JP5519841B1 - Peak detection method on 3D spectral data - Google Patents

Abstract

【Task】
Providing a peak detection method that can automatically determine the true peak, except for the input of parameters by the user, in the peak detection method on the three-dimensional spectral data, and the true peak using only the three-dimensional spectral data Provide a detection method.
[Solution]
A peak detection method for automatically detecting a peak related to a photometric value from three-dimensional spectrum data having desired measurement data on the third axis, a step of setting peak detection conditions (S20), and three-dimensional spectrum data A step of determining peak candidates for photometric values from above (S30), a step of determining a true peak based on a threshold for the peak candidates (S40), and a step of displaying three-dimensional spectrum data and a true peak (S50).
[Selection] Figure 1

Description

  The present invention relates to a peak detection method, and more particularly to an improvement of a peak detection method for detecting a peak from three-dimensional spectrum data obtained by spectroscopic measurement.

There are several types of spectrophotometers, such as an infrared spectrophotometer and a spectrofluorophotometer. The user appropriately selects a device that meets his needs from among these devices and performs spectroscopic measurement.
Here, in the conventional spectroscopic measurement, two-dimensional spectrum data is created using the wavelength of light monochromatized by the spectroscope and the photometric value obtained by the detector, and the two-dimensional spectrum is measured. It was used for characteristic analysis.
On the other hand, in the field of spectroscopic measurement in recent years, creation of three-dimensional spectrum data in which measurement data desired by the user is added to the third axis to the above two-dimensional spectrum data, and the three-dimensional spectrum data is used. Analysis is being performed.

For example, in spectrofluorescence measurement, after a series of measurements, three-dimensional spectrum data is created with the photometric value as the X axis, the wavelength of the fluorescent light as the Y axis, and the wavelength of the excitation light as the Z axis.
Based on the three-dimensional spectrum data, the user can proceed with various analyzes related to the spectrum peak, such as the behavior of the peak of the measurement target with respect to the wavelength of the excitation light and the qualitative analysis of the measurement target.

However, when determining a peak derived from an unknown measurement target, or when a user who has little experience in spectroscopic measurement determines a peak, it is necessary to spend considerable time and effort even by narrowing down the peak candidates. there were.
Therefore, at present, in the field of spectroscopic measurement, a peak detection method on three-dimensional spectral data using a computer is being developed. For example, Patent Literature 1 and Patent Literature 2 are cited as related technologies.

Patent Document 1 discloses a method for detecting a peak on three-dimensional spectral data relating to spectrofluorescence measurement.
In that method, first, a peak that seems to exist at a common position is found between the two-dimensional spectrum composed of the fluorescence wavelength and the photometric value and the two-dimensional spectrum composed of the excitation light wavelength and the photometric value, and this is used as a peak candidate. .

Subsequently, the excitation light wavelength (or a wavelength having a value multiplied by n or 1 / n with respect to the excitation light wavelength) is subtracted from the fluorescence wavelength of each peak candidate, and the value falls within the reliability threshold range. If it is, the peak candidate is set as a true peak.
However, in the true peak determination method described above, the Raman scattered light also satisfies the condition, so that a peak candidate that should be excluded is regarded as a true peak.
Therefore, the user has to manually remove this peak to be excluded from the true peak.

On the other hand, Patent Document 2 discloses a method for detecting a peak that exists in common on a three-dimensional chromatogram obtained from a plurality of samples.
In the method, first, in addition to time and photometric value, a desired wavelength or mass (m / z) parameter is measured for each sample.
Thereafter, a two-dimensional chromatogram using time and photometric values as variables for each sample and a three-dimensional chromatogram obtained by adding parameters such as wavelength to the two-dimensional chromatogram are created.

Subsequently, in each of the created chromatograms, all positions where the peak top change amount has a maximum or minimum value are determined as peak candidates. Thereafter, a peak having a correlation is determined on a three-dimensional chromatogram derived from a plurality of samples using multivariate analysis.
The method described in Patent Document 2 is an effective method for determining a common peak having a correlation among a plurality of different types of measurement objects. However, a true peak unique to one measurement object is obtained. It cannot be narrowed down automatically.

As described above, in the above-described conventional technology, it is possible to determine the peak candidate derived from the measurement target, but the user needs to be involved in narrowing down the true peak from the peak candidate. Met. Moreover, user experience was required to narrow down the true peak.
Therefore, in the field of spectroscopic measurement, it is desired to develop a peak detection method that can automatically perform all the steps other than parameter input.
In the above conventional technique, it is necessary to create two-dimensional spectrum data and determine peak candidates on the two-dimensional spectrum data. In this field, however, the steps related to the two-dimensional spectrum data are omitted. Development of a technique capable of directly detecting peaks from three-dimensional spectrum data is also desired.

