filtered data stream, comprising applying the predetermined threshold to the input data stream such that the filtered data stream contains only the segment of the peak above the predetermined threshold; and Compiling a compensated data stream including filtered data and extrapolated peak data.

Description

[0001] The present invention relates to a method for processing an input data stream comprising a signal output from an ion detector, and to a mass spectrometer.

[0002] A common method for obtaining a mass spectrum involves recording the output signal from an ion detector of a mass analyzer as a function of time using a fast analog-to-digital converter (ADC). The use of an ADC with a scanning magnetic sector mass analyzer, a scanning quadrupole mass analyzer, or an ion trap mass analyzer is known.

[0003] The ADC records the signal intensity (amplitude) at each clock cycle T nConventional ADCs can digitize a signal at a rate of, for example, 40 MHz and simultaneously record the signal intensity as a 16-bit digital value. In a TOF-based system, 2 GHz may be a typical rate, and the signal intensity can be recorded as an 8-bit digital value.

[0004] An ADC generates a data stream comprising a continuum intensity profile as a function of time corresponding to the signal (current) output from the electron multiplier, as shown in Fig. 1 illustrates.

[0005] An advantageous feature of an ADC system is its ability to record multiple simultaneous ion reception events by outputting an amplified intensity value. Therefore, multiple ion reception events can be accurately recorded using an ADC.

[0006] ADCs suffer from the problem that any noise (e.g., electrical noise, such as spontaneous detector events caused by cosmic rays, etc.) is also digitized. A distinction must be made between background noise current and current due to ion impingement.

[0007] It is known to apply a threshold to an input data stream to reduce or compensate for the effects of noise. In this thresholding technique, whenever the amplitude of the measured data stream is below a predetermined threshold, that portion(s) of the input data stream is ignored, set to zero, or otherwise marked during subsequent processing of the input data stream. Only the segment of the input data stream where the amplitude is equal to or above the threshold is of interest (and is subsequently summed to generate a final mass spectrum or m / z chromatogram).

[0008] Fig. Figure 1 illustrates the output of an ADC and shows a single ion event. The input data stream 1 comprises a single (first) peak 2 having a crest 11 and also includes background noise 3 and spontaneous detector events 12 (false peaks). The mean (average) 9 of the baseline of the input data stream 1 is also shown in Fig. 1. The baseline is the part of the data stream, not including ion events (spikes) or spontaneous detector events, that lies within the Fig. 1. The peak 2 has a leading edge 7 and a trailing edge 8, on either side of the apex 11.

[0009] According to a known technique, a threshold 4 is set to a predetermined level. The level of threshold 4 is preferably high enough to ignore expected background noise 3 and spontaneous detector events 12, but low enough not to inadvertently ignore genuine ion peaks 2. Inevitably, when using this technique, a segment 5 of the leading edge 7 and a segment 6 of the trailing edge 8 of a peak 2 are below the predetermined threshold 4. A segment 10 is above the threshold and includes part of the leading edge 7 and part of the trailing edge 8.

[0010] When applying the known technique, any parts of the input data stream 1 with an amplitude (intensity) below the predetermined threshold 4 are ignored (or set to zero or a constant or marked). The mean value of the baseline 9 of the input data stream 1 can also be subtracted from the input data stream 1. The resulting filtered data stream 20 is shown in Fig. 2 and includes only the segment 10 of the peak 2 in the input data stream 1 that is above the predetermined threshold 4.

[0011] By ignoring some of the peak data, the thresholding technique inevitably leads to inaccuracies in subsequent centroid detection and calculation of the area under peak 2. It should be noted that the area under segment 6 in Fig. 1 represents a significant part of the area under the tip 2 as a whole.

[0012] Where ion responses (peaks 2) are separated in time and where the shape and peak amplitude of each peak 2 is essentially the same, these inherent inaccuracies can be essentially constant. The resulting error can be tolerated and / or reliably compensated. For example, a correction constant (or variable) can be applied to the data for each detected peak.

[0013] However, where an input data stream 1 includes multiple ion events 2 (spikes) that at least partially overlap, segments of a spike 2 that would have been below the threshold if the spike 2 had been separated from other spikes 2 in the input data stream are now 'pushed' above the threshold when combined with a segment of a neighboring, overlapping spike 2. Accordingly, this portion of the input data stream is included in subsequent processing of the data.

