JP5840845B2 - Hazard factor determination method, hazard factor determination device, and program - Google Patents

Description

The present invention relates to a hazard factor determination method, a hazard factor determination device, and a program.

  In modern day-to-day life, there are various hazards, especially in Japan, which is becoming a super-aging society, and it is very important to avoid hazards. There is a need to do it easily.

  Here, according to the glossary of terms related to agriculture, forestry and fisheries described in Non-Patent Document 1, the harm factor means a substance in food or a state of food that may cause adverse health effects. For example, hazards include biological factors such as harmful microorganisms, pesticides, additives and chemicals in the food itself that can adversely affect human health (eg food poisoning bacteria, viruses, parasites, etc.) There are chemical factors (eg, pesticides, additives, etc.) or physical factors (eg, foreign matter, radiation, etc.). For example, examples of harmful factors in daily life include pathogenic bacteria (eg, Salmonella, pathogenic E. coli, etc.), spoilage microorganisms, viruses, parasites, pathogenic microorganisms (eg, Vibrio cholerae), biological substances (eg, shellfish poisons, Fugu poison), food additives, foreign substances (for example, hair, insects, etc.) and the like.

  Particularly in the food field, mold poisons such as deoxynivalenol (DON), nivalenol (NIV), and zearalenone (ZEA), which are examples of hazards, are a major problem worldwide.

  In addition, thorough hygiene management is expected at sites such as factories engaged in food processing and food production. In order to prevent food poisoning and the like, it is a great prevention to confirm the number of microorganisms that cause harm and to remove food residues and stains that serve as nutrient sources for microorganisms in order to prevent their growth.

  On the other hand, conventionally, the following techniques have been used as a method for detecting the above-mentioned hazard factors (for example, mold poison, food residue, dirt, etc.).

  For example, among the above-mentioned hazard factors, chemical analysis techniques such as high performance liquid chromatography (HPLC) and mass spectrometry, culture methods, etc. are used for detection of mold toxins (eg, deoxynivalenol, nivalenol, zearalenone, etc.). It was.

  However, the chemical analysis method and the culture method described above have a problem that it is necessary to perform complicated pretreatment. Furthermore, the above-described chemical analytical methods and culture methods require dedicated reagents and culture media, require time until results are obtained, the equipment is expensive, and have specialized skills for the operator of the equipment. Since it is required, there is a problem that it is difficult to make a quick and easy determination in the field. Furthermore, since the amount of mold toxins (for example, deoxynivalenol, nivalenol, zearalenone, etc.) to be detected is very low (for example, concentrations in the order of ppm or ppb), the above-described chemical analysis techniques and culture methods are used. However, there is a problem that these mold poisons cannot be detected without pretreatment such as concentration and growth, and therefore complicated pretreatment cannot be omitted.

  Of the above-mentioned hazard factors, ATP (adenosine triphosphate) wiping tests described in Non-Patent Documents 2 and 3 have been used for detection of food residues and dirt that are nutrient sources for microorganisms.

  Here, ATP wiping inspection is the most popular technique as a sanitary indicator of how much food residue and dirt have been removed. ATP (ie, bacteria, mold, yeast, animals, The energy substance (indispensable for metabolic activity) contained in cells such as plants is detected using the light emission principle of fireflies, and the degree of contamination is quantified. In addition, ATP wiping inspection was listed as a cleanliness inspection method in the “Food Hygiene Inspection Guidelines (Microbiology)” supervised by the Ministry of Health, Labor and Welfare in 2004. It is spreading as a self-management tool to judge the effect.

  However, in the ATP wipe inspection described in Non-Patent Documents 2 and 3, the range that can be inspected is 10 cm square at a time, and it is necessary to wipe the surface to be measured with an inspection kit such as a cotton swab. There is a problem that it takes time and labor to inspect over a part that does not reach or over a wide part.

  Therefore, in recent years, it has been desired to develop a method for analyzing a measurement object quickly and easily without complicated preprocessing.

  For example, in the method for discriminating flour described in Patent Document 1 by the present applicant, the fluorescence intensity is increased while gradually changing the excitation wavelength irradiated to the measurement object and the fluorescence wavelength observed from the measurement object. A fluorescent fingerprint (also known as excitation / fluorescence matrix) acquired by measurement is acquired, and the type and type of flour are determined by analyzing the acquired fluorescent fingerprint. In the component distribution visualization method described in Patent Document 2 by the same applicant, fluorescent fingerprint imaging is performed by measuring a fluorescent fingerprint in units of pixels. In addition, in the component distribution visualization method described in Patent Document 3 by the same applicant, the influence of the solvent content (or the disturbance factor for the specific component to be visualized) between the measurement object and the sample is further removed, and the measurement object The specific component of the object is analyzed more clearly.

  Here, the “fluorescence fingerprint” described in Patent Documents 1 to 3 is also called an excitation-emission matrix (EEM), and the fluorescence spectrum while continuously changing the wavelength of the excitation light applied to the sample. Means three-dimensional data obtained by measuring. Since the shape of the fluorescence fingerprint is determined in a component-specific manner like a fingerprint, the measurer can detect subtle component differences that cannot be determined only by fluorescence measurement at a single single wavelength. In addition, the measurer uses fluorescent fingerprint imaging that measures the fluorescent fingerprint for each pixel or each pixel block with positional information (that is, spatial information indicating the position of each pixel in the image) in addition to the fluorescent fingerprint information. By using, the distribution of the specific component in the measurement object can be visualized.

JP 2010-185719 A Japanese Patent No. 3706914 JP 2010-266380 A

Ministry of Agriculture, Forestry and Fisheries, “Glossary on Agriculture, Forestry and Fisheries” [searched on February 24, 2011], Internet (URL: http://www.maff.go.jp/j/use/tec_term/h.html#h11) Kikkoman Corporation, “ATP Wiping Inspection”, [February 24, 2011 search], Internet (URL: http://www.kikkoman.co.jp/bio/j/kensakiit/bfukititotop.html) Takeshi Ito, supervised by ATP wiping inspection study group, "New Sanitation Management Act ATP wiping inspection (revised version)", published by Chicken Egg Information Center, 2005

  However, in conventional methods for discriminating flour (Patent Document 1, etc.) and component distribution visualization methods (Patent Document 2, Patent Document 3, etc.), measuring and analyzing fluorescent fingerprints eliminates the need for complicated preprocessing. However, there is a problem in that it does not consider the detection of the harmful factor from the fluorescent fingerprint.

The present invention was made in view of the above problems, and by measuring and analyzing a fluorescent fingerprint, a complicated pretreatment is unnecessary, and a hazard factor quantification method capable of quickly and easily quantifying a hazard factor, It is an object to provide a hazard factor determination device and a program.

In order to achieve such an object, the hazard factor quantification method of the present invention is a hazard factor quantification method for quantifying a hazard factor from a measurement object, and when the measurement object is a powder, a pulverizer is used. A homogenization step for homogenizing the measurement object into fine pulverization, while changing the excitation wavelength to be irradiated and the fluorescence wavelength to be observed in a stepwise manner within a predetermined excitation wavelength range and a predetermined fluorescence wavelength range, Fluorescence fingerprint information acquisition step of measuring fluorescence intensity and acquiring fluorescence fingerprint information of the measurement object, and performing multivariate analysis on the fluorescence fingerprint information acquired in the fluorescence fingerprint information acquisition step, based on the results of the multivariate analysis comprises hazard quantification step, of quantifying the hazard from the measurement object, in the fluorescent fingerprint information acquisition step, it is equalized to the finely pulverized by the homogenizing step The acquired fluorescence fingerprint information of the measurement object, further, by removing the noise information other than fluorescent fingerprint of interest from the fluorescent fingerprint information acquired, representing features of the hazards contained in the measurement object Only the fluorescence fingerprint information is extracted, and in the hazard factor determination step, the matrix data obtained from the extracted fluorescence fingerprint information is rearranged to be changed to a continuous one-dimensional vector, and changed to the one-dimensional vector. The amount of the hazard factor is estimated by performing the multivariate analysis on the fluorescent fingerprint information, and the hazard factor is quantified from the measurement object based on the quantified result of the estimated hazard factor. It is characterized by.

Furthermore, hazards quantification method of the present invention, the hazards by performing the hazards quantification method described above, in the hazard quantification step, the PLS regression analysis as the multivariate analysis on the fluorescence fingerprint information And the amount of the hazard is quantified from the measurement object based on the quantitative result of the estimated hazard.

Furthermore, hazards quantification method of the present invention, the hazards quantification method described above, in the hazard quantification step, the principal component analysis is performed as the multivariate analysis on the fluorescence fingerprint information, the main component analysis The hazard factor is quantified from the measurement object based on the distribution of the main component score of the hazard factor obtained by the above.

Furthermore, hazards quantification method of the present invention, obtained in hazards quantification method described above, in the hazard quantification step, performed discriminant analysis as the multivariate analysis on the fluorescence fingerprint information, by the discrimination analysis The harm factor is quantified from the measurement object based on the distribution of the hazard factor discrimination score.

Furthermore, hazards quantification method of the present invention, the hazards quantification method described above, the hazard, it is mycotoxin and / or ATP, and said.

The hazard factor quantification method of the present invention is characterized in that, in the hazard factor quantification method described above, the mold toxin is at least one of deoxynivalenol, nivalenol, and zearalenone.

Furthermore, hazards quantification method of the present invention, the hazards quantification method described above, the measurement object, wheat, corn, food residues, and it is at least one microorganism, characterized by.

Furthermore, hazards quantification method of the present invention, the hazards quantification method described above, when the measurement object is a powder, homogenizing step for uniformizing the milling the measurement object using the grinding apparatus, In the fluorescent fingerprint information acquisition step, the fluorescent fingerprint information of the measurement object that has been homogenized into the fine pulverization in the homogenization step is acquired.

Further, the hazard factor quantification device of the present invention includes a spectral illumination device that irradiates a measurement target with a predetermined excitation wavelength, a spectral detection device that measures the measurement target with a predetermined fluorescence wavelength, and the measurement target is a powder. A hazard factor quantifying device for quantifying a hazard factor from the measurement object, comprising a pulverizing device for homogenizing the measurement object into fine pulverization in the case of a body , the predetermined excitation wavelength range and a predetermined fluorescence Fluorescence fingerprint information acquisition unit for measuring the fluorescence intensity of the measurement object while gradually changing the excitation wavelength to be irradiated and the fluorescence wavelength to be observed in a wavelength range, and acquiring the fluorescence fingerprint information of the measurement object, And a hazard factor quantification unit that performs multivariate analysis on the fluorescence fingerprint information acquired by the fluorescence fingerprint information acquisition unit, and quantifies the hazard factor from the measurement object based on the result of the multivariate analysis, Wherein the fluorescent fingerprint information acquisition unit, the pulverizing device by acquires fluorescence fingerprint information homogenized the measurement object in the milling, further, other than the fluorescent fingerprint of interest from the fluorescent fingerprint information acquired Noise information is extracted, and only the fluorescence fingerprint information representing the characteristics of the hazard included in the measurement object is extracted, and the hazard factor quantification unit obtains matrix data obtained from the extracted fluorescence fingerprint information. Rearranging, changing to a continuous one-dimensional vector, estimating the amount of the hazard by performing the multivariate analysis on the fluorescence fingerprint information changed to the one-dimensional vector, the estimated the The hazard factor is quantified from the measurement object based on a quantitative result of the hazard factor.