JP 2003-65956 A JP 2011-153966 A

  The present invention has been made in view of the above problems, and its purpose is to provide a peak detection method that can automatically determine a true peak except for input of parameters by a user, and only three-dimensional spectral data. It is to provide a true peak detection method using

As a result of the intensive studies by the present inventors on the above-described conventional problems, the present invention takes data related to spectral photometric values on the first axis, and takes data related to spectral wavelength data on the second axis, A peak detection method for automatically detecting a peak relating to the photometric value from the three-dimensional spectrum data having the desired measurement data on the third axis,
A step of setting peak detection conditions;
Determining peak candidates for the photometric value from the three-dimensional spectrum data according to the peak detection conditions;
Determining a true peak based on the threshold value for the peak candidate after setting a threshold value for the photometric value of the peak candidate;
The list of true peaks and the three-dimensional spectrum data are displayed in separate areas on the display unit, and the position of the true peak in the peak list is indicated by a predetermined mark in the display area of the three-dimensional spectrum data. And a step of displaying.

In the step of setting the peak detection condition, as the peak determination condition, either a peak shape having a convexity in the positive direction of the first axis or a convex shape in the negative direction of the first axis Either or both are selected.
In the embodiment of the present invention, a peak having a convex shape in the positive direction of the first axis is referred to as a “mountain”, and a shape of the spectral peak is convex in the negative direction of the first axis. In the following description, “valley” is used as an example having a peak, but the peak shape may be expressed using expressions other than “mountain” and “valley”.

The step of determining the peak candidate includes
Using the three-dimensional spectral data to set a peak detection map;
Selecting one measurement point from the peak detection map and determining the measurement point as the current position;
Comparing photometric values between the current position and all measurement points adjacent thereto;
A step of marking an evaluated mark for the measurement point at which the photometric value is compared;
In the photometric value comparison step, when the measurement point having the maximum or minimum photometric value is an adjacent measurement point, the step of updating the adjacent measurement point to a new current position;
Update the current position by repeating the process from the photometric value comparison process to the current position update process until the photometric value at the current position becomes maximum or minimum compared to the photometric value at the adjacent measurement point. Repeating the process,
After the step of repeating the update of the current position is completed, determining the current position information as a peak candidate and listing it;
Determine whether all of the measurement points on the map have been labeled, and if there are measurement points on the map that are not yet labeled, after setting the unlabeled measurement points at the current position, It is preferable that the method further includes a step of searching for an unlabeled measurement point, which is executed again from the comparison of the photometric values to the listing of the peak candidates.

The step of updating the adjacent measurement point to a new current position,
When the adjacent measurement point has the maximum or minimum photometric value and the measurement point is already labeled, it is preferable to move to the step of forcibly searching for an unlabeled measurement point.
Further, in the step of determining the peak candidate, selection of the maximum or minimum photometric value is a step of setting the peak detection condition,
When a peak having a convex shape in the positive direction is selected for the peak shape, a measurement point having a maximum photometric value compared to the surrounding area is a candidate for the peak,
When a peak having a convex shape in the negative direction is selected, a measurement point having a minimum photometric value compared to the surrounding area is selected as a candidate for a peak.

Determining the true peak comprises:
Setting a threshold value for the photometric value of each peak candidate;
Determining whether each peak candidate is a true peak according to the threshold;
It is preferable that the method includes a step of adding a true peak determined by the determination to a true peak list.

In the step of determining the peak,
Within the threshold, it is determined whether there is no other peak candidate connected to the peak candidate to be determined,
In the case where there are other peak candidates that are within the threshold value of the determination target peak candidate, a photometric value is compared between the determination target peak candidate and the other peak candidate,
Based on the comparison of the photometric values, if the determination target peak is recognized as a true peak, or if there is no peak candidate connected to the determination target peak candidate, the determination target peak candidate is determined as a true peak. It is preferable to move to the step of adding to the list.

The step of displaying the true peak includes
Displaying a true peak list in an independent region on the display unit;
Displaying a three-dimensional spectrum in an independent region on the display unit;
Displaying the true peak position with a predetermined symbol on the display area of the three-dimensional spectrum.

Furthermore, the process of displaying the true peak is
On the display unit, displaying the position information of the cursor position in the display region of the three-dimensional spectrum and the photometric value in an independent region;
On the display unit, displaying a cross-sectional view of two three-dimensional spectra orthogonal to the cursor position as an intersection point in an independent region;
It is preferable to include a step of adding position information and a photometric value at the cursor position to the true peak list when a position storage command is made by the user at the cursor position. is there.

  The measurement data on the third axis of the three-dimensional spectrum data set in the step of setting the peak detection condition includes the excitation light wavelength, the measurement temperature, the measurement time, the light incident angle, the light detection angle, and It is preferable that one is selected from the concentration conditions of the sample during the reaction.