[0014] This results in a difference in the mean centroid and / or tip area calculated for a given peak 2, depending on whether it is separated from all other peaks 2 or overlaps with at least one other peak (and the extent of the overlap).

[0015] It is common for a given input data stream to contain both separate and overlapping peaks, and one requirement is that such input data streams be processed and analyzed precisely. However, applying known thresholding techniques to such input data streams inherently introduces inaccuracies depending on whether the peaks are overlapping or separate.

[0016] Similar problems occur in time-of-flight (TOF) mass spectrometry, which can lead to inaccuracies in quantification and peak centroiding.

[0017] US 2007 / 0 023 642 A1 describes an apparatus and a method for processing mass spectrometric data.

[0018] An improved procedure is needed.

[0019] Accordingly, the present invention provides a method for processing an input data stream including at least one data peak according to claim 1, comprising: detecting at least one peak in the input data stream having a crest with an amplitude above a predetermined threshold, and extrapolating the segment of the peak having an amplitude above the predetermined threshold based on a shape characteristic of the peak to estimate the amplitude of the segments of the peak having a lower amplitude than said threshold.

[0020] Preferably, the step of detecting at least one peak comprises detecting at least one of the leading edge, the apex, or the trailing edge of the peak.

[0021] According to the invention, the method further comprises generating a filtered data stream, comprising applying the predetermined threshold to the input data stream such that the filtered data stream contains only the segment of the peak above the predetermined threshold.

[0022] Preferably, if the amplitude of the input data stream at a time T n less than said threshold, the amplitude in the filtered data stream at time T n set to either zero or a constant.

[0023] According to the invention, the method further comprises: compiling a compensated data stream including the filtered data and extrapolated peak data.

[0024] Preferably, the method further comprises summing the estimated amplitude of extrapolated peak data and the filtered data stream over time.

[0025] Preferably, the method further comprises estimating a shape characteristic of the tip.

[0026] Preferably, the step of estimating a shape characteristic of the peak includes evaluating the segment of the peak above the predetermined threshold.

[0027] Preferably, the step of extrapolating the peak comprises detecting the time point, T T at which the amplitude of the trailing edge of the peak in the input data stream falls below the predetermined threshold; and estimating the amplitude of the peak at time T T by applying a decay function, based on the shape characteristic, to the amplitude of the data peak at time T T-1 , above the predetermined threshold.

[0028] Preferably, the method further comprises estimating the amplitude of the peak at time T n by applying a decay function based on the shape characteristics of the peak to the estimated amplitude of the peak at time T n-1 .

[0029] Preferably, the method further comprises iteratively applying the decay function for all values ​​of n.

[0030] Preferably, the method further comprises repeating the step of iteratively applying a decay function until the amplitude of the input data stream at time T n is greater than or equal to the predetermined threshold or the estimated amplitude at time T n essentially zero, the baseline mean, or a constant.

[0031] Preferably, the step of applying the decay function comprises multiplying the amplitude of the peak at time T n-1with a constant decay value between 0 and 1.

[0032] Preferably, the step of extrapolating the peak comprises detecting the time point, T L at which the amplitude of the leading edge of the peak in the input data stream rises above the predetermined threshold; and applying a growth function based on the shape characteristic to the amplitude of the data peak at time T L, to obtain an estimated amplitude of the peak at time T L-1 to generate.

[0033] Preferably, the method further comprises estimating the amplitude of the peak at time T n-1 by applying a growth function based on the shape characteristics of the peak to the estimated amplitude of the peak at time T n .

[0034] Preferably, the method further comprises iteratively applying the growth function for all values ​​of n.

[0035] Preferably, the method further comprises repeating the step of iteratively applying a growth function until the amplitude of the input data stream at time T n-1 is greater than or equal to the predetermined threshold; or the estimated amplitude at time T n-1 less than or substantially equal to zero, the baseline mean, or a constant.

[0036] Preferably, the growth function is linear.

[0037] Preferably, the amplitude at time T n-1 estimated by subtracting a predetermined constant from the estimated amplitude at time T n is subtracted.