Furthermore, hazards quantification device of the present invention, the hazards quantification device described above, the hazard quantifying section, the as the multivariate analysis on the fluorescence fingerprint information of the hazards by performing PLS regression analysis An amount is estimated, and the hazard factor is quantified from the measurement object based on the estimated quantitative result of the hazard factor.

Furthermore, hazards quantification device of the present invention, the hazards quantification device described above, the hazard quantification unit performs principal component analysis as said multivariate analysis on the fluorescence fingerprint information, by the principal component analysis The hazard factor is quantified from the measurement object based on the obtained distribution of the main component scores of the hazard factor.

Furthermore, hazards quantification device of the present invention, the hazards quantification device described above, the hazard quantification unit performs discriminant analysis as the multivariate analysis on the fluorescence fingerprint information, obtained by the discriminant analysis Further, the hazard factor is quantified from the measurement object based on the distribution of the discrimination score of the hazard factor.

Furthermore, hazards quantification device of the present invention, the hazards quantification device described above, the hazard, it is mycotoxin and / or ATP, and said.

The hazard factor quantification device of the present invention is characterized in that, in the hazard factor quantification device described above, the mold poison is at least one of deoxynivalenol, nivalenol, and zearalenone.

The hazard factor quantification apparatus of the present invention is characterized in that, in the hazard factor quantification device described above, the measurement object is at least one of wheat, corn, food residue, and microorganism.

Further, the hazard factor quantification device of the present invention further includes a pulverization device that homogenizes the measurement object into fine pulverization when the measurement object is a powder in the hazard factor quantification device described above. The fingerprint information acquisition unit acquires fluorescent fingerprint information of the measurement object that has been uniformized into the fine pulverization by the pulverizer.

Further, the program of the present invention includes a spectroscopic illumination device that irradiates a measurement object with a predetermined excitation wavelength, a spectroscopic detection device that measures the measurement object with a predetermined fluorescence wavelength, and a case where the measurement object is powder. to the milling device to equalize the measurement object pulverisation, with a, a program to be executed by the hazard quantification device for quantifying the hazard from the measurement object, in the hazard quantification device A homogenization step of homogenizing the measurement object into fine pulverization using the pulverizer , and changing the excitation wavelength to be irradiated and the fluorescence wavelength to be observed in a stepwise manner within a predetermined excitation wavelength range and a predetermined fluorescence wavelength range. While measuring the fluorescence intensity of the measurement object, the fluorescence fingerprint information acquisition step of acquiring the fluorescence fingerprint information of the measurement object, and the fluorescence finger acquired in the fluorescence fingerprint information acquisition step Performs multivariate analysis on the information, based on the result of the multivariate analysis, hazards quantitative quantifying the hazard from the measurement object, is run at the fluorescent fingerprint information acquisition step, wherein Obtain fluorescent fingerprint information of the measurement object that has been homogenized into the fine pulverization in a homogenization step, further remove noise information other than the target fluorescent fingerprint from the acquired fluorescent fingerprint information, Only the fluorescent fingerprint information representing the characteristics of the hazard included in the measurement object is extracted, and the matrix data obtained from the extracted fluorescent fingerprint information is rearranged in the hazard factor quantification step, and the continuous one-dimensional The amount of the harm factor is estimated by performing the multivariate analysis on the fluorescence fingerprint information changed to the one-dimensional vector, and based on the quantified result of the estimated hazard factor. Te, wherein the quantifying the hazard from the measurement object.

The program of the present invention estimates the amount of the hazard factor by performing PLS regression analysis as the multivariate analysis on the fluorescent fingerprint information in the hazard factor quantification step in the program described above. The hazard factor is quantified from the measurement object based on the estimated quantitative result of the hazard factor.

Further, the program of the present invention is the program described above, wherein in the hazard factor determination step, the fluorescent fingerprint information is subjected to principal component analysis as the multivariate analysis, and the harm obtained by the principal component analysis is performed. The hazard factor is quantified from the measurement object based on the distribution of the main component score of the factor.

Further, in the program according to the present invention, the program of the present invention performs discriminant analysis as the multivariate analysis on the fluorescent fingerprint information in the hazard factor quantifying step, and the hazard factor obtained by the discriminant analysis is analyzed. The risk factor is quantified from the measurement object based on the distribution of the discrimination score.

  The program of the present invention is characterized in that, in the above program, the harm factor is mold poison and / or ATP.

  The program of the present invention is characterized in that, in the above program, the mold toxin is at least one of deoxynivalenol, nivalenol, and zearalenone.

  The program according to the present invention is characterized in that, in the program described above, the measurement object is at least one of wheat, corn, food residue, and microorganism.

  Further, the program of the present invention further executes a homogenization step of homogenizing the measurement object into fine pulverization using a pulverizer when the measurement object is a powder in the program described above, In the fluorescent fingerprint information acquisition step, fluorescent fingerprint information of the measurement object that has been homogenized into the fine pulverization in the homogenization step is acquired.

  According to the present invention, the fluorescence intensity of the measurement object is measured while changing the excitation wavelength to be irradiated and the fluorescence wavelength to be observed in a stepwise manner within a predetermined excitation wavelength range and a predetermined fluorescence wavelength range. Fluorescent fingerprint information is acquired, multivariate analysis is performed on the acquired fluorescent fingerprint information, and based on the result of the multivariate analysis, a harmful factor is detected from the measurement object, so the fluorescent fingerprint is measured and analyzed. As a result, it is possible to eliminate the need for complicated preprocessing and to quickly and easily detect the hazard factor. In this way, the present invention uses this fluorescent fingerprint to measure and extract only the fluorescence information of the target harm factor from the enormous data in a non-destructive manner, and performs the determination of the amount of the hazard factor and the presence / absence determination. There is an effect that can be.

  Further, according to the present invention, in the detection of the hazard factor, the amount of the hazard factor is estimated by performing PLS regression analysis as multivariate analysis on the fluorescence fingerprint information, and based on the quantitative result of the estimated hazard factor. Thus, since the hazard factor is detected from the measurement object, the amount of the hazard factor can be quantified quickly and easily from the fluorescent fingerprint.

  Further, according to the present invention, in the detection of the hazard factor, the principal component analysis is performed as multivariate analysis on the fluorescence fingerprint information, and based on the distribution of the principal component score of the hazard factor obtained by the principal component analysis, Since the hazard factor is detected from the measurement object, it is possible to quickly and easily determine the presence or absence of the hazard factor from the fluorescent fingerprint.

  Further, according to the present invention, in the detection of the hazard factor, the discriminant analysis is performed as a multivariate analysis on the fluorescent fingerprint information, and the measurement object is based on the distribution of the hazard factor discrimination score obtained by the discriminant analysis Therefore, it is possible to quickly and easily determine the presence or absence of a hazard factor from a fluorescent fingerprint.

  Further, according to the present invention, since the harm factor is mold poison and / or ATP, ATP that is a nutrient source of microorganisms that cause harmful mold poison and / or harm is detected quickly and easily from the fluorescent fingerprint. There is an effect that can be.

  In addition, according to the present invention, the mold toxin is at least one of deoxynivalenol, nivalenol, and zearalenone. There is an effect that the amount or the total amount can be accurately detected.

  In addition, according to the present invention, since the measurement object is at least one of wheat, corn, food residue, and microorganisms, wheat and corn that are contaminated with harmful mold poisons, and harm are caused. The food residue containing the nutrient source (ATP) of the microorganism and the microorganism itself can be used as an object to be measured, and the harmful factor (for example, mold poison, ATP, etc.) can be detected quickly and easily from the fluorescent fingerprint. .

  Further, according to the present invention, when the measurement object is powder, the measurement object is homogenized into fine pulverization using a pulverizer, and the measurement object is homogenized into fine pulverization in the acquisition of fluorescent fingerprint information. Since the fluorescence fingerprint information is acquired, it is possible to reduce the variation in the fluorescence intensity emitted from the measurement object, thereby improving the accuracy in the detection, quantification, and discrimination of the hazard factor.

FIG. 1 is a diagram showing an example of a fluorescent fingerprint as a contour graph of three-dimensional data. FIG. 2 is a bird's-eye view showing an example of the fluorescent fingerprint of FIG. FIG. 3 is a flowchart showing an example of processing of the harm factor quantification method in the present embodiment. FIG. 4 is a diagram illustrating an example of a fluorescent fingerprint of a measurement object. FIG. 5 is a diagram illustrating an example of data preprocessing performed on a fluorescent fingerprint before multivariate analysis. FIG. 6 is a diagram showing an outline of PLS regression analysis as an example of multivariate analysis. FIG. 7 is a block diagram showing an example of the configuration of the hazard factor quantification apparatus of the present embodiment. FIG. 8 is a block diagram illustrating an example of a fluorescent fingerprint acquisition apparatus. FIG. 9 is a diagram showing chemical analysis values (average) of mold-contaminated flour at four types of contamination (low, medium, high, and straw). FIG. 10 is a diagram showing fluorescent fingerprints of mold-contaminated flour (contamination level: low). FIG. 11 is a diagram showing fluorescent fingerprints of mold-contaminated flour (contamination level: medium). FIG. 12 is a diagram showing fluorescent fingerprints of mold-contaminated flour (contamination level: high). FIG. 13 is a diagram showing fluorescent fingerprints of mold-contaminated flour (contamination level: straw). FIG. 14 is a diagram showing a quantitative result of deoxynivalenol (DON) concentration in mold-contaminated wheat flour by PLS regression analysis. FIG. 15 is a diagram showing a quantitative result of nivalenol (NIV) concentration in mold-contaminated wheat flour by PLS regression analysis. FIG. 16 is a diagram showing a quantitative result of zearalenone (ZEA) concentration in mold-contaminated wheat flour by PLS regression analysis. FIG. 17 is a diagram showing the distribution of the main component score of deoxynivalenol (DON) in mold-contaminated wheat flour by principal component analysis. FIG. 18 is a diagram showing the distribution of deoxynivalenol (DON) discrimination scores in mold-contaminated flour by discriminant analysis. FIG. 19 is a diagram showing the quantification result of the ATP concentration in the ATP aqueous solution by PLS regression analysis. FIG. 20 is a diagram showing a quantitative result of the ATP coating concentration on the stainless steel plate by PLS regression analysis. FIG. 21 is a diagram showing a distribution of discriminant scores of ATP applied on stainless steel by discriminant analysis. FIG. 22 is a diagram showing the discrimination result of the degree of ATP contamination on the stainless steel plate.

