JP7631265B2 - Skin index estimation method - Google Patents

Description

The present invention relates to a technology for estimating skin indices using infrared absorption spectrum information.

The following Patent Document 1 discloses a method for measuring the amount of natural moisturizing factor (NMF) in the stratum corneum by observing a signal intensity A and a protein-derived signal intensity B in the wave number region of 1610 to 1570 cm ^-1 from an infrared absorption spectrum of the stratum corneum of the skin, and measuring the amount of natural moisturizing factor (NMF) in the stratum corneum from the ratio of signal intensity A' of the antisymmetric stretching vibration of carboxylate anion extracted from signal intensity A to signal intensity B.
Patent Document 2 listed below discloses a method for evaluating the properties of the stratum corneum of human skin for purposes other than medical treatment, in which an infrared absorption spectrum is measured by infrared spectroscopy, an amide I absorption band peak derived from the secondary structure of the protein in the infrared absorption spectrum is divided into a peak of superimposed α-helix and random coil and a peak of β-sheet, the areas of the peak of superimposed α-helix and random coil and the peak area of β-sheet are calculated, and a peak area ratio β/α defined as [peak area of β-sheet]/[peak area of superimposed α-helix and random coil] is calculated, and the calculated ratio β/α is used as an index to evaluate the properties of the stratum corneum of skin for purposes other than medical treatment.
The following Patent Documents 3 to 5 disclose a method for easily and stably evaluating the properties of the stratum corneum using infrared absorption spectra obtained from the stratum corneum obtained by tape peeling.

JP 2009-210567 A JP 2013-44525 A JP 2018-105783 A JP 2020-193948 A JP 2021-181987 A

The various methods described above evaluate the properties of the stratum corneum based on the signal intensity, peak shape, etc. of a peak in a specific wavenumber region in the infrared absorption spectrum of the stratum corneum.
The present invention provides a technique that makes it possible to estimate index values of stratum corneum characteristics or skin surface characteristics with higher accuracy by using infrared absorption spectrum information of the stratum corneum.

According to the present invention, a skin index estimation method can be provided in which one or more processors execute an acquisition step of acquiring infrared absorption spectrum information regarding the stratum corneum peeled from the skin of a subject, and an estimation step of estimating a target index value of the subject by inputting the acquired infrared absorption spectrum information into an estimation model obtained using multiple data sets of infrared absorption spectrum information regarding the stratum corneum peeled from the skin of multiple sample providers and known target index values, wherein the infrared absorption spectrum information includes multiple intensity values in a predetermined wavenumber range, and the target index value is an index value regarding the stratum corneum characteristics or skin surface characteristics.
Also, a skin index estimation device including at least one processor and a memory and capable of executing the above-mentioned skin index estimation method may be provided.

The present invention provides a technology that can estimate index values of stratum corneum characteristics or skin surface characteristics with higher accuracy using infrared absorption spectrum information of the stratum corneum.

FIG. 1 is a diagram conceptually illustrating an example of the hardware configuration of an information processing device capable of executing a skin index estimation method according to an embodiment of the present invention (the present method). 1 is a flowchart of a skin index estimation method (this method) according to the present embodiment. 1 is a graph showing an infrared absorption spectrum for the stratum corneum peeled from skin. 1 is a table showing the estimation accuracy for each target index value related to the stratum corneum delamination rate and the stratum corneum cell shape in the estimation model of Example 1. 1 is a table showing the estimation accuracy for each target index value related to stratum corneum intercellular lipids in the estimation model of Example 1. 1 is a table showing the estimation accuracy for each target index value regarding the moisture content of the stratum corneum in the estimation model of Example 1. 1 is a table showing the estimation accuracy for each target index value regarding skin surface characteristics in the estimation model of Example 1. 8 is a graph showing the weight (regression coefficient) of each wave number component (each explanatory variable) to the absorbance in the estimation model (estimation formula) with stratum corneum characteristics as the target index value. Fig. 8(a) is a graph related to the estimation model (estimation formula) with stratum corneum moisture content measured by Skicon (manufactured by Yayoi Co., Ltd.) as the target index value, Fig. 8(b) is a graph related to the estimation model (estimation formula) with transepidermal water loss (TEWL) as the target index value, Fig. 8(c) is a graph related to the estimation model (estimation formula) with the ratio (NP/NS) between the amount of ceramide NP and the amount of ceramide NS as the target index value, and Fig. 8(d) is a graph related to the estimation model (estimation formula) with stratum corneum layer peeling rate as the target index value.

Hereinafter, preferred embodiments of the present invention (hereinafter, referred to as the present embodiment) will be described. The embodiments described below are merely examples, and the present invention is not limited to the configurations of the following embodiments.