[0038] Preferably, the growth function is based on the rate of change of at least a portion of the segment of the input data stream that is above the predetermined threshold.

[0039] Preferably, if the input data stream includes two or more peaks and extrapolation of the respective peaks produces conflicting estimates for the amplitude at a time T, the highest of these estimates is selected for time T.

[0040] According to the invention, the input data stream comprises a signal output from an ion detector.

[0041] Preferably, the input data stream comprises a voltage signal, an ion signal, an ion current voltage pulse, or an electron current pulse.

[0042] Preferably, the input data stream comprises the output of an analog-to-digital converter of an ion detector of a mass spectrometer.

[0043] Preferably, the method further comprises applying a smoothing function to the input data stream using a finite impulse response or infinite impulse response filter.

[0044] Also disclosed is a method for processing an input data stream including at least one data peak, comprising: extrapolating the segment of a peak having an amplitude above the predetermined threshold based on a shape characteristic of the peak to estimate the amplitude of the segments of the peak having a lower amplitude than said predetermined threshold.

[0045] The present invention further provides a mass spectrometer according to claim 23, comprising an ion source; an analyzer; an ion detector generating a data stream; and a processor configured or programmed to: detect at least one peak in the data stream having a crest with an amplitude above a predetermined threshold; and extrapolate the segment of the peak having an amplitude above the predetermined threshold based on a shape characteristic of the peak to estimate the amplitude of segments of the peak having a lower amplitude than said threshold.

[0046] The present invention further provides a computer program or programmable logic element comprising computer readable program code means or programmed logic for causing a processor to execute a procedure for implementing the method of the present invention.

[0047] Preferably, the computer program element is embodied on a computer-readable medium.

[0048] The present invention further provides a computer-readable medium having a program stored thereon, wherein the program causes a computer to execute a procedure for implementing the method of the present invention.

[0049] Embodiments of the present invention will now be described, by way of example only, with reference to the figures, in which: Fig. Figure 1 illustrates an input data stream showing a single ion response; Fig. Figure 2 illustrates a filtered signal obtained by applying a predetermined threshold to the input data stream of Fig. 1 generated data stream; Fig. 3 illustrates a compensated data stream compiled using a method embodying the present invention; Fig. 4 and Fig. 5 illustrate an input data stream showing two overlapping ion responses; Fig. 6 illustrates a compensated data stream compiled using a method embodying the present invention; Fig. Figure 7 illustrates a data stream showing three overlapping ion responses. Fig. 8 illustrates part of the data stream from Fig. 7; and Fig. Figure 9 illustrates another data stream.

[0050] As discussed above, Fig. 1 an input data stream 1 containing noise 3 and spontaneous detector events 12. The present invention provides a method comprising detecting at least one peak 2 in the input data stream 1. A peak 2 is detected if it has a crest 11 with an amplitude above the predetermined threshold 4.

[0051] Fig. Figure 2 illustrates a filtered data stream 20 corresponding to the input data stream 1 of Fig. 1 after filtering using the predetermined threshold 4. Note that neither the noise 3 nor the spontaneous detector events 12 are included in the filtered data stream 20. In Fig. 2, the segments 5, 6 of the input data stream 1, in which the amplitude is below the predetermined threshold, were set to 0. Alternatively, those segments 5, 6 can be completely ignored, or marked.

[0052] With reference to Fig. 1 and Fig. 2, the leading edge crossing point 13 is the point at which the leading edge 7 crosses the threshold 4 (i.e., the amplitude increases from below the threshold 4 to above the threshold 4). Similarly, the trailing edge crossing point 14 is the point at which the trailing edge 8 crosses the threshold 4 (i.e., the amplitude decreases from above the threshold 4 to below the threshold 4).

[0053] A method embodying the present invention extrapolates the data peak 2 of the filtered data stream 20. In particular, a method embodying the present invention extrapolates the segment of the peak 2 above the predetermined threshold 4. The data peak 2 is extrapolated based on a shape characteristic of the peak 2 to estimate the amplitude of the peak 2 when the amplitude of the input data stream 1 is less than the predetermined threshold 4. The extrapolation is performed on both sides of the data peak 2; that is, for both the leading edge 7 and the trailing edge 8. From Fig. 1 shows that the profile of the leading edge 7 of the data tip 2 is different from the profile of the trailing edge 8 of the data tip 2.