According to the peak detection method of the present invention, since the measurement data of the third axis, the selection of the peak shape, and the setting of the threshold value are all automatically performed, the user who has not performed the spectroscopic measurement has experienced. Even true peaks can be determined automatically.
Further, since the peak detection method of the present invention determines the true peak directly from the three-dimensional spectrum data, each step related to the two-dimensional spectrum data is not required at the time of peak detection.

It is a figure which shows the basic composition of this invention. It is a flowchart which shows the detail of the peak detection method of this invention. It is a figure which shows the outline | summary of the determination operation | movement of the peak determination process of this invention. It is a figure which shows the example of a setting on the monitor by the peak display process of this invention. It is a flowchart which shows the procedure of the peak detection method of this invention. It is a figure which shows the example of a display of the three-dimensional spectrum data of this invention, and peak information. It is a figure which shows the example of a display of the three-dimensional spectrum data of this invention.

Hereinafter, the present invention will be described using embodiments.
<Embodiment>
FIG. 1 is a diagram showing the configuration of an embodiment of the present invention, and FIG. 1 (A) is a diagram showing the positional relationship of each component of the spectrofluorometer 10 that implements the peak detection method S10, and FIG. ) Is a flowchart showing each component of the peak detection method S10.

The spectrofluorometer 10 shown in FIG. 1A includes a measurement unit 20, a data processing unit 30, and a display unit 40.
Here, the measurement unit 20 includes a light source 21, a spectroscope 22, a measurement object 23, a spectroscope 24, a detector 25, a calculation device 26, and an adjustment device 27.
Here, various lamps suitable for the purpose can be used as the light source 21.
The spectroscope 22 is used when the light from the light source 21 is split, and the spectroscope 24 is used when the fluorescence from the measurement target is split.

A detector suitable for the purpose can be used as the detector 25, and the detection target of the detector 25 is fluorescence from the measurement target 23.
The computing device 26 is for converting the light intensity detected by the detector 25 into a desired photometric value.
The adjustment device 27 is a member for adjusting a measurement environment such as temperature adjustment in the measurement unit 20 and concentration adjustment of a measurement sample.
The operations of the various members 21 to 27 included in the measurement unit 20 are controlled by the data processing unit 30.

The data processor 30 includes a CPU 31 that implements the peak detection method S10 of the present invention, a memory 32 that executes the peak detection method S10 implemented in the CPU 31, and a storage means 33 that stores measurement conditions, measurement data, and the like. I have.
On the other hand, the data processing unit 30 is connected to external input means 34 such as a mouse or a keyboard, and the user can input measurement conditions, peak detection conditions, and the like via the external input means 34.
The display unit 40 includes a monitor 41 that displays various calculation processing results such as three-dimensional spectrum data output from the data processing unit 30 and a true peak list.

Subsequently, the peak detection method (S10) shown in FIG. 1B includes a step of setting peak detection conditions (S20), a step of determining peak candidates (S30), and a step of determining true peaks (S40). And a step (S50) of displaying a true peak.
Here, in the step (S20) of setting the peak detection condition, the peak detection condition on the three-dimensional spectrum data is set by the user. Also, in this step (S20), the user uses predetermined measurement data bearing the third axis of the three-dimensional spectrum data as the excitation light wavelength, measurement temperature, measurement time, light incident angle, light detection angle, and the like. It can be selected from the concentration conditions of the sample during the reaction.

In the step of determining peak candidates (S30), peak candidates relating to photometric values are determined from the three-dimensional spectrum data in accordance with the peak detection conditions set by the user.
In the step of determining the true peak (S40), after setting a threshold value for the photometric value of the peak candidate, the true peak is determined based on the threshold value.
In the step of displaying the true peak (S50), the list of true peaks and the three-dimensional spectrum data are displayed in separate areas on the monitor 41, and the true peaks are displayed on the display area of the three-dimensional spectrum. The position of is displayed.

The embodiment of the present invention is configured as described above, and the detector 25 detects data related to the photometric value, and the spectroscope 24 acquires data related to the wavelength of the fluorescence.
The spectroscope 22 and the adjusting device 27 acquire measurement data such as the wavelength of the excitation light, the temperature at the time of measurement, the measurement time, the incident angle of light, the detection angle of light, and the concentration condition of the sample during the reaction. Various data are stored in the storage means 33.
As described above, the user can set the measurement data obtained by the spectroscope 22 and the adjustment device 27 as the third axis.

In addition, the peak detection method (S10) of the present invention requires user involvement when setting peak detection conditions and setting thresholds, but all other steps can be performed automatically. Furthermore, the peak detection method (S10) can also detect the true peak directly from the three-dimensional spectrum data.
Since the true peak detected by using the peak detection method (S10) is displayed on the three-dimensional spectrum data, the user can easily confirm the position of the true peak and the shape of the true peak. You can also.