The skin index estimation method according to this embodiment (hereinafter referred to as this method) is executed by one or more processors included in one or more information processing devices.
FIG. 1 is a diagram conceptually showing an example of the hardware configuration of an information processing device 10 capable of executing the present method.
The information processing device 10 is a so-called computer, and includes a CPU 11, a memory 12, an input/output interface (I/F) 13, a communication unit 14, etc. The information processing device 10 may be a stationary PC (Personal Computer), a portable terminal such as a portable PC, a smartphone, or a tablet, or may be a dedicated computer.

The CPU 11 is a so-called processor, and may include not only a general central processing unit (CPU) but also an application specific integrated circuit (ASIC), a digital signal processor (DSP), a graphics processing unit (GPU), etc. The memory 12 is a random access memory (RAM), a read only memory (ROM), or an auxiliary storage device (such as a hard disk).
The input/output I/F 13 can be connected to user interface devices such as a display device 15 and an input device 16. The display device 15 is a device that displays a screen corresponding to drawing data processed by the CPU 11 or the like, such as an LCD (Liquid Crystal Display) or a CRT (Cathode Ray Tube) display. The input device 16 is a device that accepts input of user operations, such as a keyboard, a mouse, etc. The display device 15 and the input device 16 may be integrated and realized as a touch panel.
The communication unit 14 communicates with other computers via a communication network and exchanges signals with other devices such as a printer, etc. A portable recording medium, etc. can also be connected to the communication unit 14.

The hardware configuration of the information processing device 10 is not limited to the example of FIG. 1. The information processing device 10 may include other hardware elements not shown. Furthermore, the number of each hardware element is not limited to the example of FIG. 1. For example, the information processing device 10 may have multiple CPUs 11. Furthermore, the information processing device 10 may be realized by multiple computers consisting of multiple housings.

The information processing device 10 can execute the present method by the CPU 11 executing a computer program stored in the memory 12. This computer program is installed from a portable recording medium such as a CD (Compact Disc) or memory card, or from another computer on a network, via the input/output I/F 13 or the communication unit 14, and is stored in the memory 12.

The information processing device 10 (CPU 11) can use the estimation model and can also generate the estimation model. However, the estimation model may be generated by another computer. The estimation model may be stored in the memory 12 in the information processing device 10, or may be stored in the memory of another computer that the information processing device 10 can access by communication.
In this way, the information processing device 10 can be described as a skin index estimation device that is capable of using the estimation model and executing the present method. Details of the estimation model will be described later.

Here, this method will be described with reference to FIG.
FIG. 2 is a flowchart of the skin index estimation method (the present method) according to this embodiment.
As shown in Fig. 2, this method includes at least an acquisition step (S11) of acquiring infrared absorption spectrum information on the stratum corneum peeled from the skin of the subject, and an estimation step (S13) of estimating the target index value of the subject by inputting (S12) the infrared absorption spectrum information acquired in the acquisition step (S11) into the above-mentioned estimation model. In the example of Fig. 2, this method further includes a step (S14) of outputting the target index value of the subject acquired in the estimation step (S13).
This method is executed by an information processing device 10 (CPU 11) as exemplified in Fig. 1. Hereinafter, each step of this method will be described in detail with the CPU 11 acting as the executing entity.

The infrared absorption spectrum information acquired in the acquisition step (S11) is information on the infrared absorption spectrum (also called the infrared absorption spectrum, infrared transmission spectrum, etc.) obtained by infrared spectroscopy measurement of a specimen that includes at least the stratum corneum peeled from the skin, and includes multiple intensity values in a specified wavenumber range of the infrared absorption spectrum.

The "stratum corneum peeled off from the skin" includes not only stratum corneum collected or peeled off from the skin using hands, adhesive tape, some kind of tool, etc., but also stratum corneum that has naturally peeled off. The infrared absorption spectrum information may represent only the infrared absorption spectrum obtained for such a stratum corneum, or may represent the infrared absorption spectrum of the stratum corneum plus other components such as tape, moisture, and dust.
The "intensity value" is a value indicating the intensity of each wave number in an infrared absorption spectrum, such as absorbance, transmittance, or reflectance.

3 is a graph showing an example of an infrared absorption spectrum for the stratum corneum peeled from the skin, in which the horizontal axis indicates wave number and the vertical axis indicates absorbance (intensity value).
The infrared absorption spectrum information acquired in the acquisition step (S11) includes a plurality of intensity values extracted from a group of intensity values in a predetermined wave number region of the infrared absorption spectrum as exemplified in FIG.
The "predetermined wavenumber region" of the infrared absorption spectrum is a predetermined wavenumber region, and may be a wavenumber range that is the measurement limit according to the specifications of an apparatus for performing infrared spectrometry, or may be one or more partial wavenumber regions within the wavenumber range of the measurement limit. For example, as shown in the example of Figure 3, the predetermined wavenumber region may be a wavenumber range of 600 cm ^-1 or more and 4000 cm ^-1 or less that is the measurement limit.

