U.S. patent application number 12/155812 was filed with the patent office on 2008-10-09 for method and apparatus for chromatographic data processing.
Invention is credited to Kisaburo Deguchi, Masahito Ito.
Application Number | 20080249717 12/155812 |
Document ID | / |
Family ID | 34704459 |
Filed Date | 2008-10-09 |
United States Patent
Application |
20080249717 |
Kind Code |
A1 |
Ito; Masahito ; et
al. |
October 9, 2008 |
Method and apparatus for chromatographic data processing
Abstract
A chromatographic analyzer is provided for facilitating curve
fitting by means of the linear least-square method for a
chromatogram that contains a plurality of overlapping peaks. The
present invention is characterized by a chromatographic data
processor for executing data processing of a chromatogram obtained
by separating a sample to be measured using a column and detecting
the separated sample, wherein fitting processing is executed to
each peak in an arbitrary time region having the plurality of peaks
of the chromatogram starting from the front side of the time region
or from the back side of the time region, and the processed peaks
are subtracted from the chromatogram in the time region so that the
plurality of peaks in the chromatogram can be separated from one
another. Thus, the plurality of overlapping peaks, particularly
three or more overlapping peaks in the chromatogram can be easily
separated from one another only by defining some setting
conditions.
Inventors: |
Ito; Masahito; (Hitachinaka,
JP) ; Deguchi; Kisaburo; (Hitachinaka, JP) |
Correspondence
Address: |
DICKSTEIN SHAPIRO LLP
1825 EYE STREET NW
Washington
DC
20006-5403
US
|
Family ID: |
34704459 |
Appl. No.: |
12/155812 |
Filed: |
June 10, 2008 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
11654577 |
Jan 18, 2007 |
7403859 |
|
|
12155812 |
|
|
|
|
Current U.S.
Class: |
702/32 |
Current CPC
Class: |
G01N 30/8624 20130101;
G01N 30/8641 20130101; G01N 30/8637 20130101; G01N 30/8631
20130101 |
Class at
Publication: |
702/32 |
International
Class: |
G01N 30/86 20060101
G01N030/86 |
Claims
1-12. (canceled)
13. A liquid chromatographic analyzer comprising: a sample
separating unit which separates a sample to be measured into
components, the sample separating unit comprising: a fluid pump
which feeds an eluting solution at a constant rate; an eluting
solution storing unit which stores the eluting solution; a sample
storing unit which stores the sample; and a separation column to
which the sample and the eluting solution are fed to perform an
elution which separates the sample into the components; a detecting
unit which detects the separated sample, the detecting unit
comprising a photometer, the photometer being capable of detecting
each of the components as an output value at each elapsing time; a
data processing unit which executes data processing by obtaining
detected result from the detecting unit, the data processing unit
comprising: a unit which specifies an arbitrary retention time
region having a plurality of peaks of a chromatogram obtained as
the detected result of the detecting unit; a unit which specifies
whether fitting processing is executed from a front side end or a
back side end of the specified retention time region; and a unit
which specifies a weighting function and a waveform function used
in the fitting processing; and a display unit which displays
results obtained from the data processing unit, wherein a new
retention time region, which has peaks other than those peaks at
which the fitting processing is completed to define their peak
shapes, is displayed on the display unit, and the specifying of the
weighting function is performed with conceptual graphics of
candidate weighting functions superposed over the chromatogram
displayed on the display unit.
Description
TECHNICAL FIELD
[0001] The present invention relates to a chromatography technology
such as a liquid chromatography technology, and particularly to a
data processing method.
BACKGROUND ART
[0002] In chromatographs such as a liquid chromatographic analyzer,
a gas chromatographic analyzer or the like, a sample to be measured
is let to pass through a column to be separated into components,
and each of the separated components is detected as an output value
at each elapsing time using a photometer such as a chromatographic
detector.
[0003] Signals output from the chromatographic detector are
recorded as time sequential data having a time interval of several
hundreds ms. This is what is called a chromatogram having signal
intensity in the ordinate and retention time in abscissa. In
general, the signal intensity is converted to a digital value
I.sub.j every an arbitrary time interval (time t.sub.j) to execute
data processing.
[0004] FIG. 3 shows an example of a chromatogram obtained by
executing a body fluid amino acid analysis.