Details of each step included in the peak detection method (S10) of the present invention and the operation thereof will be described with reference to FIG.
FIG. 2 is a flowchart showing details of each step included in the peak detection method (S10), FIG. 2 (A) is a step of setting peak detection conditions (S20), and FIG. 2 (B) is a step of determining peak candidates. (S30) and FIG. 2 (C) show a step (S40) for determining a true peak, and FIG. 2 (D) shows a step (S50) for displaying a true peak, respectively.

<Step of setting peak detection conditions (S20)>
The step (S20) of setting the peak condition shown in FIG. 2A includes the step of setting the third axis (S21), the step of selecting the peak shape (S22), and the step of setting the range of peak detection ( S23).
In the step of setting the third axis (S21), the measurement data of the third axis is the wavelength of the excitation light dispersed by the spectroscope 22, or the temperature, measurement time, and light during measurement monitored by the adjustment device 27. The incident angle, the light detection angle, the concentration condition of the sample, and the like are appropriately selected.
Since the spectroscopic measurement conditions and various measurement data are stored in the above-mentioned storage means 33, different three-dimensional spectrum data can be constructed by the user reselecting the necessary measurement data from the storage means. It can be made.

In the step of selecting a peak shape (S22), a peak target is selected by the user.
Here, the target of the peak is a peak (a figure convex in the positive direction with respect to the photometric value axis) or a peak (a figure convex in the negative direction with respect to the photometric value axis). Either one or both are selected.

In the step of setting the peak detection range (S23), the user sets the cut-out range (maximum value and minimum value) of the three-dimensional spectrum data at the time of peak detection.
In the present invention, it is preferable to set a cutout range related to the photometric value, but the present invention is not limited to this, and the cutout range can be set for the fluorescence wavelength.
As described above, in the step of setting peak detection conditions (S20), each condition for peak detection is set.
In addition, the contents of the conditioning are not considered even if they are inexperienced in spectroscopic measurement.

<Step of determining peak candidates (S30)>
The step (S30) of determining peak candidates shown in FIG. 2B compares the photometric value with the step (S31) of setting a map for peak detection, the step of selecting the current position from the measurement point (S32). A step of performing (S33), a step of marking an evaluated mark (S34), a step of updating an adjacent measurement point as a new current position (S35), a step of repeatedly updating the current position (S36), A step of listing peak candidates (S37) and a step of searching for unlabeled measurement points (S38) are provided.

In the step of setting a map for peak detection (S31), the data related to the photometric value is set as the first axis, the data related to the wavelength of the fluorescent light is set as the second axis, and selected in the above step (S21). Three-dimensional spectrum data is created with the measurement data as the third axis.
At this time, using the second axis and the third axis of the three-dimensional spectrum data, a map 示 す as shown in the following (Equation 1) is created, and in the present invention, this map Χ is used for peak detection.
In the subsequent steps (S32) to (S38), peak candidates are determined based on this map Χ. Each measurement point is preferably labeled with “0” or the like in advance.

In the step of selecting the current position from the measurement points (S32), the peak position is determined by selecting the current position from one measurement point Χ _nm in the map Χ. Here, as an example, χ ₀₀ in the map Χ is set to be the initial current position.
Thereby, in the step of comparing the photometric values (S33), the photometric values are compared between the current position χ ₀₀ and all the measurement points χ ₀₁ , χ ₁₁ , χ ₁₀ adjacent thereto.
In the step of labeling the evaluated mark (S34), “1” is labeled as the evaluated mark for the measurement points χ ₀₀ , χ ₀₁ , χ ₁₁ , χ ₁₀ where the photometric values are compared. The

In the step of updating the adjacent measurement point to the new current position (S35), in the step of comparing the photometric values (S33), the measurement point having the maximum or minimum photometric value is the adjacent measurement point χ ₀₁ , If it is either χ ₁₁ or χ ₁₀ , the adjacent measurement point is updated to a new current position.

The “maximum or minimum” conditioning is determined by the step (S20).
For example, when a peak is selected as the peak shape in the step (S20), the measurement point having the largest photometric value among the adjacent measurement points is set as the new current position, and on the other hand, the peak is determined in the step (S20). When the valley is selected as the shape, the measurement point having the smallest photometric value among the adjacent measurement points is updated as a new current position.
That is, when a mountain is selected in the step (S20), and when the photometric value of the measurement point in the above example satisfies the relationship of χ ₀₀ <χ ₀₁ <χ ₁₁ <χ ₁₀ , the adjacent measurement point χ ₁₀ Becomes the new current position.