However, it is preferable that the infrared absorption spectrum information includes at least a plurality of intensity values in a first wavenumber region of 600 cm ⁻¹ or more and 1800 cm ⁻¹ or less and a plurality of intensity values in a second wavenumber region of 2800 ^{cm −1} or more and 3700 cm −1 or less. As will be described later as an example, the contribution rate (regression coefficient) of the intensity values in the first wavenumber region and the second wavenumber region to the target index value is higher than the intensity values in the other wavenumber regions, and the estimation accuracy of the estimation model can be improved.

The infrared absorption spectrum information preferably includes at least a plurality of intensity values in non-peak regions of a first wavenumber region of 600 cm ^-1 to 1800 cm ^-1 and a second wavenumber region of 2800 cm ^-1 to 3700 cm ^-1 .
The term "non-peak region" as used herein refers to a wavenumber region that is not a peak in an infrared absorption spectrum graph such as that shown in Fig. 3, and specifically refers to a wavenumber region excluding the peak region in an infrared absorption spectrum with a predetermined wavenumber resolution (e.g., 0.5 cm ^-1 ). The term "peak region" refers to a wavenumber region that exhibits a convex shape in the infrared absorption spectrum, and refers to, for example, a predetermined wavenumber range (e.g., 100 cm ^-1 ) centered on the wavenumber that exhibits the peak intensity within that wavenumber region, or a wavenumber range centered on the wavenumber that exhibits the peak intensity, from 80% to 50% of the peak intensity.
In the example of FIG. 3, the non-peak region of the first wavenumber region corresponds to wavenumber ranges around 700 to 800 cm ⁻¹ , around 1000 to 1100 cm ⁻¹ , around 1400 cm ⁻¹ , etc., and the non-peak region of the second wavenumber region corresponds to wavenumber ranges around 3000 to 3100 cm ⁻¹ , around 3400 to 3700 cm ⁻¹ , etc.
The present inventors have found that the estimation accuracy of the estimation model can be improved by using not only the peak intensity value but also a plurality of intensity values in non-peak regions as the infrared absorption spectrum information.

Moreover, the intensity value included in the infrared absorption spectrum information may be extracted regardless of whether it is a peak intensity value of the infrared absorption spectrum. For example, the infrared absorption spectrum information may include a plurality of intensity values sampled at a fixed wavenumber interval (first wavenumber interval) from a first wavenumber region of 600 cm ^-1 to 1800 cm ^-1 and at a fixed wavenumber interval (second wavenumber interval) from a second wavenumber region of 2800 cm ^-1 to 3700 cm ^-1 , regardless of whether it is a peak intensity value. In this case, the first wavenumber interval and the second wavenumber interval may be the same interval or may be different intervals. Moreover, the infrared absorption spectrum information may include a plurality of intensity values sampled at a fixed wavenumber interval from a predetermined wavenumber region (for example, 600 cm ^-1 to 4000 cm ^-1 ).
It has been demonstrated in the examples described below that the estimation accuracy of the estimation model can be improved by using multiple intensity values extracted in this manner, regardless of whether they are peak intensity values of the infrared absorption spectrum or not.

In addition, in order to improve the estimation accuracy of the estimation model, the number of intensity values included in the infrared absorption spectrum information is preferably 10 or more, more preferably 25 or more, and even more preferably 50 or more, as described below as an example.

In the acquisition step (S11), the CPU 11 may acquire the infrared absorption spectrum information from an infrared spectrometer or another computer via communication or a portable recording medium, or may acquire the infrared absorption spectrum information by processing the information acquired from the measurement device or another computer.

As described above, in this embodiment, various methods can be used to acquire infrared absorption spectrum information. For example, the stratum corneum may be collected from the skin surface using an adhesive tape having a silicone adhesive layer or an adhesive layer of another component, and the infrared absorption spectrum information may be extracted from the results of infrared spectrometry of the adhesive tape to which at least the stratum corneum is attached. In this case, the CPU 11 may acquire information on the results of the infrared spectrometry from an infrared spectrometry device or another computer, and acquire the required infrared absorption spectrum information from the result information.
The infrared absorption spectrum information may be obtained using the methods disclosed in the above-mentioned Patent Documents 3 to 5. Specifically, a method may be performed in which the stratum corneum is collected from the skin surface using an adhesive tape T1, the collected stratum corneum is transferred from the adhesive tape T1 to a tape T2 having a silicone adhesive layer, the adhesive surface of the tape T2 is subjected to ATR-IR measurement to obtain a difference spectrum between an infrared absorption spectrum A obtained by ATR-IR measurement of an adhesive surface of an unused tape T3 having a silicone adhesive layer, and the infrared absorption spectrum information is extracted from the difference spectrum. In this method, the difference spectrum is obtained by multiplying the infrared absorption spectrum A or the infrared absorption spectrum B by a constant and then subtracting it so that the peak intensities of the absorption band around 1260 cm ⁻¹ or the absorption band around 1000 to 1100 cm ⁻¹ , which are the main absorption bands derived from the silicone adhesive layer, or the peak intensities derived from both of them coincide. Note that in the Examples section, this method is referred to as the disclosed method.