Examples of preferred embodiments of [I] hazard factor quantification method and [II] hazard factor quantification apparatus and program according to embodiments of the present invention will be described below in detail with reference to FIGS. . Note that the present invention is not limited to the embodiments.

[I] Hazardous Factor Quantification Method First, fluorescent fingerprints will be described with reference to FIGS. 1 and 2. Here, FIG. 1 is a diagram showing an example of a fluorescent fingerprint as a contour graph of three-dimensional data. FIG. 2 is a bird's-eye view showing an example of the fluorescent fingerprint of FIG.

In the present embodiment, as shown in FIG. 1, the “fluorescence fingerprint” refers to an excitation wavelength λ _Ex irradiating the measurement object, a fluorescence wavelength λ _{Em of} light emitted from the measurement object, and a fluorescence intensity I _Ex of the measurement object. _{, Em} is a contour graph of three-dimensional data composed of three axes.

  Further, as shown in FIG. 2, the “fluorescence fingerprint” can be represented in a plane by plotting the fluorescence intensity at each point in a contour line with the horizontal axis representing the fluorescence wavelength and the vertical axis representing the excitation wavelength. Fluorescent fingerprints have the advantages of being able to characterize the measurement object without pretreatment such as fluorescent staining, being easy to operate and measuring in a short time, and having higher sensitivity than absorbance. Therefore, it is a widely used technique for analysis of the internal structure of food, etc., measurement of suspended matters in the atmosphere, and identification of dye raw materials. Thus, since the “fluorescence fingerprint” is fluorescence information unique to a component having a large amount of three-dimensional information, the measurer can identify the component by using the fluorescence fingerprint and Destructive measurement is possible.

Hereinafter, an example of the processing of the hazard factor quantification method will be described along the flowchart of FIG. 3 with reference to FIGS. 4 to 6 as appropriate. Here, FIG. 3 is a flowchart showing an example of processing of the hazard factor quantification method in the present embodiment.

As shown in FIG. 3, in this hazard factor quantification method, a measurement object to be detected as a hazard factor is first prepared by a measurer as a premise (step S1). Here, in step S <b> 1, when the measurement object is powder, the measurer may perform a homogenization step of homogenizing the measurement object into fine pulverization using a pulverizer.

  Here, in this embodiment, the “measurement object” means a sample that is a measurement object when a hazard factor is detected using a fluorescent fingerprint. As an example, the measurement object is, for example, wheat (for example, wheat, barley, buckwheat, rye, etc.), corn (for example, sweet corn (sweet seed), popcorn (explosive seed), flint corn (hard grain seed), waxy It may be at least one of corn (glutinous species), dent corn (horse teeth species)), food residues, and microorganisms (such as Fusarium). Here, Fusarium bacteria are microorganisms that mainly contaminate grains such as wheat and corn, and are bacteria that cause red mold disease that reduces the yield of grains. In the present embodiment, the measurement object is not limited to these, and may be another sample capable of detecting a hazard factor using a fluorescent fingerprint.

  In the present embodiment, the “hazardous factor” to be detected may be mold poison and / or ATP. Here, in this embodiment, the mold venom may be at least one of deoxynivalenol (DON), nivalenol (NIV), and zearalenone (ZEA). These mold toxins are produced by contamination of the genus Fusarium, and have acute toxicity that induces nausea, vomiting, abdominal pain, dizziness, diarrhea, headaches, etc. in humans and livestock ingested. Therefore, for example, deoxynivalenol (DON) in wheat has a provisional reference value of 1.1 ppm.

  Moreover, in this embodiment, ATP (adenosine triphosphate) is one of the hygiene management indexes that are indexes for performing hygiene management, and the hygiene management indexes indicate the degree of contamination or cleanliness. . For example, in the measuring device (Lumitester PD-20) described in Non-Patent Documents 2 and 3, a numerical value proportional to the amount of ATP is displayed as a hygiene management index. In the present invention, the amount of ATP, which is one of the hygiene management indexes, is estimated using fluorescent fingerprints. Here, for example, the measurer detects potential ATP by detecting ATP even in a poorly washed state where microorganisms have not yet grown, although there are unwashed residues that cause microorganism growth. Can detect danger. Examples of other hygiene management indicators include the number of viable bacteria and the number of microorganisms such as coliforms, Escherichia coli, Staphylococcus aureus, Vibrio parahaemolyticus, Salmonella, and the like. In this way, the hygiene management index is used when, for example, a hygiene manager of a food production / processing line confirms that dirt (for example, food residue, bacteria, dirt, etc.) remains in the cleaned place. .

  In the present embodiment, the harm factor is not limited to these, and any factor that can be detected using a fluorescent fingerprint may be used.

  Specifically, the harm factor may be a factor that causes biological harm. For example, it may be at least one of pathogenic bacteria, spoilage microorganisms, viruses, parasites, and pathogenic microorganisms. Here, the pathogenic bacteria are Salmonella, Vibrio parahaemolyticus, Campylobacter, pathogenic E. coli (including enterohemorrhagic E. coli O-157), Staphylococcus aureus, Clostridium perfringens, Bacillus cereus, Clostridium botulinum, Elysinia enterocolitica, and Listeria -It may be at least one of monocytogenes. The spoilage microorganism may be at least one of Bacillus, Clostridium, Pseudomonas, lactic acid bacteria, and yeast. The virus may also be norovirus and / or hepatitis A virus. Further, the parasite may be at least one of liver fluke, yokokawa fluke, jaw-and-mouth fluke, anisakis, tailed nematode, toxoplasma, cryptosporidium, cyclospora, and giardia. The pathogenic microorganism may be at least one of Shigella, Vibrio cholerae, and Q fever rickettsia.

  Further, the harm factor may be a factor causing chemical harm. For example, it may be at least one of a biological substance, an artificially added substance, and an incidental substance. Here, the biological substance may be at least one of mold poison, shellfish poison, puffer poison, solanine, and histamine. Moreover, food additive is mentioned as what is artificially added. Moreover, what is accidentally present may be at least one of an insecticide, a herbicide, an undesignated additive, a disinfectant, a lubricant, a paint, and a detergent.

  Further, the harm factor may be a factor that causes physical harm. For example, foreign matter may be mixed. Here, the foreign material may be at least one of a glass piece, a metal piece, a wood piece, a plastic piece, an injection needle, a shot piece, an article derived from a worker, a thread, and a wire.

In the hazard factor quantification method, next, in the fluorescence fingerprint information acquisition step, the fluorescence fingerprint of the measurement object prepared by the measurer in step S1 is measured (step S2). That is, in the fluorescence fingerprint information acquisition process of step S2, the fluorescence intensity in the measurement object is measured while changing the excitation wavelength to be irradiated and the fluorescence wavelength to be observed in a stepwise manner within a predetermined excitation wavelength range and a predetermined fluorescence wavelength range. Then, fluorescent fingerprint information is acquired. Here, in the fluorescent fingerprint information acquisition process of step S2, the fluorescent fingerprint information of the measurement object that has been homogenized and finely pulverized in the homogenization process of step S1 may be acquired. That is, in the fluorescence fingerprint information acquisition process in step S2, the excitation wavelength (for example, m wavelengths for each data acquisition interval of 10 nm within the measurement wavelength range of 200 to 900 nm) and the fluorescence wavelength are measured by an existing spectrofluorophotometer. While changing the combination of (for example, n wavelengths at a data acquisition interval of 10 nm in the measurement wavelength range of 200 to 900 nm), the fluorescence intensity of the measurement object is acquired under a total of m × n wavelength conditions. Here, the measurer may adjust the number of times of measurement (for example, three times for each sample). Here, FIG. 4 is a diagram illustrating an example of the fluorescent fingerprint of the measurement object.

  As described above, in the fluorescence fingerprint information acquisition step of step S2, the fluorescence intensity is acquired for the measurement object under the m × n wavelength condition in which the combination of the excitation wavelength to be irradiated and the fluorescence wavelength to be observed is different, and the fluorescence fingerprint of the measurement object is acquired. Get information.

Further, in the fluorescent fingerprint information acquisition process of step S2, data preprocessing is performed on the acquired fluorescent fingerprint before the multivariate analysis performed in the hazard factor determination process of the next step S3.

  Hereinafter, with reference to FIG. 4 and FIG. 5, data preprocessing performed on the fluorescent fingerprint before multivariate analysis will be described. Here, FIG. 5 is a diagram illustrating an example of data preprocessing performed on a fluorescent fingerprint before multivariate analysis.

As shown in FIG. 4, the fluorescent fingerprint acquired in the fluorescent fingerprint information acquisition step of step S2 includes high-dimensional fluorescent intensity data consisting of parameters of a total of m × n wavelength conditions (for example, 5041 wavelength conditions). Furthermore, noise information (for example, scattered light of excitation light, its secondary light, tertiary light, quaternary light, etc.) is included. Therefore, before the multivariate analysis performed in the hazard factor quantification process in the next step S3, the noise information is removed in the fluorescence fingerprint information acquisition process in step S2, and the acquired high-dimensional (for example, 5041 wavelength condition) fluorescence. It is desirable to use low-dimensional (for example, 2063 wavelength condition) fluorescence intensity data obtained by removing information other than the target fluorescence fingerprint from the intensity data.

  Therefore, as shown in FIG. 5, in the fluorescence fingerprint information acquisition step of step S <b> 2, excitation light confusion light that becomes noise information and its secondary light and tertiary light data (for example, data shown in FIG. 5 (i)). ) Is deleted. Further, in the fluorescence fingerprint information acquisition step of step S2, data in a range where the fluorescence wavelength is shorter than the excitation wavelength (for example, data shown in FIG. 5 (ii)) is removed. This is a process for analyzing only fluorescence intensity data of a fluorescence wavelength longer than the excitation wavelength because the fluorescence wavelength emitted from the measurement object is longer than the excitation wavelength. Although not shown in the drawing, in the fluorescence fingerprint information acquisition step of step S2, data of a region where the luminance value of the excitation light source is low (for example, excitation wavelength 240 nm or less) and a region where the sensitivity of fluorescence and detector is low (for example, fluorescence wavelength) (800 nm or more) data may be deleted. As described above, in the fluorescent fingerprint information acquisition step of step S2, only the fluorescent fingerprint information representing the characteristics of the harm factor included in the measurement object is extracted.