In step of repeating the updating of the current position (S36), the photometric values of the current position chi _xy is compared to photometric value of the measurement point adjacent to its current location chi _xy, until the maximum or minimum, compares the photometric value The process from the step (S33) to the step (S35) for updating the current position is repeated.
That is, when a mountain is selected in the step (S20) and the photometric value of the current position _{ｘ xy satisfies} χ _{x + 1, y + 1} <... <Χ _{x-1, y-1} <χ _{x, y} Then, the process proceeds to the next step (S37).
In the step of listing peak candidates (S37), it is assumed that the position of the current position χ _xy when the above loop processing is completed is the peak candidate, and the position (measurement point) information and photometric value of the current position χ _xy. Add information to the list.

In the subsequent step of searching for unlabeled measurement points (S38), it is determined whether or not all of the measurement points existing in the map _nm are labeled with “1” and are not yet labeled ( That is, when a measurement point (having “0”) exists in the map _nm , the peak point is listed from the photometric value comparison step (S33) with the measurement point as the current position (S37). Repeat until.
As a result, “1” is labeled at all the measurement points in the map _nm , and all current positions having the maximum or the minimum as compared with the adjacent measurement points can be listed as peak candidates.

In the step of updating the current position (S35), if “1” is already labeled at the measurement point to be updated, the process proceeds to the step of forcibly searching for an unlabeled measurement point (S38). It is particularly preferable to set so as to shift.
That is, by this operation, it is possible to prevent the current position from being updated to the measurement point once compared with the photometric value. Thereby, the time concerning the process (S30) which determines a peak candidate can be reduced.

Further, when both the peaks and valleys have been selected as the peak shape in the step (S20), the step (S30), until first all measurement points in the map chi _nm is labeled "1", Mountain It can be set to search for a peak candidate having the shape of Thereafter, in a state where all measurement points in the map _nm are returned to “0”, a peak candidate having a valley shape is now searched.
In the step (S30) of the present invention, the order is not limited to that particular order. First, after determining the peak candidate having the valley shape, the peak candidate having the mountain shape is determined again. It is also possible to change the setting.

As described above, in the step (S30) of determining the peak candidates, the peak candidates are determined according to the conditions input in the step (S20).
Thereby, in the step (S30), only measurement points having the maximum and / or minimum photometric value in a certain area can be automatically picked up.
Further, in the step (S30), it is not necessary to use an advanced arithmetic processing such as an optimization method, and it is not necessary to consider a problem such as a local minimum.

<Step of determining peak (S40)>
The step (S40) of determining a peak shown in FIG. 2C includes a step of setting a threshold (S41), a step of determining a peak (S42), and a step of adding a true peak to a list (S43). It has.
Here, in the step of setting a threshold value (S41), a threshold value for determining whether or not the peak is a true peak is set with reference to the photometric value of each peak candidate.
Next, in the step of determining a peak (S42), it is determined whether or not each peak candidate is a true peak according to the threshold value.
In the step of adding a true peak to the list (S43), the true peak determined by the above determination is added to the list of true peaks.

In the step of determining the peak (S42), it is first determined whether or not there is another peak candidate connected to the peak candidate to be determined within the threshold value based on the photometric value of each peak candidate. Is done.
Here, if there are other continuous peak candidates within the threshold value of the peak candidate to be determined, the photometric values are compared between the peak candidate to be determined and the other peak candidate.
Then, if the determination target peak is recognized as a true peak based on the comparison of the photometric values, or if there is no peak candidate connected to the determination target peak candidate, the determination target peak candidate is determined to be true. It moves to the process (S43) which determines with the peak of this and adds it to said true peak list | wrist.
This determination operation will be described in more detail with reference to FIG.

FIG. 3 is a diagram showing an outline of the peak determination step (S42) according to the present invention. FIGS. 3A and 3B show other consecutive peak candidates within the threshold range of the target peak candidate. 3C and 3D show examples in which other peak candidates are present within the threshold value of the target peak candidate.
3 (A) to 3 (D) are those when a peak is selected as the peak shape in the step (S20) of setting the peak detection condition, but the same is true even when a valley is selected. Is determined. In the following description, it is assumed that the peak candidate to be determined is α and the other peak candidates are β.

There is no other continuous peak candidate β within the threshold range set with reference to the peak top of the peak candidate α ₁ shown in FIG. In this case, based on the above determination, the peak candidate alpha ₁ is a true peak, is added to the list.

There are other peak candidates β ₂ around the peak candidate α ₂ shown in FIG. However, within the scope of the set threshold the peak top of a peak candidate alpha ₂ as a reference, another of mountain apart. That is, these peak candidates do not have continuity. Again, peak candidate alpha ₂ is a true peak, is added to the list.

There are other peak candidates β ₃ around the peak candidate α ₃ shown in FIG. Also within the set threshold peak top of a peak candidate alpha ₃ as a reference, another mountain is continuous. Therefore, the photometric values are compared between the peak candidate α ₃ and the peak candidate β ₃ . As a result, since the peak candidate α ₃ has a photometric value larger than that of the peak candidate β ₃ , the peak candidate α ₃ is added to the list as a true peak.