The estimation model used in step (S12) is a model obtained using multiple data sets related to multiple sample donors.
Each data set is a combination of infrared absorption spectrum information for the stratum corneum peeled from the skin of a sample donor and a known target index value for that sample donor.
The infrared absorption spectrum information is used as the value of the explanatory variable that is the input of the estimation model. The infrared absorption spectrum information is as described in step (S11). However, the infrared absorption spectrum information used to construct the estimation model is obtained for the stratum corneum of the sample provider. However, the subject may be included in the sample provider.

The objective index value is an index value related to the stratum corneum characteristics or skin surface characteristics, and is used as the value of the objective variable that is the output of the estimation model. The known objective index values contained in the above-mentioned dataset are obtained in advance by performing evaluation, measurement, etc. on each sample provider.

The index value relating to the stratum corneum characteristics is an index value that indicates the characteristics of the stratum corneum, for example, an index value relating to the delamination rate of the stratum corneum, an index value relating to the cell shape of the stratum corneum, an index value relating to the intercellular lipids of the stratum corneum, an index value relating to the moisture content of the stratum corneum, etc.
The index value relating to the skin surface characteristics is an index value obtained by measuring a replica that reproduces the skin surface shape or the skin surface, and includes an index value indicating the morphology of the skin surface. Examples of the index value indicating the morphology of the skin surface include index values of scale, smoothness and wrinkle state of the skin obtained by measuring the surface shape of the skin using a microscope or the like, roughness of the skin surface obtained by measuring the replica, wrinkle depth, number of wrinkles, wrinkle volume, wrinkle area and wrinkle area ratio.
In addition, an index value relating to the moisture content of the stratum corneum, such as the transepidermal water loss (TEWL) described below, is also an index value obtained by measuring the skin surface, and therefore can also be considered as an index value relating to skin surface characteristics. However, in this specification, the index value relating to the moisture content of the stratum corneum is considered to be an index value relating to the stratum corneum characteristics, since it is one of the values expressing the physical properties of the "stratum corneum."

The index values relating to the delamination rate of the stratum corneum and the cell shape of the stratum corneum can be obtained, for example, by analyzing an image of the stratum corneum captured using tape or the like.
The stratum corneum layer peeling rate can be obtained by determining the area of the peeled part and the area of the layer peeling part by binarizing the stratum corneum image, and dividing the area of the layer peeling part by the area of the peeled part.
The index value relating to the shape of the stratum corneum cells can be obtained by measuring the shape index values of multiple stratum corneum cells per person and performing statistical processing such as averaging the multiple measured shape index values. For example, multiple stratum corneum cell images are taken per person, multiple stratum corneum cells are identified per stratum corneum cell image, and the shape index value of each identified stratum corneum cell is measured and statistical processing is performed using the number of stratum corneum cell images, thereby obtaining an index value relating to the shape of the stratum corneum cells of a subject. Examples of index values relating to the shape of the stratum corneum cells include the area of the stratum corneum cells, the length of the stratum corneum cells, and the circularity of the stratum corneum cells.

The index value for intercellular lipids is the amount of a component of the intercellular lipids or the ratio of the amount of each component, for example, the total amount of intercellular lipids, the amount of ceramide, cholesterol or free fatty acid in the intercellular lipids, or the amount of each ceramide molecular species in the intercellular lipids or the ratio of the amount of each ceramide molecular species.
The index value for the intercellular lipids can be measured, for example, by an analytical method combining liquid chromatography (LC) and mass spectrometry (MS). Examples of the index value for the intercellular lipids include the total amount of intercellular lipids, the amount of ceramide, the amount of cholesterol, the amount of free fatty acid, the amount of each ceramide molecular species (NDS, NS, NH, NP, ADS, AS, AH, AP, EOS, EOH, EOP, etc.), and the ratio of the amount of each ceramide molecular species (NP/NS, etc.).

The index value for the moisture content of the stratum corneum is an index value indicating the amount of moisture contained in the stratum corneum (also referred to as stratum corneum moisture content). However, the index value here is not limited to a value strictly indicating the amount of moisture contained only in the stratum corneum, and may be an index value that mainly includes the amount of moisture contained in the stratum corneum, as well as the amount of moisture contained in tissues below the stratum corneum in the epidermis.
The moisture content of the stratum corneum can be measured, for example, by a conductance measurement method, a capacitance measurement method, or the like.
In addition, in this specification, the transepidermal water loss (TEWL), which is known as an evaluation value indicating the barrier function of the skin and means the amount of water that evaporates through the stratum corneum, is also positioned as an index value related to the water content of the stratum corneum. The evaluation value here includes an index value indicating not only the barrier function from the inside of the skin to the outside, but also the barrier function from the outside of the skin to the inside. However, as described above, the transepidermal water loss (TEWL) can also be positioned as an index value related to the skin surface characteristics.