Then, in the fluorescent fingerprint information acquisition step of step S2, the matrix data obtained from the fluorescent fingerprint information as shown in FIG. 5 is rearranged and changed to a continuous one-dimensional vector. The fluorescence fingerprint information changed to the one-dimensional vector (for example, 2063 fluorescence intensity data) is used for multivariate analysis in the hazard factor quantification step of the next step S3.

Returning to FIG. 3, in the hazard factor quantification method, in the hazard factor quantification step, multivariate analysis (for example, PLS regression) is performed on the fluorescence fingerprint information of the measurement object acquired in the fluorescence fingerprint information acquisition step of step S <b> 2. Analysis, principal component analysis, discriminant analysis, etc.), and based on the result of the multivariate analysis, a hazard factor is detected from the measurement object (step S3).

Specifically, in the hazard factor quantification step in step S3, the amount of the hazard factor is estimated by performing PLS regression analysis as a multivariate analysis on the fluorescence fingerprint information acquired in the fluorescence fingerprint information acquisition step in step S2. Based on the quantified result of the estimated hazard factor (for example, FIG. 14 to FIG. 17, FIG. 19 and FIG. 20), the hazard factor is detected from the measurement object. Further, in the hazard factor quantification process of step S3, principal component analysis is performed as multivariate analysis on the fluorescence fingerprint information acquired in the fluorescence fingerprint information acquisition process of step S2, and the hazard factor obtained by the principal component analysis is determined. Based on the distribution of the main component scores (for example, FIG. 17), the harm factor may be detected from the measurement object. Further, in the hazard factor quantification step of step S3, the fluorescence fingerprint information acquired in the fluorescence fingerprint information acquisition step of step S2 is subjected to discriminant analysis as multivariate analysis, and the hazard factor discrimination score obtained by the discriminant analysis is obtained. Based on the distribution (for example, FIG. 21), a hazard factor may be detected from the measurement object. The hazard factor quantification results (for example, FIGS. 14 to 17, 19, and 20) obtained as a result of the multivariate analysis in the hazard factor quantification step of step S 3, the principal component scores of the hazard factors are obtained. An example of the distribution (for example, FIG. 17) and the distribution of the hazard factor discrimination score (for example, FIG. 19) will be described in detail in an embodiment described later.

Here, with reference to FIG. 6, the process which estimates the quantity of a harm factor by performing PLS regression analysis as an example of the multivariate analysis performed in the hazard factor determination process of step S3 is demonstrated. FIG. 6 is a diagram showing an outline of PLS regression analysis as an example of multivariate analysis.

As shown in FIG. 6, in the hazard factor determination step of step S3, first, the fluorescence fingerprint information (m × n fluorescence intensity data in FIG. 6) acquired in the fluorescence fingerprint information acquisition step of step S2 is used as an explanatory variable. Based on the PLS (Partial Last Squares) algorithm, the explanatory variable is converted into a latent variable (T _{1 to} T _k in FIG. 6), and the latent variable is substituted into the regression equation, thereby harming the target variable. Estimate the factor concentration (ie, the amount of the hazard). Then, in the hazard factor determination step of step S3, the fluorescence fingerprint information of the measurement object of unknown concentration is applied to a calibration curve created based on a known or previously measured hazard factor, and the hazard included in the measurement object Quantify the concentration of the factor. Thus, in the hazard factor determination step, the amount of the hazard factor is estimated by performing PLS regression analysis on the fluorescence fingerprint information (m × n fluorescence intensity data in FIG. 6), and the estimated Based on the quantification result of the hazard factor (for example, FIG. 14 to FIG. 17, FIG. 19 and FIG. 20), the hazard factor is detected from the measurement object.

In this embodiment, PLS regression analysis, principal component analysis, and discriminant analysis are given as examples of multivariate analysis performed in the hazard factor determination step of S3. However, the present invention is not limited to these, and any statistical analysis processing is performed. You may go.

For example, in the hazard factor determination step of S3, the statistical analysis processing may be performed by multivariate analysis or data mining. Further, in the hazard factor determination step of S3, multivariate analysis or data mining may be performed using at least one method of data structure analysis, discriminant analysis, pattern classification, multidimensional data analysis, regression analysis, and learning machine. Good. Further, in the hazard factor determination step of S3, the data structure analysis may be performed using at least one technique among principal component analysis, factor analysis, correspondence analysis, and independent component analysis. Further, in the hazard factor determination step of S3, the discriminant analysis may be performed using a linear discriminant analysis method or a nonlinear discriminant analysis method. In the hazard factor determination step of S3, pattern classification may be performed by cluster analysis or multidimensional scaling. In the hazard factor determination step of S3, the regression analysis may be performed using a linear regression, nonlinear regression, or multiple regression analysis technique. In the hazard factor determination step of S3, the learning machine may perform at least one of a neural network, a support vector machine, a self-organizing map, group learning, and a genetic algorithm.

  In addition, in this embodiment, various conditions such as a measurement object, a measurement condition of a fluorescent fingerprint, and a statistical analysis method can be selected as appropriate without departing from the object of the present invention.

[II] Hazard Factor Detection Device Next, the configuration of the hazard factor detection device of the present invention will be described with reference to FIG. 7 and FIG. In addition, although the hazard factor detection apparatus in embodiment of this invention can be used suitably for the above-mentioned hazard factor detection method mentioned above, the measuring apparatus used for this hazard factor detection method is not limited to this. .

Here, FIG. 7 is a block diagram showing an example of the configuration of the hazard factor quantification apparatus of the present embodiment, and conceptually shows only the portion related to the present invention in the configuration.

As shown in FIG. 7, the hazard factor quantification apparatus 20 includes at least a fluorescent fingerprint acquisition apparatus 10. The fluorescent fingerprint acquisition device 10 is a device that acquires fluorescent fingerprint information, and includes a spectral illumination device 11 and a spectral detection device 12. The hazard factor quantification device 20 is a device that detects the hazard factor of the measurement target 13 from the fluorescence fingerprint information acquired by the fluorescence fingerprint acquisition device 10, and includes a memory 21, a control unit 23, and a calculation processing unit 24. The measurer inputs the measurement conditions and the like to the hazard factor determination device 20 by using the keyboard / mouse 22. Although not shown, the hazard factor quantification device 20 may include a crushing device that uniformizes the measurement object 13 into fine crushing. Here, the pulverizer may be, for example, a coarse pulverizer, a hammer mill pulverizer, a jet mill pulverizer, or a cyclone mill pulverizer.

  Here, FIG. 8 is a block diagram showing an example of the fluorescence fingerprint acquisition apparatus 10, and conceptually shows only the part related to the present invention in the configuration.

  As shown in FIG. 8, the fluorescent fingerprint acquisition apparatus 10 includes a spectral illumination device 11 and a spectral detection device 12. The spectroscopic illumination device 11 is a device that irradiates the measurement target 13 with excitation light having a predetermined wavelength to generate fluorescence from the components of the measurement target 13. The spectral illumination device 11 has excitation wavelength variable means for arbitrarily changing the excitation wavelength to be irradiated.

The spectroscopic detection device 12 is a device that acquires the fluorescence intensity of the measurement object 13 at a predetermined fluorescence wavelength and transmits the fluorescence intensity information to the hazard factor quantification apparatus 20. The spectroscopic detection device 12 selectively captures a specific fluorescence wavelength among the fluorescence emitted from the measurement object 13 and measures the fluorescence intensity. The spectroscopic detection device 12 includes fluorescence wavelength variable means for arbitrarily changing the observed fluorescence wavelength.

Here, returning to FIG. 7, the hazard factor quantification apparatus 20 will be described.

As shown in FIG. 7, the hazard factor quantification device 20 acquires the fluorescence fingerprint information detected by the fingerprint detection device 124 of the spectroscopic detection device 12, and is necessary for the analysis by the data preprocessing shown in FIGS. Only the fluorescent fingerprint information is extracted (in FIG. 7, the processing by the fluorescent fingerprint information acquisition unit 24-1). Next, multivariate analysis (for example, PLS regression analysis, principal component analysis, discriminant analysis, etc.) is performed on the acquired fluorescence fingerprint information of the measurement object 13, and the measurement object 13 is based on the result of the multivariate analysis. From FIG. 7, a hazard factor is detected (in FIG. 7, processing by the hazard factor quantification unit 24-2).

The memory 21 is transferred from the fingerprint detection device 124 to the hazard factor quantification device 20, and the fluorescence fingerprint information acquired by the fluorescence fingerprint information acquisition unit 24-1 of the calculation processing unit 24 or the hazard factor determination unit 24 of the calculation processing unit 24. -2 stores regression equations and discriminants used when performing multivariate analysis, concentration data of hazard factors measured in advance, which are referred to as a control (comparison control) of multivariate analysis results, and the like.

  The control unit 23 irradiates the spectral illumination device 11 with the excitation wavelength and the spectral detection device 12 so as to acquire the fluorescence fingerprint of the measurement object 13 with the excitation wavelength range, fluorescence wavelength range, and wavelength pitch input by the measurer. Gives an instruction to adjust the observed fluorescence wavelength, and instructs the calculation processing unit 24 to perform the process.

The fluorescent fingerprint information transferred from the fluorescent fingerprint acquisition apparatus 10 is stored in the memory 21 of the hazard factor quantification apparatus 20. When the measurer instructs the calculation processing unit 24 to perform processing through the keyboard / mouse 22, first, the fluorescent fingerprint information acquisition unit 24-1 of the calculation processing unit 24 uses the fluorescent fingerprint information stored in the memory 21. Fluorescent fingerprint information necessary for analysis is extracted by data preprocessing. Next, the hazard factor quantification unit 24-2 of the calculation processing unit 24 performs multivariate analysis on the extracted fluorescent fingerprint information, and based on the result of the multivariate analysis, the hazard factor is determined from the measurement object. Detect.

In the present embodiment, the hazard factor quantification unit 24-2 of the calculation processing unit 24 estimates the amount of the hazard factor by performing PLS regression analysis as multivariate analysis on the fluorescent fingerprint information, and the estimated hazard factor. Based on the quantification result, a hazard factor may be detected from the measurement object. The hazard factor quantification unit 24-2 of the calculation processing unit 24 performs principal component analysis as multivariate analysis on the fluorescence fingerprint information, and based on the distribution of the principal component scores of the hazard factors obtained by the principal component analysis. Thus, the hazard factor may be detected from the measurement object. Further, the hazard factor quantification unit 24-2 of the calculation processing unit 24 performs discriminant analysis as multivariate analysis on the fluorescent fingerprint information, and performs measurement based on the distribution of the hazard factor discrimination score obtained by the discriminant analysis. A hazard may be detected from the object. Further, the hazard factor quantification unit 24-2 of the calculation processing unit 24 may store the result of the multivariate analysis in the memory 21, output it on the display 30, or a printer (not shown). You may print via.