[0054] Ion detection systems generally include avalanche devices for multiplying the incoming individual ions into sufficient electron current to be easily detected by the facility's electronic detection system. Existing detection systems include, for example, a photomultiplier tube, a multi-channel plate (MCP), or an electron multiplier. The invention is described only with reference to a PMT-based system. The photomultiplier tube is effectively a pulsed output current source, and therefore the waveform at the ADC input comprises a very rapid rise to a peak value, which then decays essentially exponentially. For example, the output may rise from the mid-level (or zero) to a peak value over only one ADC sample time unit; and may then decay to 90% of its previous value with each additional ADC sample time unit.

[0055] As a result, the profile of the leading edge 7 of the data tip 2 is Fig. 1 is essentially linear. The profile of the trailing edge 8 of the data peak 2 is essentially exponential. Methods for extrapolating both the leading edge 7 and the trailing edge 8 will now be described.

[0056] Fig. Figure 3 illustrates a compensated data stream 30, wherein the filtered data stream 20 of Fig. 2 has been extrapolated using a method embodying the present invention. Leading edge extrapolation

[0057] Referring to Fig. 1, only the segment 10 of the data peak 2 above the predetermined threshold 4 is stored for subsequent operation - ie for generating the filtered data stream 20 of Fig. 2. A method embodying the present invention uses the rate of increase of the amplitude of the leading edge 7 of the peak 2 when above the threshold 4 to estimate the amplitude of the leading edge 7 when below the threshold 4.

[0058] It is understood that this method involves backward extrapolation. Therefore, the estimated amplitude of the input data stream 1, if below the predetermined threshold 4, can only be calculated after the amplitude of the input data stream 1 has risen above the predetermined threshold 4.

[0059] A method for extrapolating the leading edge 7 of a data peak 2 embodying the present invention comprises identifying the leading edge crossover point 13 of a data peak 2. The leading edge crossover point 13 is identified when the amplitude of the data peak 2 increases from below the predetermined threshold 4 to above the predetermined threshold 4. The time at which the leading edge crossover point 13 occurs is referred to herein as T L referred to (see Fig. 2).

[0060] A method embodying the present invention subsequently estimates the amplitude of the data peak 2 in a previous time interval, T L-1 , by applying a growth function to the measured amplitude of data peak 2 at time T L The growth function is based on a shape characteristic of the tip. The amplitude of the leading edge 7 at time T L-1 is lower than the amplitude at time T L .

[0061] The growth function may be predetermined based on the expected shape characteristic of the tip. In one embodiment, a shape characteristic may be determined for each tip to be measured.

[0062] For example, the shape characteristic can be determined or estimated by evaluating the amplitude of the leading edge 7 at two or more sample times when it is above or equal to the predetermined threshold 4. The rate of change of the amplitude between the two sample times can be used to estimate the rate of linear increase of the amplitude of the leading edge 7. This is the growth function.

[0063] The method iteratively applies the growth function to the estimated amplitude of the peak at a given time T n to obtain an estimated amplitude of the peak at time T n-1 Thus, the amplitude of the leading edge 7 at time T L-2estimated by applying the growth function to the estimated amplitude of the leading edge at time T L-1 is applied.

[0064] A method embodying the present invention continues to extrapolate the leading edge 7 backward until the estimated amplitude is less than or substantially equal to the baseline mean 9 (or zero); or when the amplitude of the input data stream 1 at time T is equal to or greater than the predetermined threshold 4 (for example, because the data stream includes the trailing edge 8 of another peak 2 immediately before the leading edge 7 to be extrapolated).

[0065] Fig. Figure 3 illustrates a segment 15 of the estimated amplitude for the leading edge 7 of the data peak 2 when below the predetermined threshold, estimated according to a method embodying the present invention. Note that the segment 15 has been extrapolated backward to the mean 9 of the baseline because the input data stream 2 remained below the threshold 4 during the period of the extrapolated segment 15.