The estimation model is constructed using a set of data of such known target index values and infrared absorption spectrum information for multiple sample providers. In this embodiment, an estimation model constructed using multivariate analysis is exemplified. That is, the estimation model in this embodiment is obtained by performing multivariate analysis on multiple data sets related to multiple sample providers.
Well-known multivariate analysis techniques include, for example, multiple regression analysis, principal component analysis, partial least squares regression (PLS), and principal component regression analysis (PCR). In this embodiment, the specific technique of multivariate analysis used is not limited as long as it is possible to construct an estimation model in which the infrared absorption spectrum information is an explanatory variable and the objective index value is an objective variable. The estimation model in this embodiment can also be expressed as an estimation formula.

The estimation model may be composed of a convolutional neural network including a deep neural network, and may be realized by a combination of a computer program and parameters, a combination of multiple functions and parameters, etc. In addition, when the estimation model is constructed with a neural network, and the input layer, intermediate layer, and output layer are regarded as a single neural network unit, it may refer to one neural network or a combination of multiple neural networks. The model may be composed of a combination of multiple estimation formulas, or may be composed of one estimation formula. The estimation model may be composed of a combination of a neural network and another model such as a regression formula or a discriminant formula, and may be configured to input a value obtained as an output of the neural network into the other model and output the value obtained.

In this embodiment, the estimation model may be a single estimation formula in which the infrared absorption spectrum information including a plurality of intensity values is used as an explanatory variable and a single objective index value is used as an objective variable, or may be a set of two or more estimation formulas in which different objective index values are used as objective variables. In the latter case, two or more objective index values are obtained in step (S13).
Furthermore, when the estimation model is constructed as an AI model consisting of a neural network, infrared absorption spectrum information including multiple intensity values is input, and one or multiple target index values are output.

In step (S14), CPU 11 outputs the target index value acquired in step (S13). CPU 11 can display the target index value on display device 15, can output the target index value to a printer device connected via input/output I/F 13 or communication unit 14, and can record or transmit the target index value to a portable recording medium or another computer via input/output I/F 13 or communication unit 14. In this way, the output destination and output method of the target index value are not limited in any way.
Furthermore, the CPU 11 may output the target index value acquired in step S13 as a numerical value as it is, or may output the acquired target index value after normalizing it by digit alignment or the like. The present embodiment does not limit the output format of the target index value.

As described above, in this embodiment, infrared absorption spectrum information relating to the stratum corneum peeled off from the skin of the subject, the infrared absorption spectrum information including a plurality of intensity values in a relatively wide predetermined wavenumber range including at least a first wavenumber range of 600 cm ^-1 or more and 1800 cm ^-1 or less and a second wavenumber range of 2800 ^{cm -1} or more and 3700 cm -1 or less, is input into an estimation model, and the target index value of the subject is estimated.
The target index value estimated from the infrared absorption spectrum information on the stratum corneum is an index value related to the stratum corneum characteristics or the skin surface characteristics.

In the above-mentioned conventional methods, attention was focused on peaks in a specific wavenumber region in the infrared absorption spectrum of the stratum corneum when evaluating the properties of the stratum corneum. However, the present inventor has newly discovered that it is possible to estimate the characteristics of the stratum corneum or the skin surface with higher accuracy by turning his attention to a relatively wide, specific wavenumber region in the infrared absorption spectrum of the stratum corneum that includes at least the above-mentioned first and second wavenumber region, and using multiple intensity values extracted from such a wide wavenumber region (intensity values in non-peak regions, multiple intensity values sampled at regular intervals unrelated to peaks, etc.).
As a result, according to this embodiment, by using multiple intensity values in the specified wave number region in the infrared absorption spectrum of the stratum corneum, it is possible to estimate an index of the skin surface including the stratum corneum with high accuracy.
Furthermore, such skin surface indexes can be easily estimated by simply obtaining infrared absorption spectrum information on the stratum corneum peeled from the skin of the subject, without performing individual processes for obtaining indices of the skin surface including the stratum corneum, such as measuring the skin surface using various measuring devices or taking images, etc. Furthermore, since such infrared absorption spectrum information can be obtained from the stratum corneum in a state where it is peeled from the skin of the subject without contacting the subject, this embodiment enables remote evaluation and is also useful in terms of hygiene.

Some or all of the above-described embodiments and variations may also be specified as follows. However, the above-described embodiments and variations are not limited to the following descriptions.

<1> One or more processors,
An acquisition step of acquiring infrared absorption spectrum information regarding the stratum corneum peeled from the skin of the subject;
an estimation step of estimating the target index value of the subject by inputting the acquired infrared absorption spectrum information into an estimation model obtained by using a plurality of data sets of infrared absorption spectrum information on the stratum corneum peeled from the skin of a plurality of sample donors and known target index values;
Run
the infrared absorption spectrum information includes a plurality of intensity values in a predetermined wave number region;
The target index value is an index value related to a stratum corneum characteristic or a skin surface characteristic.
Skin index estimation method.