  Hereinafter, embodiments of the present invention will be described with reference to FIGS. The present invention is not limited to these examples.

[Example 1]
In Example 1, as shown in FIG. 9, mold-contaminated flour at four types of contamination (low, medium, high, and straw) was used as the measurement object. Here, FIG. 9 is a diagram showing chemical analysis values (average) of mold-contaminated flour at four types of contamination (low, medium, high, and straw). Specifically, first, fluorescence fingerprint information of mold-contaminated flour at four types of contamination (low, medium, high, and straw) was acquired by the processing of the fluorescence fingerprint information acquisition unit 24-1. Then, by the processing of the hazard factor quantification unit 24-2, the PLS regression analysis is performed on the fluorescence fingerprint information of each mold-contaminated flour, so that the hazard factors (DON, NIV, and ZEA of the mold poison in the first embodiment). The amount was estimated, and the hazard was detected from the measurement object based on the quantitative result of the estimated hazard. Details of the first embodiment will be described below with reference to FIGS.

(1-1. Preparation of measurement object)
In the present Example 1, mold-contaminated flour in four types of contamination (low, medium, high, and straw) was used as the measurement object. And each mold-contaminated wheat flour 240mg was put into the powder cell, and it enclosed with the quartz window. In addition, the mold-contaminated flour sample prepared in Example 1 is a total of 60 points (4 contaminated areas × 15 points each).

(1-2. Acquisition of fluorescent fingerprint information)
Then, the powder cell which enclosed each mold-contaminated wheat flour 240mg was set to the fluorescence fingerprint acquisition apparatus 10 (Hitachi spectrofluorometer F-7000). And the fluorescence fingerprint information of each mold | contamination flour was acquired by the process of the fluorescence fingerprint information acquisition part 24-1. Here, the measurement conditions are as follows.
Measurement wavelength range: excitation wavelength (Ex) 200 to 900 nm / fluorescence wavelength (Em) 200 to 900 nm
-Data acquisition interval: 10 nm

  Specifically, the fluorescence fingerprint information of each mold-contaminated flour shown in FIGS. 10 to 13 was acquired by the processing of the fluorescence fingerprint information acquisition unit 24-1 (however, the fluorescence fingerprint information displayed in FIGS. 10 to 13). Is before data preprocessing). Here, FIG. 10 is a diagram showing fluorescent fingerprints of mold-contaminated flour (contamination level: low). Moreover, FIG. 11 is a figure which shows the fluorescence fingerprint of mold-contaminated wheat flour (contamination degree: medium). FIG. 12 is a diagram showing fluorescent fingerprints of mold-contaminated flour (contamination level: high). Moreover, FIG. 13 is a figure which shows the fluorescence fingerprint of mold-contaminated flour (contamination degree: straw).

(1-3. Detection of hazards)
Subsequently, PLS regression analysis is applied to the fluorescence fingerprint information of each acquired mold-contaminated flour to estimate the concentration of the mold poison (DON, NIV, ZEA in Example 1) contained in each mold-contaminated flour. went.

That is, by the processing of the hazard factor quantification unit 24-2, the hazard factor (this example 1) is obtained by performing PLS regression analysis on the fluorescence fingerprint information of each mold-contaminated flour (contamination degree: low, medium, high, and straw). The amount of mold poisons DON, NIV, and ZEA) was estimated. Specifically, it is obtained by preprocessing the fluorescence fingerprint information (that is, the fluorescence fingerprint) of each mold-contaminated flour (contamination level: low, medium, high, cocoon) by the processing of the hazard factor quantification unit 24-2. The luminance value of m × n fluorescent fingerprints) is used as an explanatory variable, the explanatory variable is converted into a latent variable, and the latent variable is substituted into a regression equation, whereby the hazard variable concentration of the objective variable (that is, FIG. 14). "DON predicted value (ppm)", "NIV predicted value (ppb)" shown in FIG. 15, and "ZEA predicted value (ppm)" shown in FIG. 16) were estimated.

Then, by the processing of the hazard factor quantification unit 24-2, for example, as shown in FIGS. 14 to 16, based on the quantification result of the estimated hazard factor (DON, NIV, ZEA of mold poison in the first embodiment). A hazard was detected from the measurement object (mold-contaminated flour in four types of contamination (low, medium, high, and straw) in Example 1). Here, FIG. 14 is a figure which shows the fixed_quantity | quantitative_assay result of the deoxynivalenol (DON) density | concentration in the mold-contaminated wheat flour by PLS regression analysis. Moreover, FIG. 15 is a figure which shows the fixed_quantity | quantitative_assay result of the nivalenol (NIV) density | concentration in mold-contaminated wheat flour by PLS regression analysis. Moreover, FIG. 16 is a figure which shows the fixed_quantity | quantitative_assay result of the zearalenone (ZEA) density | concentration in mold-contaminated wheat flour by PLS regression analysis.

  Hereinafter, with reference to FIG. 14 to FIG. 16, the quantitative results of each mold poison (DON, NIV, ZEA in Example 1) will be described.

Specifically, first, as shown in FIG. 14, the concentration of mold toxin (deoxynivalenol (DON) in FIG. 14) estimated by the processing of the hazard factor quantification unit 24-2 (in FIG. 14, “DON predicted value”). (Ppm) ") is in good agreement with the deoxynivalenol (DON) concentration (" DON actual measurement value (ppm) "in FIG. 14) of the chemical analysis values shown in FIG. It shows that the concentration of each hazard factor (deoxynivalenol (DON) in FIG. 14) contained in the measurement object can be quantified.

As shown in FIG. 14, the coefficient of determination (R ² ) of the predicted DON value (ppm) (in FIG. 14, each plot indicating the four types of contamination (low, medium, high, and haze)) is 0.990. became. Here, the closer the determination coefficient is to 1, the higher the accuracy. Therefore, it was shown that the accuracy of the DON predicted value (ppm) estimated by the processing of the hazard factor quantification unit 24-2 is very high. Therefore, according to the present Example 1, it is possible to measure the low-concentration mold toxin (deoxynivalenol (DON) in FIG. 14) quantitatively as a hazard factor from the fluorescent fingerprint of mold-contaminated flour quickly and easily by PLS regression analysis. It was shown that there is.

Subsequently, as shown in FIG. 15, the concentration of mold toxin (nivalenol (NIV) in FIG. 15) estimated by the processing of the hazard factor quantifying unit 24-2 (“NIV predicted value (ppb)” in FIG. 15)) Is in good agreement with the nivalenol (NIV) concentration of chemical analysis values shown in FIG. 9 (“NIV actual measurement value (ppb)” in FIG. 15) (in FIG. 15, four kinds of contamination levels (low, medium, Each plot showing high and wrinkles) shows that the concentration of the hazard factor (nivalenol (NIV) in FIG. 15) contained in the measurement object can be quantified.

As shown in FIG. 15, the coefficient of determination (R ² ) of the predicted NIV value (ppb) (in FIG. 15, each plot showing the four types of contamination (low, medium, high, and wrinkle)) is 0.977. became. Here, the closer the determination coefficient is to 1, the higher the accuracy. Therefore, it was shown that the accuracy of the predicted NIV value (ppb) estimated by the processing of the hazard factor quantification unit 24-2 is very high. Therefore, according to the present Example 1, it is possible to quickly and easily measure a low concentration of mold venom (nivalenol (NIV) in FIG. 15) as a harmful factor from a fluorescent fingerprint of mold-contaminated flour by PLS regression analysis. It was shown that.

Subsequently, as shown in FIG. 16, the concentration of mold poison (in FIG. 16, zearalenone (ZEA)) estimated by the processing of the hazard factor quantification unit 24-2 (in FIG. 16, “ZEA predicted value (ppm)”) Is in good agreement with the zearalenone (ZEA) concentration of chemical analysis values shown in FIG. 9 (“ZEA actual measurement value (ppm)” in FIG. 16) (in FIG. 16, four kinds of contamination levels (low, medium, Each plot showing high, 甚) shows that the concentration of the hazard factor (Zearalenone (ZEA) in FIG. 16) contained in the measurement object can be quantified.

As shown in FIG. 16, the coefficient of determination (R ² ) of the predicted ZEA value (ppm) (in FIG. 16, each plot showing the four types of contamination (low, medium, high, haze)) is 0.989. became. Here, the closer the determination coefficient is to 1, the higher the accuracy. Therefore, it was shown that the accuracy of the predicted ZEA value (ppm) estimated by the processing of the hazard factor quantification unit 24-2 is very high. Therefore, according to the present Example 1, it is possible to quantitatively measure a low concentration of mold poison (Zearalenone (ZEA) in FIG. 16) as a harmful factor from a fluorescent fingerprint of mold-contaminated flour quickly and easily by PLS regression analysis. It was shown that.

[Example 2]
In Example 2, as in Example 1, mold-contaminated flour at four types of contamination (low, medium, high, and straw) was used as the measurement object. However, in the present Example 2, as an example, only deoxynivalenol (DON) among mold poisons was set as the detection target of the harm factor. Specifically, first, fluorescence fingerprint information of mold-contaminated flour at four types of contamination (low, medium, high, and straw) was acquired by the processing of the fluorescence fingerprint information acquisition unit 24-1. Then, the principal component analysis is performed on the fluorescent fingerprint information of each mold-contaminated flour by the processing of the hazard factor quantification unit 24-2, and the measurement object is based on the distribution of the principal component scores obtained by the principal component analysis. From the above, a harmful factor (deoxynivalenol (DON) of mold poison in Example 2) was detected. Details of the second embodiment will be described below with reference to FIG.

(2-1. Preparation of measurement object)
In the present Example 2, as in Example 1, mold-contaminated flour in four types of contamination (low, medium, high, and straw) was used as the measurement object. And each mold-contaminated wheat flour 240mg was put into the powder cell, and it enclosed with the quartz window. As in Example 1, the total number of mold-contaminated flour samples prepared in Example 2 is 60 points (4 contaminated areas × 15 points each).

(2-2. Acquisition of fluorescence fingerprint information)
Subsequently, in the same manner as in Example 1, a powder cell in which 240 mg of each mold-contaminated flour was encapsulated was set in the fluorescence fingerprint acquisition apparatus 10 (Hitachi spectrofluorometer F-7000). And the fluorescence fingerprint information of each mold | contamination flour was acquired by the process of the fluorescence fingerprint information acquisition part 24-1. The measurement conditions are the same as in Example 1. Specifically, the fluorescence fingerprint information of each mold-contaminated flour shown in FIGS. 10 to 13 was obtained by the processing of the fluorescence fingerprint information acquisition unit 24-1.