[0005] As shown in FIG. 3, peaks of 11 components from Gly
(glycine) to Tyr (tyrosine) densely exist in the range of retention
time from 23 to 34 (min). In such a case, area-quantitative
calculation is conventionally executed using a vertically dividing
method in which a vertical line is drawn from each minimum point
between peaks, that is, what is called "a root". However, this
method produces an error as large as several tens % to cause an
incorrect result when the peaks are strongly overlapped with each
other. Therefore, when the chromatogram of such a kind needs to be
quantitatively analyzed in a high accuracy, it has been general
that the analyzing time is lengthened to improve the separation
degree.
[0006] On the other hand, in order to perform quantitative
calculation without lengthening the analyzing time even if peaks
are overlapped so strongly with each other, quantitative
calculation is tried to be performed using numerical analysis in a
manner like data processing. This method is, for example, a
non-linear least-square method.
[0007] In the case of using the non-linear least-square method, at
least three independent parameters (A: area, T.sub.R: retention
time, .sigma.: standard deviation) are used as variables in order
to execute fitting of a peak for one component. Therefore, in order
to execute fitting of peaks for a plurality of components, it is
necessary to calculate three parameters of A.sub.i, T.sub.Ri,
.sigma..sub.i for each of the components (i).
[0008] The conventional examples of using the non-linear
least-square method are disclosed in Japanese Patent Application
Laid-Open No.6-324029 and Japanese Patent Application Laid-Open
No.63-151851.
[0009] These examples disclose that overlapping peaks on a
chromatogram are curve-fit using a waveform function such as the
Gaussian function or an EMG function (exponentially modified
Gaussian function) which can express an asymmetric waveform of a
peak. As shown in these examples, the overlapping peaks can be
separated into individual peak waveforms, and the quantitative
calculation can be performed by obtaining peak sizes such as a peak
area and so on corresponding to a component of each of the
peaks.
DISCLOSURE OF INVENTION
[0010] However, in the conventional examples using the non-linear
least-square method, the separation of the peak waveforms is
applied to two or three overlapping peaks, but not applied to the
more number of overlapping peaks.
[0011] The reason is that in the case of separating the overlapping
peak waveforms using curve fitting through the non-linear
least-squire method, as the number of peak components is increased
to 3, 4, 5, . . . , there occurs a phenomenon that the fitting
processing is difficult to be converged or that the separation of
peaks can not correctly performed (the error is increased).
[0012] For example, in the case of the chromatogram shown in FIG.
3, when fitting is tried to the 11 components from Gly to Tyr at a
time, the 33 parameters of 11.times.3 must be determined at a time.
This is very difficult calculation processing to the non-linear
least-square method, and accordingly various kinds of techniques
are necessary in order to solve this problem. Therefore, when the
curve-fitting is executed using the non-linear least-square method
in the case of existing many overlapping peaks on a chromatogram, a
measuring operator must specify calculation regions (time windows)
for 2 or 3 peaks seeming to be converged one by one. This process
expenses much time and much effort, and in addition, there is a
problem in the reliability of the calculation result because the
regions are artificially determined.
[0013] An object of the present invention is to provide a
chromatographic analyzer capable of easily executing curve fitting
using the non-linear least-square method to a chromatogram having a
plurality of overlapping peaks.
[0014] The present invention to attain the above object is
characterized by a chromatographic data processor for executing
data processing of a chromatogram obtained by separating a sample
to be measured using a column and detecting the separated sample,
wherein fitting processing is executed to each peak in an arbitrary
time region having the plurality of peaks of the chromatogram
starting from the front side of the time region or from the back
side of the time region, and the processed peaks are subtracted
from the time region of the chromatogram so that the plurality of
peaks in the chromatogram can be separated from one another.
[0015] The object, the operation and the effect of the present
invention will be described in detail in the section of DESCRIPTION
OF THE PREFERRED EMBODIMENTS to be described later.
BRIEF DESCRIPTION OF DRAWINGS
[0016] FIG. 1 is a flowchart showing fitting processing in
accordance with the present invention.
[0017] FIG. 2 is a diagram showing the outline of a chromatographic
analyzer.
[0018] FIG. 3 is a chart showing a chromatogram for 70 minutes of a
body fluid amino acid analysis method.
[0019] FIG. 4 is a chart showing an example of a display of a
chromatogram in which a time window is set.
[0020] FIG. 5 is a view showing an example of a display of a
dialogue box for setting a waveform function.
[0021] FIG. 6 is diagrams showing examples of weighting
functions.
[0022] FIG. 7 is a view showing an example of a display of a
dialogue box for selecting a weighting function and a direction of
fitting.