<2> The infrared absorption spectrum information includes at least a plurality of intensity values in a first wavenumber region of 600 cm ⁻¹ or more and 1800 cm ⁻¹ or less and a plurality of intensity values in a second wavenumber region of 2800 ^{cm −1} or more and 3700 cm ⁻¹ or less.
The skin index estimation method according to <1>.
<3> The infrared absorption spectrum information includes at least a plurality of intensity values in non-peak regions of the first wavenumber region and the second wavenumber region,
The skin index estimation method according to <2>.
<4> The infrared absorption spectrum information includes a plurality of intensity values sampled at a constant wavenumber interval from the first wavenumber region and at a constant wavenumber interval from the second wavenumber region,
The skin index estimation method according to <2> or <3>.
<5> The target index value is a stratum corneum delamination rate, a stratum corneum cell area, a stratum corneum cell length, or a stratum corneum cell circularity rate,
The skin index estimating method according to any one of <1> to <4>.
<6> The target index value is a total amount of intercellular lipids, an amount of ceramide, an amount of cholesterol, or an amount of free fatty acid, an amount of each ceramide molecular species in intercellular lipids, or a ratio of an amount of each ceramide molecular species in intercellular lipids.
The skin index estimating method according to any one of <1> to <4>.
<7> The target index value is the stratum corneum moisture content or the transepidermal water loss,
The skin index estimating method according to any one of <1> to <4>.
<8> The target index value is an index value of a scale state, a skin smoothness or a wrinkle state obtained by measuring a surface morphology of the skin, or an index value of a skin surface roughness, a wrinkle depth, a wrinkle number, a wrinkle volume, a wrinkle area or a wrinkle area ratio obtained by measuring a replica that reproduces the surface shape of the skin,
The skin index estimating method according to any one of <1> to <4>.

<9> A skin index estimation device including at least one processor and a memory,
A skin index estimation device capable of executing the skin index estimation method according to any one of <1> to <8>.

The following provides several examples to explain the above in more detail. However, the description of each example does not impose any limitations on the above.

In Example 1, 390 women aged from their 20s to their 50s served as sample donors, and a data set of infrared absorption spectrum information and target index values for each sample donor was obtained.
Infrared absorption spectrum information was obtained by the techniques disclosed above.
As the infrared absorption spectrum information, all (6950) absorbances in the wave number region from 600 cm ^-1 to 4000 cm ^-1 (wave number resolution 0.5 cm ^-1 ) output from an infrared spectrometer were used.
Regarding the target index value, multiple types of target index values were obtained from each sample provider using an acquisition method corresponding to each target index value. The types of target index values will be described later.

The collected datasets were divided into 7:3, with the former (70%) being used to build an estimation model and the latter (30%) being used to verify the estimation model.
The estimation model was constructed by applying partial least squares (PLS) to multiple datasets for constructing the estimation model, and the model (estimation equation) that minimized the predicted residual sum of squares (PRESS) in K-fold cross-validation (K = 5) was selected.
The constructed estimation model was then verified using multiple validation data sets. Specifically, the correlation between the target index value (estimated index value) obtained by inputting the infrared absorption spectrum information of the validation data set into the estimation model and the known target index value of the data set was verified.

Figures 4 to 7 are tables showing the estimation accuracy for each objective index value of the estimation model, and the correlation coefficient (R value) and P value indicating the estimation accuracy are shown for each objective index value.

Fig. 4 is a table showing the estimation accuracy for each objective index value related to the stratum corneum delamination rate and the stratum corneum cell shape. In Fig. 4, the objective index values related to the stratum corneum cell shape include the stratum corneum cell area, the stratum corneum cell length, and the stratum corneum cell circularity. The stratum corneum cell length indicates the maximum length of a single stratum corneum cell, and the stratum corneum cell circularity is a value that approaches 1 for a perfect circle, and approaches 1 as the cell approaches a perfect circle.
In this example, four images of the stratum corneum were taken per person, and approximately 10 to 15 stratum corneum cells were identified in each stratum corneum image. The shape index value of each identified stratum corneum cell was measured and the average of the four stratum corneum images was taken to obtain the actual measured values of each target index in Figure 4.
As shown in Figure 4, the P values for the estimation accuracy of the target index values of the stratum corneum delamination rate, stratum corneum cell area, stratum corneum cell length, and stratum corneum cell circularity rate were sufficiently small (at a significance level of 1%), and the correlation coefficients were 0.25 or higher.