(2-3. Detection of hazards)
Subsequently, a principal component analysis was performed on the fluorescence fingerprint information of each acquired mold-contaminated flour. Specifically, the principal component analysis is performed on the fluorescence fingerprint information of each mold-contaminated flour by the processing of the hazard factor quantification unit 24-2, and the measurement object represented by the main component (in the second embodiment, Principal component score plots (for example, the distribution of principal component scores shown in FIG. 17) were created by plotting principal component scores, which are values of mold-contaminated flour at four types of contamination, as coordinates.

Then, as shown in FIG. 17 by the processing of the hazard factor quantification unit 24-2, the distribution of the main component scores of the hazard factor obtained (deoxynivalenol (DON) of mold poison in Example 2) (FIG. 17). On the basis of the plots shown in FIG. 17, a hazard factor (mold poison deoxynivalenol (DON) indicated by a plot group (contamination level: 甚) in FIG. 17) was detected from the measurement object. Here, FIG. 17 is a diagram showing a distribution of main component scores of deoxynivalenol (DON) in mold-contaminated wheat flour by principal component analysis.

  As shown in FIG. 17, in the upper right part of the figure, a plot group consisting of a round mark, a square mark, and a cross mark (contamination level: low, medium, high), and in the lower right part of the figure, Y It was shown that the two plot groups, including the plot group consisting of the marked plots (contamination level: 甚), were clearly separated and dispersed. Thereby, it was shown that the plot group (contamination level: 甚) marked with Y is clearly different from other samples and can be discriminated. Therefore, according to the present Example 2, it is possible to determine the degree of the hazard (deoxynivalenol of mold poison (contamination level: 甚)) in the main component analysis from the fluorescent fingerprint of the mold-contaminated flour. Indicated.

[Example 3]
In Example 3, as in Example 1 and Example 2, mold-contaminated flour at four types of contamination (low, medium, high, and straw) was used as the measurement object. In Example 3, similarly to Example 2, as an example, only deoxynivalenol (DON) among mold poisons was detected as a hazard factor. Specifically, first, fluorescence fingerprint information of mold-contaminated flour at four types of contamination (low, medium, high, and straw) was acquired by the processing of the fluorescence fingerprint information acquisition unit 24-1. Then, by the processing of the hazard factor quantifying unit 24-2, the fluorescence fingerprint information of each mold-contaminated flour is discriminated and analyzed, and the hazard factor obtained by the discriminant analysis (in this example 3, deoxynivalenol of mold toxin). Based on the score distribution of (DON)), the degree of the harm factor was determined from the measurement object. Details of the third embodiment will be described below with reference to FIG.

(3-1. Preparation of measurement object)
In Example 3, as in Example 2, mold-contaminated flour in four types of contamination (low, medium, high, and straw) was used as the measurement object. And each mold-contaminated wheat flour 240mg was put into the powder cell, and it enclosed with the quartz window. As in Example 2, the total number of mold-contaminated flour samples prepared in Example 3 is 60 points (4 contaminated areas × 15 points each).

(3-2. Acquisition of fluorescent fingerprint information)
Subsequently, in the same manner as in Example 2, a powder cell in which 240 mg of each mold-contaminated flour was encapsulated was set in the fluorescence fingerprint acquisition apparatus 10 (Hitachi Spectrofluorometer F-7000). And the fluorescence fingerprint information of each mold | contamination flour was acquired by the process of the fluorescence fingerprint information acquisition part 24-1. The measurement conditions are the same as in Example 2. Specifically, the fluorescence fingerprint information of each mold-contaminated flour shown in FIGS. 10 to 13 was obtained by the processing of the fluorescence fingerprint information acquisition unit 24-1.

(3-3. Detection of hazards)
Subsequently, discriminant analysis was performed on the fluorescence fingerprint information of each acquired mold-contaminated flour. Specifically, by the processing of the hazard factor quantification unit 24-2, the discriminant analysis was performed by substituting fluorescent fingerprint information of each mold-contaminated flour as a parameter into a discriminant prepared in advance.

Then, as shown in FIG. 18, by the processing of the harm factor quantification unit 24-2, the distribution of the discrimination score of the obtained hazard factor (mold poison deoxynivalenol (DON) in Example 3) (in FIG. 18). Based on the four types of pollution levels (low, medium, high, and flaw plots), hazard factors were detected from the measurement object. Here, FIG. 18 is a diagram showing the distribution of the discrimination score of deoxynivalenol (DON) in mold-contaminated flour by discriminant analysis.

  As shown in FIG. 18, a plot group consisting of a Y-marked plot (contamination level: に) in the lower left part of the figure, and a cross-marked plot (contamination level: high) in the upper right part of the figure. In the lower right part of the figure, there are four plot groups, a plot group consisting of a circle-marked plot (contamination level: low) and a plot group consisting of a square-marked plot (contamination level: medium). It was shown that it was divided and distributed. As a result, it was possible to discriminate between four types of contamination, and in particular, the plot group of Y (contamination level: 甚) was far from the other three types, indicating that the concentration was high. Therefore, according to the present Example 3, four types of contamination levels (low, medium, high, cocoon) of the hazard factor (deoxynivalenol of mold poison in this Example 3) are determined by discriminant analysis from the fluorescent fingerprint of mold-contaminated flour. Shown to be discernable.

[Example 4]
In Example 4, the quantitative property was verified using an ATP aqueous solution (ATP reagent) as a measurement object. Specifically, first, fluorescent fingerprint information of the ATP aqueous solution was acquired by the processing of the fluorescent fingerprint information acquisition unit 24-1. Then, the amount of the hazard factor (ATP in the fourth embodiment) is estimated by performing the PLS regression analysis on the fluorescent fingerprint information of the ATP aqueous solution by the processing of the hazard factor determination unit 24-2, and the estimated hazard factor Based on the quantitative results, the hazards were detected from the measurement object. Details of the fourth embodiment will be described below with reference to FIG.

(4-1. Preparation of measurement object)
In the present Example 4, ATP aqueous solution in 13 types of density | concentration was prepared as a measuring object. Specifically, a 100 mM ATP aqueous solution was prepared as a stock solution using an ATP freezing / drying reagent and 25 mM Tris buffer. Then, using the same 25 mM Tris buffer, 13 concentrations (ie, (1) 1.0 × 10 ⁻⁴ [mol / l], (2) 5.0 × 10 ⁻⁵ [mol / l]) (3) 1.0 × 10 ⁻⁵ [mol / l], (4) 5.0 × 10 ⁻⁶ [mol / l], (5) 1.0 × 10 ⁻⁶ [mol / l], ( 6) 5.0 × 10 ⁻⁷ [mol / l], (7) 1.0 × 10 ⁻⁷ [mol / l], (8) 5.0 × 10 ⁻⁸ [mol / l], (9) 4. 1.0 × 10 ⁻⁸ [mol / l], (10) 5.0 × 10 ⁻⁹ [mol / l], (11) 1.0 × 10 ⁻⁹ [mol / l], (12) 0 × 10 ⁻¹⁰ [mol / l], (13) 1.0 × 10 ⁻¹⁰ [mol / l]). And 200 microliters of each ATP aqueous solution was inject | poured into the quartz cuvette cell.

(4-2. Acquisition of fluorescence fingerprint information)
Subsequently, the quartz cuvette cell into which 200 μL of each ATP aqueous solution was injected was set in the fluorescence fingerprint acquisition apparatus 10 (Hitachi spectrofluorometer F-7000). And the fluorescence fingerprint information of each ATP aqueous solution was acquired by the process of the fluorescence fingerprint information acquisition part 24-1. That is, the fluorescence fingerprint of ATP, which is a hygiene management index indicating the degree of contamination (or cleanliness), was measured by the processing of the fluorescence fingerprint information acquisition unit 24-1. Here, the measurement conditions are as follows.
Measurement wavelength range: excitation wavelength (Ex) 200 to 900 nm / fluorescence wavelength (Em) 200 to 900 nm
-Data acquisition interval: 10 nm
・ Number of measurements: 2 times for each sample

(4-3. Detection of hazards)
Subsequently, PLS regression analysis was applied to the obtained fluorescence fingerprint information of each ATP aqueous solution to estimate the concentration of ATP contained in each ATP aqueous solution.

That is, the amount of the hazard factor (ATP in Example 4) was estimated by performing PLS regression analysis on the fluorescence fingerprint information of each ATP aqueous solution by the processing of the hazard factor quantification unit 24-2. Specifically, the fluorescent fingerprint information of each ATP aqueous solution (that is, the luminance value of mxn fluorescent fingerprints obtained by preprocessing the fluorescent fingerprint data) is obtained by the processing of the hazard factor quantification unit 24-2. By converting the explanatory variable into a latent variable, and substituting the latent variable into the regression equation, the hazard variable concentration of the objective variable (ie, “log (ATP [M]) predicted value LV05 shown in FIG. 19) is used. Where LV05 indicates that there are five latent variables.

Then, by the processing of the hazard factor quantification unit 24-2, for example, as shown in FIG. 19, based on the quantification result of the estimated hazard factor (ATP in the fourth embodiment), the measurement object (the fourth embodiment). In Fig. 1, the hazard factor was detected from the ATP aqueous solution at 13 concentrations. Here, FIG. 19 is a figure which shows the fixed_quantity | quantitative_assay result of the ATP density | concentration in ATP aqueous solution by PLS regression analysis. In FIG. 19, since the amount of ATP corresponds to the number of bacterial growth, both axes are expressed logarithmically.

As shown in FIG. 19, the determination coefficient (R ² ) of “log (ATP [M]) predicted value LV05” (indicated by a circle in FIG. 19) was 0.9903. Here, the closer the determination coefficient is to 1, the higher the accuracy. Therefore, it was shown that the accuracy of the log (ATP [M]) predicted value estimated by the processing of the hazard factor quantification unit 24-2 is very high. Therefore, according to Example 4, it was shown that a low concentration of ATP can be quantitatively measured as a hazard factor quickly and easily from a fluorescent fingerprint of an ATP aqueous solution by PLS regression analysis.

[Example 5]
In Example 5, assuming a stainless steel device installed at a food processing site such as a factory, ATP traces on the stainless steel were verified. Specifically, first, fluorescent fingerprint information of the stainless steel plate coated with the ATP solution was acquired by the processing of the fluorescent fingerprint information acquisition unit 24-1. Then, the amount of the hazard factor (ATP in this embodiment 5) is estimated by performing PLS regression analysis on the fluorescent fingerprint information of the stainless steel plate coated with the ATP solution by the processing of the hazard factor quantification unit 24-2. Based on the quantitative result of the estimated hazard factor, the hazard factor was detected from the measurement object. Details of the fifth embodiment will be described below with reference to FIG.