[0023] FIG. 8 is a view showing a dialogue box for instructing
execution of the fitting processing.
[0024] FIG. 9 is an illustrating chart showing a chromatogram in
which a first fitting region is set.
[0025] FIG. 10 is an illustrating chart showing the chromatogram in
which a first component is cut out.
[0026] FIG. 11 is a view showing an example of a display of a
fitting result.
[0027] FIG. 12 is a chart showing an example of graphically setting
the weighting function.
BEST MODE IN WHICH THE INVENTION IS CARRIED OUT
[0028] An embodiment of the present invention will be described
below.
[0029] FIG. 2 is a diagram showing the outline of a liquid
chromatographic analyzer to which the present invention is applied.
An eluting solution 1 is initially pumped to a column 3 using a
fluid pump 2 by an instruction from a controller 5. A sample supply
portion 8 is arranged between the fluid pump 2 and the column 3,
and a sample is supplied from a sampler 7 containing the sample to
the eluting solution by an instruction of the controller 5. The
sample is separated by the column 3 to be detected using a detector
4 such as a UV detector. A chromatogram of the detected data is
transmitted to the controller 5 to be data-analyzed, and the result
is displayed on a display 6 or printed by a printer 9.
[0030] Data processing in the controller 5 will be described below
when the chromatogram of FIG. 3 is obtained as the detected
result.
[0031] In the present embodiment, the data processing of the
chromatogram is executed mainly according to the following
procedure.
[0032] Step 1: specification of a time interval to execute fitting
thereto.
[0033] Step 2: selection of a weighting pattern.
[0034] Step 3: selection of a fitting direction.
[0035] Step 4: clicking of a fitting execution button.
[0036] Step 5: displaying and outputting of the result.
[0037] FIG. 1 is a flowchart showing the detailed process of the
data processing of the chromatogram described above. In the present
embodiment, description will be made on a case where the fitting
processing is executed two components by two components.
[0038] Initially, a chromatogram to be executed fitting is selected
from detected results obtained to start the processing (100). Here,
the variable i is set to 1 (one), that is i=1.
[0039] Next, an interval of retention time to execute fitting
processing thereto (a time window) is set (101). Here, the peak
group of 11 components from Gly to Tyr in the chromatogram of FIG.
4 is set to the time window.
[0040] The setting of time window is performed by dragging the time
abscissa of the chromatogram displayed on the display 9 using a
cursor. Otherwise, a peak starting and a peak ending the fitting
may be selected by picking the peaks Gly and Tyr of the
chromatogram displayed. Further, it is possible to employ a method
of inputting a starting time and an ending time.
[0041] Next, a waveform function used for the fitting calculation
is set. As the waveform function, for example, "Gaussian" or "EMG"
may be selected. This selection can be performed by selecting a
waveform function used using a dialogue box on the display 9 shown
in FIG. 5.
[0042] Although the fitting in the present embodiment is executed
two components by two components, it is necessary to make the
influence of signals of a second and later waveforms in order to
accurately determine a waveform of a peak in the front side of the
fitting direction, that is, a first peak. Therefore, a weighting
function shown in FIG. 6 (a) is set (103).
[0043] The weighting function is a function for weighting each of
the two peaks. The example of FIG. 6 (a) is a function that the
weight w is set to 1 (one), that is, w=1 from a peak start time 21
to a peak end (root) time 23 of the first peak, and then the weight
w is linearly slanted to be reduced from 1 to 0 from a peak start
time 23 to a peak end (root) time 25 of the second peak. As the
weighting function, various kinds of weighting functions may be
employed. Some of the examples are described below.
[0044] FIG. 6 (b): a weighting function having a curvilinear
gradient. FIG. 6 (c): a weighting function having an S-shaped
curvilinear gradient. FIG. 6 (d): a weighting function having a
linear gradient from a peak start point of a first peak to a peak
end point of a second peak. FIG. 6 (e): a weighting function having
a weight of 0.5 at an end point of a second peak. FIG. 6 (f): a
weighting function having a linear gradient from a maximum point of
a second peak to an end point of the second peak. FIG. 6 (g): a
weighting function having a linear gradient from a maximum point of
a first peak to an end point of a second peak by particularly
attaching importance to the left-hand side of the first peak. The
weight of 0.5 is set to an end point of the first peak (a start
point of the second peak. FIG. 6 (h): a weight function having a
flat weight of 0.5 from an end point of a first peak (a start point
of a second peak) to an end point of the second peak. FIG. 6 (i): a
weighting function which is equivalent to a case without any
weight.