On the other hand, the peak candidate α ₄ shown in FIG. 3D is continuous with other peak candidates β ₄ . Therefore, although a comparison of the photometric value between the two peak candidate is performed, in this example, since towards the other peak candidate beta ₄ is larger photometric value of the peak top in comparison with the peak candidate alpha _4, peak candidate alpha ₄ is not added to the peak list.

As described above, in the step of determining a peak (S40), only a peak having high reliability as a peak can be extracted from the peak candidates.
Further, in the step (S40), since everything except the threshold setting is automatically performed, it is possible to reduce the burden on the user.

<Step of displaying true peak (S50)>
The step (S50) of displaying a true peak shown in FIG. 2D includes a step of displaying a peak list (S51), a step of displaying a three-dimensional spectrum (S52), and a step of displaying a true peak position. (S53), a step of displaying cursor position information (S54), a step of displaying a cross-sectional view of the three-dimensional spectrum data at the cursor position (S55), and a step of adding cursor position information to the peak list (S56) ). FIG. 4 shows a setting example of a screen displayed on the monitor 41 by executing this process.

In the step of displaying the peak list (S51), the true peak list is displayed in the region PL (upper left of FIG. 4), and in the step of displaying the three-dimensional spectrum (S52), the three-dimensional spectrum data is stored in the region SD (in FIG. 4). Display in the center).
In the step of displaying the true peak position (S53), the true peak position in the peak list is displayed by a symbol such as a plus mark in the three-dimensional spectrum display area SD.
The user can confirm the true peak information by looking at the region PL and the region SD.

  Further, in the step of displaying the cursor position (S54), the display region CP (FIG. 5) in which the coordinates of the cursor position and the photometric value in the display region SD of the three-dimensional spectrum are provided independently of the region PL and the region SD. 4 in the upper right). The user can always check the photometric value information at the cursor position by looking at this area CP.

In the step of displaying the cross-sectional view of the three-dimensional spectrum data at the cursor position (S55), the cross-sectional views of two orthogonal three-dimensional spectra with the cursor position in the region SD as an intersection are represented as independent regions CC (FIG. 4). It is preferable to display in the lower center). As a result, the user can visually confirm the information on the cursor position.
In the step of adding to the peak list (S56), the position of the current cursor position in the display area SD and the photometric value information can be added to the true peak list when a position storage command is issued by the user. .

In the step of displaying the three-dimensional spectrum of the present invention (S52), the three-dimensional spectrum display method can be selected from a three-dimensional solid diagram, a contour map, and a color map.
As described above, in the step of displaying the true peak (S50), the position of the true peak determined in the above step can be confirmed on the three-dimensional spectrum.
In addition, even inexperienced persons in spectroscopic measurement can use the region CP and the region CC to confirm the state around the true peak in detail.

The above is the detailed description of the step (S20) of setting the peak condition provided in the method (S10) for detecting the peak of the present invention to the step of displaying the true peak (S50). Hereinafter, the specific operation of the peak detection method (S10) of the present invention will be described using examples.
<Example>
Measurement conditions Equipment used: spectrofluorometer 10
Measurement sample: Rhodamine B
First measurement data: Photometric value of fluorescent light Second measurement data: wavelength of fluorescent light Third measurement data: wavelength of excitation light, measurement time, ambient temperature, incident angle,
Detection angle, sample concentration conditions, etc./Peak detection conditions Data on the third axis: Select the wavelength of the excitation light Peak shape: Mountain Peak detection range: Maximum value = 10000, Minimum value = 0
Threshold: 100
・ Results Number of detected true peaks: 3

In this example, after setting the above measurement conditions, the spectrofluorometer 10 was used to execute the spectroscopic measurement and peak detection method (S10). In the present embodiment, measurement is performed by a general spectrofluorimetric method, and the description thereof is omitted.
Hereinafter, details of peak detection using the peak detection method (S10) of the present invention for each measurement data stored in the storage means 33 by the spectral fluorescence measurement will be described with reference to the procedure shown in FIG.

The flowchart shown in FIG. 5 shows the procedure of the peak detection method (S10) of the present invention, and starts from selection of measurement data.
Here, “wavelength of excitation light” is selected as the third measurement data (S21), “mountain” is selected as the peak shape (S22), and “maximum value = 10000, minimum value = 0” as the peak detection range. Is input (S23).
Thereby, the step (S20) of setting the peak detection condition is completed, and the peak detection method (S10) shifts to the step (S30) of determining the peak candidate, and information such as necessary measurement data is stored from the storage means 33. Is called.