(5-1. Preparation of measurement object)
In this Example 5, similarly to Example 4, ATP aqueous solutions at 13 types of concentrations were prepared as measurement objects. And 10 microliters of each ATP aqueous solution was dripped on the stainless steel plate (SUS plate) using the 20 mL capacity | capacitance micropipette. Specifically, when the stainless steel plate was installed in the powder cell holder in the fluorescent fingerprint acquisition apparatus 10, 10 μL of each ATP aqueous solution was dropped onto the position on the stainless steel plate to be the excitation light irradiated region. Then, the stainless steel plate was covered so that dust and the like did not adhere, and was naturally dried indoors. Here, the temperature in the naturally dried room was about 26 ° C., and the humidity was about 30%. In this manner, two stainless steel plates were prepared for each ATP aqueous solution, and 13 types of ATP solutions with different concentrations were applied and dried.

(5-2. Acquisition of fluorescent fingerprint information)
Subsequently, the stainless steel plate coated with 10 μL of each ATP solution was set in the fluorescence fingerprint acquisition device 10 (Hitachi spectrofluorometer F-7000). And the fluorescence fingerprint information of each stainless steel plate was acquired by the process of the fluorescence fingerprint information acquisition part 24-1. The measurement conditions are the same as in Example 4.

(5-3. Detection of hazards)
Subsequently, PLS regression analysis was applied to the obtained fluorescent fingerprint information of each stainless steel plate, and the ATP coating concentration of each stainless steel plate was estimated.

That is, the amount of the hazard factor (ATP in Example 5) was estimated by performing the PLS regression analysis on the fluorescent fingerprint information of each stainless steel plate by the processing of the hazard factor quantification unit 24-2. Specifically, the fluorescent fingerprint information of each stainless steel plate (that is, the luminance value of mxn fluorescent fingerprints obtained by preprocessing the fluorescent fingerprint data) is obtained by the processing of the hazard factor quantifying unit 24-2. By converting the explanatory variable into a latent variable, and substituting the latent variable into the regression equation, the hazard variable concentration of the objective variable (that is, “log (ATP [M]) predicted value LV09 shown in FIG. Here, LV09 indicates that there are nine latent variables.

Then, by the processing of the hazard factor quantification unit 24-2, for example, as shown in FIG. 20, based on the quantification result of the estimated hazard factor (ATP in the fifth embodiment), the measurement object (the fifth embodiment). In FIG. 1, the hazard factor was detected from a stainless plate coated with an ATP aqueous solution at 13 different concentrations. Here, FIG. 20 is a figure which shows the fixed_quantity | quantitative_assay result of the ATP application density | concentration on a stainless steel plate by PLS regression analysis. In FIG. 20, both axes are represented by logarithmic transformation as in FIG.

As shown in FIG. 20, the determination coefficient (R ² ) of “log (ATP [M]) predicted value LV09” (the circled plot in FIG. 20) was 0.9780. Here, the closer the determination coefficient is to 1, the higher the accuracy. Therefore, it was shown that the accuracy of the log (ATP [M]) predicted value estimated by the processing of the hazard factor quantification unit 24-2 is very high. Therefore, according to Example 5, it was shown that a low concentration of ATP can be quantitatively measured as a harmful factor by a PLS regression analysis from a fluorescent fingerprint of a stainless steel plate coated with an ATP aqueous solution.

[Example 6]
In Example 6, as in Example 5, a stainless steel plate coated with each ATP aqueous solution was used as the measurement object. Specifically, first, fluorescent fingerprint information of a stainless steel plate coated with each ATP solution was acquired by processing of the fluorescent fingerprint information acquisition unit 24-1. Then, by the processing of the hazard factor quantification unit 24-2, discriminant analysis is performed on the fluorescent fingerprint information of the stainless steel plate coated with each ATP solution, and the hazard factor obtained in the discriminant analysis (in the sixth embodiment, Based on the score distribution of (ATP), the degree of the harm factor was determined from the measurement object. Details of the sixth embodiment will be described below with reference to FIGS. 21 and 22.

(6-1. Preparation of measurement object)
In Example 6, as in Example 5, a stainless steel plate coated with each ATP solution and dried was prepared. Details are the same as those in the fifth embodiment, and a description thereof will be omitted.

(6-2. Acquisition of fluorescence fingerprint information)
Subsequently, in the same manner as in Example 5, the stainless steel plate coated with 10 μL of each ATP solution was set in the fluorescence fingerprint acquisition apparatus 10 (Hitachi spectrofluorometer F-7000). And the fluorescence fingerprint information of each stainless steel plate was acquired by the process of the fluorescence fingerprint information acquisition part 24-1. The measurement conditions are the same as in Example 5.

Here, the stainless steel plate used as a measurement object (sample) was classified into three discrimination groups based on the following criteria.
Low (safety): ATP contamination amount = 10 ⁰ to 10 ² [RLU], ATP amount = 0 to 10 ⁻¹⁴ [mol / assay]
Mid (Attention required): ATP contamination amount = 10 ² to 10 ³ [RLU], ATP amount = 10 ⁻¹⁴ to 10 ⁻¹³ [mol / assay]
・ High (danger): ATP contamination amount = 10 ³ to 10 ⁵ [RLU], ATP amount = 10 ⁻¹² to 10 ⁻⁹ [mol / assay]

Here, the measured value by the measuring device (Lumitester PD-20) of a nonpatent literature 2 and 3 was employ | adopted for the amount of ATP contamination. [RLU] (Relativistic Light Unit) indicates a unit of the amount of luminescence measured by the measuring device (Lumitester PD-20) described in Non-Patent Documents 2 and 3. In Example 6, according to the ATP wiping inspection described in Non-Patent Documents 2 and 3, the amount of ATP contained in the whole food residue and microorganisms is set as the light emission amount [RLU]. For example, when the RLU value is less than 10 ² , the microorganism detection rate is low, and thus “safe” is set. In addition, when the RLU value is 10 ² to 10 ³ , there is a possibility of microbial contamination, so “Caution” is required. In addition, when the RLU value is more than 10 ^3, there is almost certain microbial contamination, so it was set as “dangerous”.

(6-3. Detection of hazards)
Subsequently, discriminant analysis was performed on the obtained fluorescent fingerprint information of each stainless steel plate. Specifically, a discriminant analysis was performed for the purpose of discriminating into the above three groups using the fluorescence fingerprint information of each stainless steel plate as a parameter by the processing of the hazard factor quantification unit 24-2.

Then, as shown in FIG. 21, by the processing of the hazard factor quantification unit 24-2, the distribution of discrimination scores of the obtained hazard factors (ATP in the sixth embodiment) (in FIG. 21, three discrimination criteria ( Based on each plot showing High (danger), Mid (attention required), and Low (safety)), a hazard factor was detected from the measurement object. Here, FIG. 21 is a diagram showing a distribution of discriminant scores of ATP applied on stainless steel by discriminant analysis.

  As shown in FIG. 21, the left side portion of the figure consists of a plot group consisting of a cross mark (discrimination criterion: High), and the center portion of the figure consists of a round mark plot (discrimination criterion: Low). It was shown that three plot groups, which are a plot group and a plot group consisting of a triangle mark plot (discrimination criterion: Mid), are clearly divided and distributed on the right side of the figure. Further, as shown in FIG. 22, the prediction result by the discriminant analysis shown in FIG. 21 (that is, a plot group of 14 crosses determined to be high (dangerous) and a low (safe) 6 The prediction results of the six circle-marked plot groups and the six triangle-marked plot groups determined to be Mid (attention required) are the true values defined in advance (High (danger): 14), Low (Safety): 6 and Mid (attention required): 6). Here, FIG. 22 is a figure which shows the discrimination | determination result of the ATP contamination degree on a stainless steel plate.

  Therefore, according to the sixth embodiment, it was shown that the degree of the hazard (ATP in the sixth embodiment) can be determined by discriminant analysis from the fluorescent fingerprint of the stainless steel plate coated and dried with each ATP solution. As described above, according to the sixth embodiment, as a result of performing the discriminant analysis on the same fluorescent fingerprint data as that used for the PLS regression analysis in the fourth and fifth embodiments, it is erroneously discriminated by a model using six variables. It was shown that the hazard could be identified without any damage.

[Other embodiments]
Although the embodiments of the present invention have been described so far, the present invention is not limited to the above-described embodiments, but can be applied to various different embodiments within the scope of the technical idea described in the claims. It may be implemented.

  In addition, among the processes described in the embodiment, all or part of the processes described as being automatically performed can be performed manually, or the processes described as being performed manually can be performed. All or a part can be automatically performed by a known method.

  In addition, unless otherwise specified, the processing procedures, control procedures, specific names, information including registration data for each processing, parameters such as search conditions, screen examples, and database configurations shown in the above documents and drawings Can be changed arbitrarily.

In addition, regarding the hazard factor quantification apparatus 20, the illustrated components are functionally conceptual, and need not be physically configured as illustrated.

For example, all or some of the processing functions provided in each device of the hazard factor quantification apparatus 20, particularly the processing functions performed by the calculation processing unit 24, are performed by a CPU (Central Processing Unit) and the CPU. It may be realized by a program to be interpreted and executed as hardware by wired logic. The program is recorded on a recording medium to be described later, and is mechanically read by the hazard factor determination device 20 as necessary. That is, a memory 21 such as a ROM or an HD stores a computer program for giving instructions to the CPU in cooperation with an OS (Operating System) and performing various processes. This computer program is executed by being loaded into the RAM, and constitutes a control unit in cooperation with the CPU.

Further, this computer program may be stored in an application program server connected to the hazard factor quantification apparatus 20 via an arbitrary network, and may be downloaded in whole or in part as necessary. It is.

  In addition, the program according to the present invention may be stored in a computer-readable recording medium, and may be configured as a program product. Here, the “recording medium” is any memory card, USB memory, SD card, flexible disk, magneto-optical disk, ROM, EPROM, EEPROM, CD-ROM, MO, DVD, Blu-ray Disc, etc. Of “portable physical media”.

  The “program” is a data processing method described in an arbitrary language or description method, and may be in any format such as source code or binary code. The “program” is not necessarily limited to a single configuration, but is distributed in the form of a plurality of modules and libraries, or in cooperation with a separate program represented by an OS (Operating System). Including those that achieve the function. Note that a well-known configuration and procedure can be used for a specific configuration for reading a recording medium, a reading procedure, an installation procedure after reading, and the like in each device described in the embodiment.

  Various databases and the like stored in the memory 21 are storage means such as a memory device such as a RAM and a ROM, a fixed disk device such as a hard disk, a flexible disk and an optical disk, and various programs and programs used for providing websites. Stores tables, databases, files for web pages, etc.

Further, the hazard factor quantification apparatus 20 may be configured as an information processing apparatus such as a known personal computer or workstation, or may be configured by connecting an arbitrary peripheral device to the information processing apparatus. Further, the hazard factor quantification apparatus 20 may be realized by installing software (including programs, data, etc.) that realizes the method of the present invention in the information processing apparatus.