[0045] In the concrete, the weighting function is set by displaying
a dialogue box shown in FIG. 7 on the display 9 and selecting a
weighting pattern. Therein, by selecting an "ARBITRARY SET" and
pushing an "OPTION" button, the above-described various kinds of
pattern function type graphs shown in FIG. 6 can be set. The
various kinds of patterns of FIG. 6 are pre-stored in a memory such
as a hard disk in the controller 5, and can be easily read out from
the memory to be used for the processing by being specified using a
dialogue box of FIG. 7.
[0046] Next, a direction of executing fitting calculation is
determined (104).
[0047] That is, describing the example of FIG. 3, it is decided
whether the processing is executed from the front side (the Gly
side) or the back side (the Tyr side). In the concrete, the setting
of the direction executing the fitting calculation is set using the
dialogue box shown in FIG. 7.
[0048] An example of executing the processing from the Gly side
will be described below.
[0049] By the above, setting of the conditions necessary for the
fitting processing has been completed.
[0050] Then, a dialogue box shown in FIG. 8 is displayed after
selecting an "OK" button in the dialogue box of FIG. 7, and the
fitting calculation processing is started by pushing an "EXECUTION"
button.
[0051] As the calculation processing is started, number of peaks
within the set time window is detected to be set to Imax (106).
[0052] In the concrete, inflection points inside the set time
window are detected to obtain maximum points and minimum points,
and an interval for each of the peaks is defined by determining the
maximum point as the apex and the minimum points as the end point
and the starting point of the peak. Therein, the roots (the minimum
points) of the peak and a retention time having the points are
stored in relation to each other.
[0053] Next, a difference between Imax and i is calculated (107).
Therein, if the difference is larger than 3 (three components), the
processing proceeds to Process 108.
[0054] When it is judged in Process 107 that the difference is
larger than 3 (three components), a width of the weighting function
set in Process 103 is adjusted so as to meet with a width between
the roots detected in Process 105 (108).
[0055] FIG. 9 is an illustrating chart showing a chromatogram in
which a region of first fitting processing is set. The width of the
weighting function is adjusted so as to meet with the width of the
fitting region as shown in FIG. 9.
[0056] Next, the fitting processing is executed to the two
components of the front side in the fitting direction (109).
[0057] The fitting processing will be described below in
detail.
[0058] As described above, the Gaussian or EMG function is used as
the waveform function f(t) for the fitting. Firstly, in a case of
employing the Gaussian function, the waveform function is shown in
Equation 1.
f ( t ) = g 1 ( t ) + g 2 ( t ) + at + b = A 1 2 .pi. .sigma. 1 exp
{ - ( t - t R 1 ) 2 2 .sigma. 1 2 } + A 2 2 .pi. .sigma. 2 exp { -
( t - t R 2 ) 2 2 .sigma. 2 2 } + at + b ( g i ( t ) : the suffix i
denotes a component i . ) ( Equation 1 ) ##EQU00001##
[0059] There, the term (at+b) denotes a base line. Though the
waveform function becomes Equation 1 shown above because the
fitting processing is executed two components by two components in
the present embodiment, the fitting processing may be executed
three components by three components if the capability of
non-linear least-square method can be secured. Of course, extension
to the EMG function may be possible. In a case of employing the EMG
function, the term g.sub.i(t) of the waveform function in Equation
1 is replaced by Equation 2.
g 1 EMG ( t ) = A i 2 .pi. .sigma. i .tau. i .intg. 0 t exp { - ( t
- t Ri - t ' ) 2 2 .sigma. i 2 - t ' .tau. i } t ' ( Equation 2 )
##EQU00002##
[0060] In the least-square method, each of the fitting parameters
A.sub.i, t.sub.RI, .sigma..sub.i and .tau..sub.i is determined so
as to minimize the following S.sub.1 of Equation 3 or the following
S.sub.2 of Equation 4. There, I.sub.j is a signal intensity of a
measured chromatogram, j is a suffix expressing time t.sub.j, and N
is number of data points in a time interval. Either of S.sub.1 and
S.sub.2 may be used.