In the step (S30), first, the photometric value of the fluorescent light is used as the first axis, the wavelength of the fluorescent light is used as the second axis, and the wavelength of the excitation light is used as the third axis. And are set as position coordinates, respectively.
By developing each measurement point on the coordinates, a map _nm indicating the coordinates of the three-dimensional spectrum data is created (S31).
Each measurement point on the coordinates is labeled with “0” indicating that the photometric value is not compared.

The initial current position χ ₀₀ is designated from the map _{ｎｍ nm} . After that, the photometric values are compared between the χ ₀₀ and all the measurement points adjacent thereto (S33), and each evaluated measurement point is labeled “1”. (S34).
After labeling, refers to the result of the comparison of the photometric value, a photometric value of _xy chi adjacent measurement point, when larger than the photometric value of the current position chi _00, a new measurement points chi _xy adjacent current The position is updated (S35).
Thereafter, the updating of the current position and the comparison of the photometric values are repeatedly performed until the photometric value of the new current position χ _xy becomes a position larger than the photometric value of the measurement point adjacent to the current position χ _xy (S36). ), The current position where the work is completed is listed as a peak candidate (S37).

Then, after listing one peak candidate, it is determined whether all the measurement points on the map _nm have the label “1”. If there is still a measurement point having the label “0”, The measurement point _{ｘ x′y ′} having the _unlabeled is the current position, and the process returns to the step (S33). The steps (S32) to (S37) are repeated until “1” is labeled on all the measurement points on the map _nm .
As described above, when “1” is already labeled at the measurement point scheduled to be updated, the process is forcibly shifted to the step of searching for an unlabeled measurement point (S38). Duplicate counting of peak candidates can be prevented.
A list of peak candidates could be obtained by this step (S30).

The peak determination step (S40) starts with a threshold value (100 in this embodiment) being input by the user (S41), and then the peak top of each peak candidate is used as a reference from the peak candidate list. The determination of the true peak is started according to the threshold range (S42).
If the current determination target is not a true peak, the determination is repeated with a different peak candidate as a new target (S42).
Then, the peak candidate recognized as a true peak is added to the target peak candidate in the true peak list (S43).
As a result of performing the above process for all the peak candidates in the list, three true peaks were obtained.

Then, in the step (S50), the list of three true peaks created in the step (S40) for determining the true peak is the step (S30) for determining candidates for the peak in the region PL of FIG. The generated three-dimensional spectrum data displays the position of three true peaks on the area SD in FIG. 6 and on the area SD (S51) to (S53).
Although the three-dimensional spectrum data shown in the region SD of FIG. 6 is shown as a color diagram, the data is shown as the three-dimensional solid diagram of FIG. 7A or the contour map of FIG. It can be shown inside.

In addition, the area CP of FIG. 6 displays cursor position information and photometric value information on the area SD (S54). The area CC in FIG. 6 displays a two-dimensional spectrum cross section (the cross sections A and B in the area SD) shown in the area SD (S55).
Furthermore, when an instruction to add a peak list is given, the position of the cursor position and photometric value information can be added to the true peak list (S56).
Then, peaks 1 to 3 obtained by the current peak detection are displayed on the monitor screen shown in FIG. Here, each position of the peaks 1 to 3 is displayed as a plus mark on the area SD.
Further, the shapes of the peaks 1 to 3 can be determined based on the spectrum data of the region SD and the region CC, and in this case, these three peaks were considered to be appropriate as true peaks.

As described above, by using the peak detection method (S10) of the present invention, the user can determine the true peak from the three-dimensional spectrum data without depending on experience, and the peak detection result is displayed on the display screen. Can be confirmed.
In the above embodiment, the wavelength of the excitation light is treated as measurement data on the third axis. However, the temperature during measurement, the time during measurement, the incident angle of light, the detection angle of light, and the concentration of the sample during reaction From the conditions, desired parameters can be called from the storage means 33 to create three-dimensional spectrum data.
And the peak detection method (S10) of this invention can detect a peak directly from the three-dimensional spectrum data.

  In the field of spectroscopic measurement using a spectroscopic measurement device, the present invention automatically performs everything except selection of measurement data for the third axis, selection of peak shape, and setting of a threshold value. Even a person can determine a true peak automatically and easily. Further, the present invention does not require a step of creating two-dimensional spectrum data and a step of determining peak candidates on the two-dimensional spectrum data at the time of peak detection.