[0066] Although the growth function in the example described above is linear, the method can assume any growth function. For example, it can be determined or estimated that the leading edge 7 increases essentially exponentially, in which case an exponential growth function can be adopted to extrapolate the leading edge 7. Trailing edge extrapolation

[0067] Extrapolating the trailing edge 8 of the data peak 2 is similar to extrapolating the leading edge 7 in that the amplitude (either measured or estimated) in one time interval is used to estimate the amplitude in a neighboring time interval.

[0068] A method embodying the present invention for extrapolating the trailing edge 8 of a data peak 2 comprises identifying the trailing edge crossover point 14 of a data peak 2. The trailing edge crossover point 14 is identified when the amplitude of the data peak 2 decreases from above the predetermined threshold 4 to below the predetermined threshold 4. The time at which the trailing edge crossover point 14 occurs is referred to herein as T T referred to (see Fig. 2).

[0069] A method embodying the present invention estimates the amplitude of the data peak 2 in a subsequent time interval, T T+1, by applying a decay function to the amplitude of data peak 2 at time T T The decay function is based on the shape characteristics of a peak 2.

[0070] To estimate the amplitude of the peak at a time T n the method iteratively applies the decay function to the estimated amplitude of peak 2 in a previous time interval T n-1 to.

[0071] A method embodying the present invention continues to extrapolate the trailing edge 8 forward until the estimated amplitude at a time T n is substantially equal to the mean 9 of the baseline (or zero); or the measured amplitude of the input data stream at a time T n is greater than or equal to the predetermined threshold 4 (for example, because the data stream 1 includes the leading edge 7 of another peak 2 immediately after the trailing edge 8 of the peak to be extrapolated).

[0072] Preferably, the decay function used to extrapolate the trailing edge 8 is a constant. The step of applying the decay function comprises multiplying the amplitude of the peak at a time T n-1 with a constant decay value between 0 and 1 to determine the amplitude at time T n to appreciate.

[0073] Fig. Figure 3 illustrates a segment 16 of the estimated amplitude for the trailing edge 8 of data peak 2 when below the predetermined threshold, estimated according to a method embodying the present invention. Note that segment 15 has been extrapolated backward to the mean 9 of the baseline because the input data stream 1 remained below threshold 4 during the period of the extrapolated segment 16.

[0074] Although the decay function in the example described above is exponential, the method can adopt any decay function. For example, it can be determined or estimated that the trailing edge 8 decreases essentially linearly, in which case a linear decay function can be adopted to extrapolate the trailing edge 8. Compensated data stream

[0075] The Fig. 3 illustrated compensated data stream 30 comprises the filtered data stream 20 of Fig. 2, segment 15 of the estimated amplitude of the leading edge 7 and segment 16 of the estimated amplitude of the trailing edge 8.

[0076] The compensated data stream 30 of Fig. 3 correlates more closely with the input data stream of Fig. 1, but without the noise 3 and any spontaneous detector events 12. The baseline mean 9 has also been subtracted. As a result, centroid detection and calculation of the area of ​​peak 2 of the compensated data stream 30 are more precise than the conventional thresholding method in isolation. Multiple ion events

[0077] A method embodying the present invention is particularly advantageous when used to process an input data stream comprising at least partially overlapping ion events. That is, where the trailing edge 8 of one tip 2 at least partially overlaps the leading edge 7 of another tip.

[0078] Fig. Figure 4 illustrates a data stream 100 and shows two peaks 2A, 2B. The first data peak 2A is identical in intensity and shape to that shown in Fig. 1 shown data peak 2.

[0079] In isolation, the second peak 2B is also identical in intensity and shape to that of the first peak 2A. However, since the leading edge 7 of the second peak 2B overlaps with the trailing edge 8 of the first peak 2A, the respective amplitudes of both peaks 2A, 2B are compounded during the overlap in the input data stream 100. As a result, more of the second data peak 2B is above the predetermined threshold 4 than the first data peak 2A. As described above, this leads to inaccuracies when using a conventional thresholding method.

[0080] Fig. Figure 5 schematically illustrates how data from the first and second peaks 2A, 2B are compounded together to obtain the Fig. 4 to generate the input data stream 100. As in Fig. 1, a segment 5 of the leading edge 7 of the data peak 2A is below the predetermined threshold. However, the trailing edge 8 of the first peak 2A does not fall below the predetermined threshold 4 because it compounds with the leading edge 7 of the second data peak 2B and is thus "pushed" above the threshold 4.