S 1 = j = 1 N w j ( f ( t j ) - I j ) 2 ( Equation 3 ) S 2 = j = 1
N w j 2 ( f ( t j ) - I j ) 2 ( Equation 4 ) ##EQU00003##
[0061] The function g.sub.i(t) having a smaller value of t.sub.R
(that is, the earlier retention time) in the f(t) of Equation 1
obtained here corresponds to the first peak Gly, and accordingly
the waveform of the first component is determined (110).
[0062] Next, the waveform of the first component g.sub.i(t) is
subtracted from the original measured chromatogram to form a
chromatogram cut out the first peak Gly (111). FIG. 10 is an
illustrating chart showing the chromatogram in which the first peak
Gly has been cut out. It is preferable not to subtract the base
line at+b from the original chromatogram because it does not
influence the peak waveform.
[0063] Next, the variable i is incremented, and the processing is
returned to Process 107.
[0064] Then, the fitting processing for the peaks Ala and Cit is
executed to determine the second component waveform of the second
peak Ala according to the same procedure as that of the fitting
processing for the first peak.
[0065] After that, the peaks Ala, Cit . . . are successively
subtracted until the peaks for two components Leu (leucine) and Tyr
in the back side end. When the remaining becomes two components
(i=2), the processing proceeds to Process 113 to determine the peak
waveforms of the final two components by executing the peak fitting
for the two components without weighting (114). Then, the peak
waveforms for the two components are subtracted from the original
chromatogram (115) to complete the processing.
[0066] By the series of the processing described above, all the
waveforms in the time window have been separated from one another.
Further, according to the above-described processing, since the
remaining chromatogram after cutting out all the peak waveforms
becomes a base line, the base line can be obtained at a time.
[0067] After completion of the above-described processing,
quantitative calculation (for example, concentration) is executed
using parameters for each of the peaks obtained by the
above-described processing. Then, a display for each of the peaks
(each of the components) shown in FIG. 11 is displayed on the
display 9 as the calculation result.
[0068] There, because the uncertainties (deviations) of the
parameters for each of the components can be obtained by the
above-described processing, the uncertainty of each of the measured
quantitative values can be calculated based on the above-described
uncertainties.
[0069] The point that the uncertainties of the measured
quantitative values can be calculated is also one of the advantages
of the present embodiment. the calculation of obtaining the
uncertainties of the measured quantitative values is executed
according to the error propagation equation.
[0070] Further, the quality of the fitting can be judged from the
magnitude of the uncertainty of the measured quantitative value (it
can be said that the quality is better when the uncertainty is
smaller). For example, the uncertainty can be used for judging
quality of the calculation result of the overlapping portion when
the fitting processing executed from both of the front side end and
the back side end which is to be described later.
[0071] In the present invention, the above-described processing
executing separating the peaks while the peaks are successively
being cutting out is called as "sequencing", and the
chromatographic data processor for executing the sequencing is
called as "a chromatographic peak sequencer".
[0072] Although the fitting calculation described above is executed
from the front side end, the fitting calculation may be executed
from the back side end by setting the fitting direction of the
dialogue box of FIG. 7. In this case, the function g.sub.2(t)
expressing the second peak is subtracted from the chromatogram, and
the shape of the weighting function is reversed left to right. In
the processing direction described above, the peaks are cut out
from the peaks having the larger retention time, that is, in order
of the peaks Tyr, Leu, Ile (isoleucine) in the case of FIG. 4.
[0073] Ideally, the fitting processing may be executed from either
of the front side and the back side. However, actually, it is
preferable that the fitting processing is executed from the both
sides toward a peak having the strongest overlapping intensity.
When such fitting processing is performed, the button "AUTOMATIC
JUDGMENT" in the dialogue box of FIG. 7 is selected.
[0074] For example, in the case of the chromatogram of FIG. 3, the
forward sequencing processing is executed from the peak Gly to the
peak Cysthi (Cystathionine) or Ile and the backward sequencing
processing is executed to the peak Lue or Ile because overlapping
of the peaks Lue and Ile is strong. By doing so, in regard to the
peaks Lue and Ile, which fitting result should be employed and
determined can be judged by comparing the appropriateness of the
fitting results between the forward processing and the backward
processing.
[0075] Therein, the judgment can be executed by using the
uncertainties (the deviations) of the fitting parameters A.sub.i,
t.sub.Ri and .sigma..sub.i themselves as the indexes for judging
the appropriateness or by using the statistical index .chi..sub.2
or number of repetitive calculating times for watching the
convergence rate of .chi..sub.2. Further, more comprehensive
judgment can be performed by combining these parameters. Any way,
the peak area value A.sub.i needs to be accurately calculated in to
execute quantitative value calculation, and accordingly importance
should be placed on the uncertainty of the peak area value
A.sub.i.