Fig. 5 is a table showing the estimation accuracy for each target index value related to stratum corneum intercellular lipids. The target index values shown in Fig. 5 are the amounts of ceramide NDS, ceramide NS, ceramide NH, ceramide NP, ceramide ADS, ceramide AS, ceramide AH, ceramide AP, ceramide EOS, ceramide EOH, and ceramide EOP, the total amount of ceramide, the total amount of free fatty acid, the amount of cholesterol, the amount of cholesterol sulfate, the total amount of stratum corneum intercellular lipids, and the ratio of the amount of ceramide NP to the amount of ceramide NS (NP/NS).
As shown in Figure 5, with regard to the estimation accuracy of the target index value related to intercellular lipids, except for the amount of ceramide NS and ceramide AS, the P value is sufficiently small (at the 1% significance level) and the correlation coefficient is 0.25 or more. With regard to the estimation accuracy of the amount of ceramide AS, the correlation coefficient is 0.201 at the 5% significance level, and with regard to the estimation accuracy of the amount of ceramide NS, the correlation coefficient is 0.180 at the 10% significance level.

Fig. 6 is a table showing the estimation accuracy for each objective index value related to the moisture content of the stratum corneum. The objective index values shown in Fig. 6 are the transepidermal water loss (TEWL), the moisture content of the stratum corneum obtained by the capacitance measurement method, and the moisture content of the stratum corneum obtained by the conductance measurement method.
As shown in FIG. 6 , the actual values of transepidermal water loss (TEWL) were measured using a Tewameter (MPA580, Courage+Khazaka, Germany), the actual values of the items labeled "Corneometer-Capacitance" were measured using a Corneometer (MPA580, Courage+Khazaka, Germany) as a capacitance measurement method, and the actual values of the items labeled "Skicon-Conductance" were measured using Skicon (manufactured by Yayoi) as a conductance measurement method.
As shown in FIG. 6, with regard to the estimation accuracy of the target index value related to the stratum corneum moisture content, the P value was sufficiently small (at the 1% significance level) and the correlation coefficient was 0.45 or more.

Figure 7 is a table showing the estimation accuracy for each objective index value related to skin surface characteristics. The objective index values shown in Figure 7 are index values indicating the scale condition (SESC), skin smoothness (SESM), and wrinkle condition (SEW) obtained by surface morphology measurement of the cheek skin, and the arithmetic mean roughness (Ra), root mean square roughness (Rq), total mean wrinkle depth, ten-point mean roughness (Rz), number of wrinkles, total wrinkle volume, total wrinkle area, and wrinkle area ratio obtained by surface morphology measurement of a replica of the corner of the eye.

The actual measurement values of the items marked as "Visioscan" in Fig. 7 were measured from the skin of the cheek using a Visioscan (Courage+Khazaka, Germany). The actual measurement values of the items marked as "replica" in Fig. 7 were obtained by applying a silicone agent for skin replicas (SILFLO, manufactured by FLEXICO) to the corners of the eyes, hardening it, and then peeling it off to create a concave-convex replica that reproduced the surface shape of the skin, and analyzing the replica with a three-dimensional skin analysis system PRIMOS (Canfield Scientific).
The arithmetic mean roughness (Ra) is the average value obtained by adding up the absolute values of the distances from the mean line to the measurement curve, the root mean square roughness (Rq) is the root mean square at a reference length, the total mean wrinkle depth is the average depth of wrinkles extracted from a three-dimensional shape that has been shape-corrected for the surface shape of the replica, the ten-point mean roughness (Rz) is the sum (distance) between the average of the five highest peaks and the average of the five lowest valleys, the number of wrinkles is the total number of wrinkles extracted from a three-dimensional shape that has been shape-corrected for the surface shape of the replica, the total wrinkle volume is the sum of the volumes of each wrinkle extracted from a three-dimensional shape that has been shape-corrected for the surface shape of the replica, the total wrinkle area is the sum of the areas of each wrinkle extracted from a three-dimensional shape that has been shape-corrected for the surface shape of the replica, and the wrinkle area ratio is the ratio of the total wrinkle area to the area of the region from which wrinkles are extracted.

As shown in Figure 7, the estimation accuracy of the target index value related to skin surface characteristics shows a correlation coefficient of 0.2 or more at the 5% or 1% significance level, except for "Visioscan-SESM", "Replica - total wrinkle area", and "Replica - wrinkle area ratio". The estimation accuracy of "Replica - wrinkle area ratio" shows a correlation coefficient of 0.185 at the 5% significance level, and the estimation accuracy of "Visioscan-SESM" and "Replica - total wrinkle area" shows an estimation accuracy of 0.16 or more at the 10% significance level.

Thus, according to Example 1, it has been demonstrated that index values relating to the stratum corneum ^{characteristics} and skin surface characteristics can be estimated with high accuracy by using a large number (6,950) of absorbance values at a certain interval (wavenumber resolution 0.5 ^{cm -1} ) in the wavenumber region from 600 cm -1 to 4,000 cm -1 in the infrared absorption spectrum of the stratum corneum.