[0081] Point 101 denotes the point of the leading edge 7 of the second peak 2B, which would have been the leading edge crossing point if the second peak 2B had been separated from the first peak 2A. However, since a portion of the leading edge 7 of the second peak 2B is compounded with the segment 6 of the trailing edge 8 of the first peak 2A, the amplitude of the input data stream 100 is caused to rise above the predetermined threshold 4. The beginning of the leading edge 7 of the second peak 2B coincides with the point at which the amplitude of the trailing edge 8 of the first peak 2A is equal to the predetermined threshold 4. As a result, the entire segment 5 of the second data peak 2B is above the predetermined threshold 4. The hatched area below point 101 in Fig. 5 denotes the amount of the second data peak 2B that was made to exceed the predetermined threshold by being compounded with the trailing edge 8 of the first peak 2A.

[0082] Applying a predetermined threshold to the data stream from Fig. 4 and Fig. 5, without the extrapolation of the present invention, would mean that segment 5 of the leading edge 7 of the first data peak and segment 6 of the trailing edge 8 of the second data peak 2B would be removed from the filtered data stream, but the corresponding segment 5 of the leading edge 7 of the second data peak and the corresponding segment 6 of the trailing edge 8 of the first data peak 2A would still be included. This causes inaccuracies.

[0083] Point 102 denotes the point of the trailing edge 8 of the second peak 2B, which would have been the trailing edge crossing point if the second peak 2B had been separated from the first peak 2A. However, since a portion of the trailing edge 8 of the second peak 2B is compounded with the trailing edge 8 of the first peak 2A, the amplitude of the input data stream 100 is caused to rise above the predetermined threshold 4.

[0084] Accordingly, the area of ​​the data recorded with respect to the second peak 2B is larger than that for the first peak 2A. This is despite the fact that, if the first 2A and the second 2B peaks were separated in time, the area of ​​the part 10 of the peaks 2A, 2B above the predetermined threshold 4 would be identical.

[0085] If data peaks 2A and 2B overlap, this results in a nonlinear relationship between the actual ion current and the reported ion current. For time-of-flight (TOF) instruments, this would also result in m / z shifts depending on the amount of data stream above the threshold.

[0086] By adopting a method embodying the present invention, the leading edge 7 of the first peak 2A and the trailing edge 8 of the second peak 2B are extrapolated. The trailing edge 8 of the first peak 2A and the leading edge 7 of the second peak 2B are not extrapolated because they are compounded to be above the predetermined threshold.

[0087] Extrapolation is only performed when the input data stream 100 is below the threshold.

[0088] Fig. Figure 6 illustrates the resulting compounded data stream 130 compiled by processing the input data stream 100 of Fig. 4 using a method embodying the present invention.

[0089] Fig. Figure 7 illustrates an input data stream 200 comprising three data peaks 2A, 2B and 2C. The size and shape of the first and second peaks in isolation are shown in Fig. 4 and Fig. 5. The size and shape of the third tip 2C, in isolation, is also identical to that of the first 2A and second 2B tips.

[0090] Note that the trailing edge 8 of the second peak 2B falls below the predetermined threshold 4, and that the beginning of the leading edge 7 of the third peak 3B is also below the predetermined threshold 4. At point 15, the leading edge 7 of the third peak 3C compounds with the end of the trailing edge 8 of the second peak to cause the input data stream 200 to rise toward the predetermined threshold 4.

[0091] The amplitude of the data stream 200 between the trailing edge crossing point 14 of the second peak 2B and the leading edge crossing point 13 of the third peak 2C is set to zero (or ignored or flagged) in the filtered data stream (not shown). Consequently, the method embodying the present invention then extrapolates the second peak 2B forward from the trailing edge crossing point 14B. According to one embodiment, the trailing edge extrapolation continues iteratively until the amplitude of the input data stream 200 is greater than or equal to the predetermined threshold 4 or the estimated amplitude at that time is substantially equal to the baseline mean 9.

[0092] Fig. Figure 8 shows an enlarged portion of the data stream 200 of Fig. 7 between the trailing edge crossing point 14B of the second tip 2B and the trailing edge crossing point 13C of the third tip 2C.