Another Embodiment of Setting the Weighting Function
[0076] Although the weighting function set in the above-described
process 102 has been set on the dialogue box, the weighting
function may be set through a program setting method in which a
table method shown in Table 1 is used or through a graphical
setting method in which setting is performed by displaying and
superposing the weighting function on a chromatogram as shown in
FIG. 12.
TABLE-US-00001 TABLE 1 Weighting program (expressing FIG. 6 (g))
Weighting Characteristic point value Shape 0 starting point of
first 1 drawing peak a vertical line horizontal line apex of first
peak 1 gradient (straight line) root between peaks 0.5 gradient
(convex) end point of second peak 0
[0077] The setting program of Table 1 shows the example of setting
the weighting function of FIG. 6 (g). The weighting function is
defined by that a measurer inputs "weighting values" and "shapes".
Table 1 is displayed on the display.
[0078] FIG. 12 shows the method of setting the weighting function
by superposing on the chromatogram on the display. For example, a
cursor is moved to a point near a node 31 using a pointing device,
and a characteristic point of the first peak starting point is
displayed by clicking on the point. By doing so, the node 31 is
determined. When the node 31 needs to be moved, the node is picked
and then released. That is, the node may be dragged. Next, a node
32 is determined by clicking on a point near the node 32 and
vertically dragging the node 31 to a point having a weight of 1.
Therein, a vertical line is automatically drawn from the node 31 to
the node 32 and determined. Similarly, nodes 33 and 34 are
determined to set the weighting function. When the weighting
function needs to be changed, changing of setting can be performed
by clicking an arbitrary node to highlight the node. Further, a
node can be deleted. Furthermore, a new node can be added by
clicking an arbitrary point. When a line section between nodes is
changed to a curvilinear line, the line section to be changed is
clicked to be highlighted (the displaying color is changed), and
the set attribution is changed by light-hand side clicking.
Therein, concave, convex or the like can be specified.
Another Embodiment of Sequencing
[0079] In the embodiment of sequencing described above, the
measurement is once performed to obtain a chromatogram, and after
that the data processing is performed. However, it is possible to
simultaneously execute the sequencing processing with measuring of
the sample. An embodiment in such a case will be described
below.
[0080] The simultaneous sequencing processing with measurement is
progressed as follows.
[0081] Step 1: setting of a time interval (a time window) to
execute fitting thereto.
[0082] Step 2: setting of a weighting pattern.
[0083] Step 3: setting of a fitting direction.
[0084] Step 4: measuring of a sample.
[0085] Step 5: executing of fitting processing.
[0086] Step 6: displaying and outputting of the result.
[0087] In steps 1 to 3, a time program is set before starting
measurement as shown in Table 2. A time window is set by specifying
a time period from the fitting start time to the fitting end time.
In addition, the time window may be also set by selecting a fitting
start peak and a fitting end peak as by specifying peak names "Gly"
to "Tyr".
TABLE-US-00002 TABLE 2 Time Program Time Command Status 0.0
Waveform function Gaussian 0.0 Weighting function Straight line
gradient 0.0 Fitting direction Automatic judgment 23.0 Fitting
start ON 34.0 Fitting end ON
[0088] According to the chromatographic peak sequencer in
accordance with the present invention, fitting of a chromatogram
having a plurality of overlapping peaks can be automatically
executed only by setting some number of conditions.
[0089] Further, a highly accurate base line can be calculate at the
same time. From the viewpoint of existing the highly accurate base
line determining function, the chromatographic peak sequencer in
accordance with the present invention is considered to be an
excellent data processor. However, because the fitting is executed
by assuming the shape of peak waveform as a shape of Gaussian or
EMG function, an attention should be paid on this point when a
shape of an actual peak waveform is significantly different from
the shapes of these analytical functions. In such a case, it is
necessary to pre-store specific waveform functions by obtaining an
isolated peak waveform for each component from a standard
sample.
[0090] According to the present invention, in a chromatogram having
a plurality of overlapping peaks, particularly, having three or
more overlapping peaks, the peaks can be easily separated only by
setting some number of setting conditions. By doing so, the
accuracy of quantitative analysis and qualitative analysis can be
improved. Further, the capability of determining the base line can
be also improved. Furthermore, measurer's labor for the data
processing can be substantially reduced.
* * * * *