DESCRIPTION OF SYMBOLS 10 Spectrofluorometer 20 Measurement part 30 Data processing part 40 Display part S10 Peak detection method S20 The process of setting peak detection conditions S30 The process of determining a peak candidate S31 The process of setting the map for peak detection S32 The measurement point of the present position Step S33 for selecting a photometric value S34 Step for marking an evaluated mark S35 Step for updating an adjacent measurement point to the current position S36 Step for repeatedly updating the current position S37 Step for listing peak candidates S38 Not yet Step S40 for searching for a measurement point of a label Step S41 for determining a true peak Step S41 for setting a threshold value Step S42 for determining a peak Step S43 for adding a true peak to the list S50 Step for displaying a true peak S51 Step for true peak Step S52 for displaying a list Step S5 for displaying three-dimensional spectrum data 3 Step S54 for Displaying True Peak Position Step S54 for Displaying Cursor Position Information Step S55 for Displaying Section View at Cursor Position Step for Adding Cursor Position Information to List

Claims (8)

Peaks relating to the photometric values are taken from the three-dimensional spectral data in which the data relating to the photometric value of the spectrum is taken on the first axis, the data relating to the wavelength of the spectrum is taken on the second axis and the desired measurement data is taken on the third axis. A peak detection method for automatically detecting
A step of setting peak detection conditions;
Determining peak candidates related to the photometric value from the three-dimensional spectrum data according to the peak detection condition;
After setting a threshold for the photometric value of the peak candidate, determining a true peak based on the threshold for the peak candidate;
The list of true peaks and the three-dimensional spectrum data are displayed in separate areas on the display unit, and the true peaks in the peak list are displayed with predetermined marks in the display area of the three-dimensional spectrum data. A process,
The step of determining the peak candidate includes
Setting a peak detection map using the three-dimensional spectral data;
Select map or found one measuring point for the peak detection, and determining the measuring point to the current position,
Comparing photometric values between the current position and all measurement points adjacent thereto;
A step of marking an evaluated mark for the measurement point at which the photometric value is compared;
In the photometric value comparison step, when the measurement point having the maximum or minimum photometric value is an adjacent measurement point, the step of updating the measurement point to a new current position;
Update the current position by repeating the process from the photometric value comparison process to the current position update process until the photometric value at the current position becomes maximum or minimum compared to the photometric value at the adjacent measurement point. Repeating the process,
After the step of repeating the update of the current position is completed, determining the current position information as a peak candidate and listing it;
It is determined whether all the measurement points on the map are labeled, and when there are measurement points that are not yet labeled on the map, the measurement points that are not labeled are set at the current position, and then again. And a step of searching for unlabeled measurement points for executing from the comparison of the photometric values to the listing of the peak candidates, and a method for detecting a peak on three-dimensional spectrum data. The peak detection method according to claim 1,
In the step of setting the peak detection condition, as the peak determination condition, either a peak shape having a convexity in the positive direction of the first axis or a convex shape in the negative direction of the first axis A method for detecting a peak on three-dimensional spectrum data, characterized in that either or both are selected. The peak detection method according to claim 1 or 2,
The step of updating the current position includes:
If the adjacent measurement point has the maximum or minimum photometric value and the measurement point is already labeled, the process is forcibly shifted to a step of searching for an unlabeled measurement point. Peak detection method on original spectral data. In the peak detection method according to claims 1 to 3,
In the peak candidate determination step, selection of the maximum or minimum photometric value is the peak condition setting step,
When a peak having a convex shape in the positive direction is selected for the peak shape, a measurement point having a maximum photometric value compared to the surrounding area is a candidate for the peak,
A three-dimensional spectrum characterized in that, when a peak having a convex shape in the negative direction is selected as the peak shape, a measurement point having a minimum photometric value compared to the surrounding area is a target of a peak candidate. The peak detection method on the data. In the peak detection method according to claims 1 to 4,
Determining the peak comprises:
Setting a threshold for the photometric value of each peak candidate;
Determining whether each peak candidate is a true peak according to the threshold;
Adding a true peak determined by the determination to a true peak list, and detecting a peak on three-dimensional spectrum data. The peak detection method according to claim 1,
The step of displaying the true peak is as follows.
Displaying a true peak list in an independent region on the display unit;
Displaying a three-dimensional spectrum in an independent region on the display unit;
And a step of displaying the position of the true peak with a predetermined symbol on the display region of the three-dimensional spectrum. The peak detection method according to claim 6,
The process of displaying the true peak is
On the display unit, displaying a measurement point and a photometric value corresponding to a cursor position in the display region of the three-dimensional spectrum in an independent region;
On the display unit, displaying a cross-sectional view of two three-dimensional spectra orthogonal to the cursor position as an intersection point in an independent region;
Adding a position information and photometric value information at the cursor position to the true peak list when a position storage command is made by a user at the cursor position. To detect peaks on 3D spectral data. In the peak detection method according to claim 1,
The measurement data bearing the third axis of the three-dimensional spectrum data set in the step of setting the peak detection condition includes the wavelength of the excitation light, the temperature at the time of measurement, the measurement time, the incident angle of the light, the detection angle of the light, and the reaction A method for detecting a peak on three-dimensional spectral data, wherein one of the concentration conditions of a sample in the sample is selected .

Images