  Furthermore, the specific form of distribution / integration of the devices is not limited to that shown in the figure, and all or a part of them may be functional or physical in arbitrary units according to various additions or according to functional loads. Can be distributed and integrated. That is, the above-described embodiments may be arbitrarily combined and may be selectively implemented.

As described in detail above, according to the present invention, by measuring and analyzing a fluorescent fingerprint, a complicated pretreatment is unnecessary, and a hazard factor quantification method capable of quickly and easily detecting a hazard factor, Since the hazard factor determination device and the program can be provided, it is extremely useful in various fields such as the food field.

  As described above, the present invention can accurately measure a low-concentration hazard factor in spite of non-destructive and non-contact measurement, so that inspection in a farm, storage, food processing site, manufacturing site, etc. It can be widely used in various fields such as medical research (eg, hospitals, operating room equipment, etc.) that require a sanitary environment, and basic research in terms of ATP quantification. It is a highly versatile technique. In addition, the present invention can be applied to fluorescent fingerprint imaging by reducing the measurement time by narrowing down the measurement conditions. Furthermore, since the present invention facilitates a wide range of inspections at factories and the like, it can also be applied to 100% inspection and real-time monitoring of the sanitary condition of food surfaces.

10 Fluorescent fingerprint acquisition device 11 Spectral illumination device
110 Light source
112 Spectroscopic device 12 Spectroscopic detection device
122 Spectrometer
124 Fingerprint Detection Device 13 Measurement Object 20 Hazard Factor Determination Device 21 Memory 22 Keyboard / Mouse 23 Control Unit 24 Calculation Processing Unit
24-1 Fluorescent fingerprint information acquisition unit
24-2 Hazard Factor Quantification Unit 30 Display

Claims (22)

A hazard factor quantification method for quantifying a hazard factor from a measurement object,
A homogenization step of homogenizing the measurement object into fine pulverization using a pulverizer when the measurement object is powder;
Fluorescence fingerprint information of the measurement object is measured by measuring the fluorescence intensity of the measurement object while gradually changing the excitation wavelength and the observation fluorescence wavelength in a predetermined excitation wavelength range and a predetermined fluorescence wavelength range. Fluorescent fingerprint information acquisition step for acquiring, and
A hazard factor quantification step of performing multivariate analysis on the fluorescence fingerprint information acquired in the fluorescence fingerprint information acquisition step, and quantifying the hazard factor from the measurement object based on a result of the multivariate analysis,
Including
In the fluorescent fingerprint information acquisition step,
Obtain fluorescent fingerprint information of the measurement object that has been homogenized into the finely pulverized in the homogenizing step,
Furthermore, noise information other than the target fluorescent fingerprint is removed from the acquired fluorescent fingerprint information, and only the fluorescent fingerprint information representing the characteristics of the hazard included in the measurement object is extracted,
In the hazard factor determination step,
The matrix data obtained from the extracted fluorescent fingerprint information is rearranged and changed to a continuous one-dimensional vector, and the multivariate analysis is performed on the fluorescent fingerprint information changed to the one-dimensional vector. A hazard factor quantification method characterized in that an amount of a hazard factor is estimated, and the hazard factor is quantified from the measurement object on the basis of the estimated risk factor quantification result. In the hazard factor determination step,
The amount of the hazard factor is estimated by performing PLS regression analysis as the multivariate analysis on the fluorescent fingerprint information, and the hazard factor is calculated from the measurement object based on the quantitative result of the estimated hazard factor. Quantifying,
The hazard factor quantification method according to claim 1, wherein: In the hazard factor determination step,
The principal component analysis is performed as the multivariate analysis on the fluorescent fingerprint information, and the hazard factor is quantified from the measurement object based on the distribution of the principal component score of the hazard factor obtained by the principal component analysis. about,
The hazard factor quantification method according to claim 1, wherein: In the hazard factor determination step,
Performing discriminant analysis as the multivariate analysis on the fluorescent fingerprint information, and quantifying the hazard from the measurement object based on the distribution of the discriminant score of the hazard obtained by the discriminant analysis,
The hazard factor quantification method according to claim 1, wherein: The hazard is
Be mold poison and / or ATP,
The hazard factor quantification method according to any one of claims 1 to 4, characterized in that: The mold poison is
At least one of deoxynivalenol, nivalenol and zearalenone,
The hazard factor quantification method according to claim 5, wherein: The measurement object is
Be at least one of wheat, corn, food residues, and microorganisms;
The hazard factor quantification method according to any one of claims 1 to 6, characterized in that: A spectroscopic illumination device that irradiates a measurement object with a predetermined excitation wavelength, a spectroscopic detection device that measures the measurement object with a predetermined fluorescence wavelength, and finely pulverizes the measurement object when the measurement object is powder A hazard factor quantification device for quantifying a hazard factor from the measurement object, comprising a crushing device for uniformizing
Fluorescence fingerprint information of the measurement object is measured by measuring the fluorescence intensity of the measurement object while gradually changing the excitation wavelength and the observation fluorescence wavelength in a predetermined excitation wavelength range and a predetermined fluorescence wavelength range. A fluorescence fingerprint information acquisition unit for acquiring
A hazard factor quantification unit that performs multivariate analysis on the fluorescence fingerprint information acquired by the fluorescence fingerprint information acquisition unit, and quantifies the hazard factor from the measurement object based on a result of the multivariate analysis,
With
The fluorescent fingerprint information acquisition unit
Obtain fluorescent fingerprint information of the measurement object that has been homogenized into the fine pulverization by the pulverizer,
Furthermore, noise information other than the target fluorescent fingerprint is removed from the acquired fluorescent fingerprint information, and only the fluorescent fingerprint information representing the characteristics of the hazard included in the measurement object is extracted,
The hazard factor quantitative section
The matrix data obtained from the extracted fluorescent fingerprint information is rearranged and changed to a continuous one-dimensional vector, and the multivariate analysis is performed on the fluorescent fingerprint information changed to the one-dimensional vector. A hazard factor quantification apparatus characterized by estimating an amount of a hazard factor and quantifying the hazard factor from the measurement object based on the estimated quantitative result of the hazard factor. The hazard factor quantitative section
The amount of the hazard factor is estimated by performing PLS regression analysis as the multivariate analysis on the fluorescent fingerprint information, and the hazard factor is calculated from the measurement object based on the quantitative result of the estimated hazard factor. Quantifying,
The hazard factor quantification device according to claim 8 , characterized in that: The hazard factor quantitative section
The principal component analysis is performed as the multivariate analysis on the fluorescent fingerprint information, and the hazard factor is quantified from the measurement object based on the distribution of the principal component score of the hazard factor obtained by the principal component analysis. about,
The hazard factor quantification device according to claim 8 , characterized in that: The hazard factor quantitative section
Performing discriminant analysis as the multivariate analysis on the fluorescent fingerprint information, and quantifying the hazard from the measurement object based on the distribution of the discriminant score of the hazard obtained by the discriminant analysis,
The hazard factor quantification device according to claim 8 , characterized in that: The hazard is
Be mold poison and / or ATP,
The hazard factor quantification device according to any one of claims 8 to 11 , characterized by: The mold poison is
At least one of deoxynivalenol, nivalenol and zearalenone,
The harm factor quantification device according to claim 12 , characterized in that: The measurement object is
Be at least one of wheat, corn, food residues, and microorganisms;
The harm factor quantification device according to any one of claims 8 to 13 , characterized by: A spectroscopic illumination device that irradiates a measurement object with a predetermined excitation wavelength, a spectroscopic detection device that measures the measurement object with a predetermined fluorescence wavelength, and finely pulverizes the measurement object when the measurement object is powder a milling device to uniform, with a, a program to be executed by the hazard quantification device for quantifying the hazard from the measurement object, the
In the hazard factor quantification device,
A homogenization step of homogenizing the measurement object into fine pulverization using the pulverizer;
Fluorescence fingerprint information of the measurement object is measured by measuring the fluorescence intensity of the measurement object while gradually changing the excitation wavelength and the observation fluorescence wavelength in a predetermined excitation wavelength range and a predetermined fluorescence wavelength range. Fluorescent fingerprint information acquisition step for acquiring, and
A hazard factor quantification step of performing multivariate analysis on the fluorescence fingerprint information acquired in the fluorescence fingerprint information acquisition step, and quantifying the hazard factor from the measurement object based on a result of the multivariate analysis,
And execute
In the fluorescent fingerprint information acquisition step,
Obtain fluorescent fingerprint information of the measurement object that has been homogenized into the finely pulverized in the homogenizing step,
Furthermore, noise information other than the target fluorescent fingerprint is removed from the acquired fluorescent fingerprint information, and only the fluorescent fingerprint information representing the characteristics of the hazard included in the measurement object is extracted,
In the hazard factor determination step,
The matrix data obtained from the extracted fluorescent fingerprint information is rearranged and changed to a continuous one-dimensional vector, and the multivariate analysis is performed on the fluorescent fingerprint information changed to the one-dimensional vector. A program characterized by estimating the amount of a hazard factor and quantifying the hazard factor from the measurement object based on the estimated quantitative result of the hazard factor. In the hazard factor determination step,
The amount of the hazard factor is estimated by performing PLS regression analysis as the multivariate analysis on the fluorescent fingerprint information, and the hazard factor is calculated from the measurement object based on the quantitative result of the estimated hazard factor. Quantifying,
The program according to claim 15 , wherein: In the hazard factor determination step,
The principal component analysis is performed as the multivariate analysis on the fluorescent fingerprint information, and the hazard factor is quantified from the measurement object based on the distribution of the principal component score of the hazard factor obtained by the principal component analysis. about,
The program according to claim 15 , wherein: In the hazard factor determination step,
Performing discriminant analysis as the multivariate analysis on the fluorescent fingerprint information, and quantifying the hazard from the measurement object based on the distribution of the discriminant score of the hazard obtained by the discriminant analysis,
The program according to claim 15 , wherein: The hazard is
Be mold poison and / or ATP,
The program according to any one of claims 15 to 18 , characterized by: The mold poison is
At least one of deoxynivalenol, nivalenol and zearalenone,
The program according to claim 19 , wherein: The measurement object is
Be at least one of wheat, corn, food residues, and microorganisms;
The program according to any one of claims 15 to 20 , characterized in that: When the measurement object is a powder, a homogenization step for homogenizing the measurement object into fine pulverization,
Is executed further,
In the fluorescent fingerprint information acquisition step,
Obtaining fluorescent fingerprint information of the measurement object homogenized into the finely pulverized in the homogenizing step,
The program according to any one of claims 15 to 21 , characterized in that:

Images