[0093] When the data stream 200 crosses the trailing edge crossing point 14B of the second peak 2B, a method embodying the present invention then extrapolates the peak 2B using a decay function, the result of which is correlated with the decreasing portion of the dashed line 210 in Fig. 8. According to a method embodying the present invention, the extrapolation continues until a point 211 at which the amplitude of the input data stream 200 exceeds the predetermined threshold 4.

[0094] With reference to Fig. 8, however, it should be noted that the extrapolated data stream 210 does not accurately estimate the leading edge 7 of the third peak 2C.

[0095] According to an embodiment of the present invention, the leading edge 7 of the third tip 2C is also extrapolated backwards.

[0096] With reference to Fig. 8, the dashed line 212 denotes the estimated amplitude by extrapolating the leading edge 7C of the third peak 13C.

[0097] Between T TB and time T LC The highest estimated amplitude is used at all times. Consequently, the estimated amplitude of the Fig. 8 illustrates the portion of the input data stream 200 more closely to the “V-shaped” nature of the input data stream 200.

[0098] Accordingly, at any time T, if conflicting amplitude estimates exist based on the extrapolation of multiple peaks in the input data stream, the highest estimated amplitude at that time T is used.

[0099] Fig. Figure 9 illustrates another input data stream comprising multiple data peaks, some of which are separated in time and some of which overlap.

[0100] Line 250 denotes the input data stream. Line 251 denotes the threshold. Line 252 is the sum of the input data stream using conventional thresholds. Line 253 denotes the sum of the input data stream when compensated using a method embodying the present invention. Line 254 denotes the percentage difference (error) between the sum of the input data stream using conventional thresholds and the 'ideal' sum of the input data stream. The threshold is 25, and the decay function is 0.5.

[0101] Note that the sum 252 of the input data stream using conventional thresholds is necessarily lower than the sum 253 of the compensated data stream generated according to a method embodying the present invention. This is because the compensated data stream includes data from extrapolating the portions of the input data stream that are below the threshold; whereas the traditional thresholding technique ignores the portion of the input data stream below the threshold.

[0102] As a result, the error between applying a conventional thresholding technique to the data stream and a method embodying the present invention is greatest when the input data stream is below the predetermined threshold.

[0103] In embodiments of the present invention, both a filtered data stream and a data stream comprising the extrapolated peak data are first generated and then compiled to produce a compensated data stream. Subsequently, the compensated data stream is summed over time.

[0104] Alternatively, instead of first compiling a discrete data stream based on a particular operation and then summing the values ​​over time, the input data stream may instead be summed substantially in real time. Accordingly, in one embodiment, the present invention includes summing the input data stream when above the predetermined threshold and summing the extrapolated peak when the input data stream is below the predetermined threshold.

[0105] As used in this specification and these claims, the terms "comprises" and "comprising," and variations thereof, mean that the specified features, steps, or integers are included. These terms should not be construed to exclude the presence of additional features, steps, or components.

[0106] The features disclosed in the above description or the following claims or the accompanying drawings, expressed in their specific forms or as a means of performing the disclosed function or a method or process for achieving the disclosed result, as applicable, may be used singly or in any combination of such features to realize the invention in various forms thereof.

Claims

[1] A method for processing an input data stream including at least one data peak, comprising: Detecting at least one peak in the input data stream having a crest with an amplitude above a predetermined threshold, wherein the input data stream comprises a signal output from an ion detector; extrapolating the segment of the peak having an amplitude above the predetermined threshold based on a shape characteristic of the peak to estimate the amplitude of the segments of the peak having a lower amplitude than said threshold; Generating a filtered data stream comprising applying the predetermined threshold to the input data stream such that the filtered data stream contains only the segment of the peak above the predetermined threshold; and Compile a compensated data stream including the filtered data and extrapolated peak data. [2] A method of processing a data stream according to claim 1, wherein the step of detecting at least one peak comprises detecting at least one of the leading edge, the vertex, or the trailing edge of the peak. [3] A method for processing a data stream according to claim 1, wherein when the amplitude of the input data stream at a time T n less than said threshold, the amplitude in the filtered data stream at time T n is set either to zero or a constant. [4] A method of processing a data stream according to any one of the preceding claims, further comprising: summing the estimated amplitude of extrapolated peak data and the filtered data stream over time. [5] A method according to any preceding claim, further comprising estimating a shape characteristic of the tip. [6] The method of claim 5, wherein the step of estimating a shape characteristic of the peak includes evaluating the segment of the peak above the predetermined threshold. [7] A method according to any preceding claim, wherein the step of extrapolating the peak comprises: Detecting the time, T T at which the amplitude of the trailing edge of the peak in the input data stream falls below the predetermined threshold; and Estimate the amplitude of the peak at time T T by applying a decay function, based on the shape characteristic, to the amplitude of the data peak at time T T-1 , above the predetermined threshold. [8] A method according to claim 7, comprising: Estimate the amplitude of the peak at time T nby applying a decay function based on the shape characteristics of the peak to the estimated amplitude of the peak at time T n-1 . [9] A method according to claim 8, comprising iteratively applying the decay function for all values ​​of n. [10] The method of claim 9, comprising repeating the step of iteratively applying a decay function until the amplitude of the input data stream at time T n is greater than or equal to the predetermined threshold, or the estimated amplitude at time T n essentially zero, the baseline mean, or a constant. [11] A method according to any one of claims 7 to 10, wherein applying the decay function comprises multiplying the amplitude of the peak at time T n-1 with a constant decay value between 0 and 1. [12] A method according to any preceding claim, wherein the step of extrapolating the peak comprises: Detecting the time point, T L at which the amplitude of the leading edge of the peak in the input data stream rises above the predetermined threshold; and Applying a growth function based on the shape characteristic to the amplitude of the data peak at time T L to obtain an estimated amplitude of the peak at time T L-1 to generate. [13] A method according to claim 12, comprising: Estimate the amplitude of the peak at time T n-1 by applying a growth function based on the shape characteristics of the peak to the estimated amplitude of the peak at time T n . [14] The method of claim 13, comprising iteratively applying the growth function for all values ​​of n. [15] The method of claim 14, comprising repeating the step of iteratively applying a growth function until the amplitude of the input data stream at time T n-1 is greater than or equal to the predetermined threshold; or the estimated amplitude at time T n-1 less than or substantially equal to zero, the baseline mean, or a constant. [16] A method according to any one of claims 12 to 15, wherein the growth function is linear. [17] Method according to claim 16, wherein the amplitude at time T n-1 is estimated by subtracting a predetermined constant from the estimated amplitude at time T n is subtracted. [18] A method according to any one of claims 12 to 17, wherein the growth function is based on the rate of change of at least a portion of the segment of the input data stream that is above the predetermined threshold. [19] A method according to any preceding claim, wherein, if the input data stream includes two or more peaks and extrapolation of the respective peaks produces conflicting estimates for the amplitude at a time T, the highest of these estimates is selected for time T. [20] A method according to any preceding claim, wherein the input data stream comprises a voltage signal, an ion signal, an ion current voltage pulse or an electron current pulse. [21] A method according to any preceding claim, wherein the input data stream comprises the output of an analog-to-digital converter of an ion detector of a mass spectrometer. [22] The method of claim 21, further comprising applying a smoothing function to the input data stream using a finite impulse response or infinite impulse response filter. [23] Mass spectrometer comprising: an ion source; an analyzer; an ion detector that generates a data stream; and a processor configured or programmed to: Detecting at least one peak in the data stream having a crest with an amplitude above a predetermined threshold, wherein the input data stream comprises a signal output from an ion detector; extrapolating the segment of the peak having an amplitude above the predetermined threshold based on a shape characteristic of the peak to estimate the amplitude of the segments of the peak having a lower amplitude than said threshold; Generating a filtered data stream comprising applying the predetermined threshold to the input data stream such that the filtered data stream contains only the segment of the peak above the predetermined threshold; and Compile a compensated data stream including the filtered data and extrapolated peak data. [24] A computer program or programmable logic element comprising computer readable program code means or programmed logic for causing a processor to execute a procedure for implementing the method of any one of claims 1 to 22. [25] A computer program element according to claim 24, embodied on a computer-readable medium. [26] A computer-readable medium having a program stored thereon, wherein the program causes a computer to execute a procedure for implementing the method according to any one of claims 1 to 22.

Images