Fig. 8 is a graph showing the weight (regression coefficient) of each wave number component (each explanatory variable) to the absorbance in the estimation model (estimation formula) with the stratum corneum characteristics as the target index value. Fig. 8(a) is a graph related to the estimation model (estimation formula) with the stratum corneum moisture content measured by Skicon (manufactured by Yayoi Co., Ltd.) as the target index value, Fig. 8(b) is a graph related to the estimation model (estimation formula) with the transepidermal water loss (TEWL) as the target index value, Fig. 8(c) is a graph related to the estimation model (estimation formula) with the ratio (NP/NS) between the amount of ceramide NP and the amount of ceramide NS as the target index value, and Fig. 8(d) is a graph related to the estimation model (estimation formula) with the stratum corneum layer peeling rate as the target index value.
In FIG. 8, a wave number component having a large absolute value on the vertical axis of the graph indicates a high contribution rate to the target index value.

According to FIG. 8, since there are multiple peaks in the first wavenumber region of 600 cm ⁻¹ or more and 1800 cm ⁻¹ or less and the second wavenumber region of 2800 ^{cm −1} or more and 3700 cm ⁻¹ or less, it can be seen that the contribution rate of the intensity values in the first wavenumber region and the second wavenumber region to the target index value is higher than the intensity values in the other wavenumber regions.
Therefore, as in this embodiment, by using infrared absorption spectrum information including at least a plurality of intensity values in a first wavenumber region of 600 cm ^{-1 or} more and 1800 cm ^- 1 or less and a plurality of intensity values in a second wavenumber region of 2800 cm-1 or more and 3700 cm ^-1 or less, it has been demonstrated that the estimation accuracy of the target index value by the estimation model can be improved.

In the above-mentioned Example 1, an example was shown in which infrared absorption spectrum information was obtained by the above-mentioned disclosed method, that is, a method in which the stratum corneum is transferred from adhesive tape T1 to tape T2 having a silicone adhesive layer, the adhesive surface of tape T2 is subjected to ATR-IR measurement to obtain an infrared absorption spectrum A, and the adhesive surface of unused tape T3 having a silicone adhesive layer is subjected to ATR-IR measurement to obtain an infrared absorption spectrum B, and the infrared absorption spectrum information is extracted from the difference spectrum.

In Example 2, infrared absorption spectrum information was obtained by a method different from the disclosed method employed in Example 1. That is, in Example 2, infrared absorption spectrum information was obtained by collecting stratum corneum from the skin surface with adhesive tape T1, transferring the collected stratum corneum from the adhesive tape T1 to tape T2 having a silicone adhesive layer, and measuring the adhesive surface of tape T2 with an infrared spectrometer. Therefore, it can be said that the infrared absorption spectrum information in Example 2 includes at least spectral information related to the components of adhesive tape T2, stratum corneum components, and moisture.

In Example 2, an estimation model was constructed in the same manner as in Example 1, and the correlation between the target index value (estimated index value) obtained by inputting the infrared absorption spectrum information of the validation data set (30% of the data set) into the estimation model and the known target index value of that data set was verified. Then, the correlation verified in Example 1 was compared with the correlation verified in Example 2.

As a result, for the estimation model (estimation equation) with the stratum corneum moisture content (Skicon-conductance) as the target index value, the correlation coefficient obtained in Example 1 was 0.491 and the P value was 2.49E-08 (see FIG. 6 ), and the correlation coefficient obtained in Example 2 was 0.520 and the P value was 2.68E-09.
Regarding the estimation model (estimation equation) in which the stratum corneum moisture content (corneometer-capacitance) was used as the target index value, the correlation coefficient obtained in Example 1 was 0.577 and the P value was 2.15E-11 (see FIG. 6 ), and the correlation coefficient obtained in Example 2 was 0.606 and the P value was 1.14E-12.
Regarding the estimation model (estimation equation) with transepidermal water loss (TEWL) as the target index value, the correlation coefficient obtained in Example 1 was 0.460 and the P value was 2.91E-07 (see FIG. 6 ), and the correlation coefficient obtained in Example 2 was 0.530 and the P value was 1.53E-09.
Regarding the estimation model (estimation formula) in which the ratio of the amount of ceramide NP to the amount of ceramide NS (NP/NS) was used as the target index value, the correlation coefficient obtained in Example 1 was 0.607 and the P value was 1.00E-12 (see FIG. 5 ), and the correlation coefficient obtained in Example 2 was 0.573 and the P value was 3.17E-11.
Regarding the estimation model (estimation formula) in which the stratum corneum desquamation rate was used as the target index value, the correlation coefficient obtained in Example 1 was 0.453 and the P value was 3.25E-07 (see FIG. 4 ), and the correlation coefficient obtained in Example 2 was 0.507 and the P value was 6.50E-09.
Therefore, even when infrared absorption spectrum information obtained by a method different from that of Example 1 is used, as in Example 2, it was found that the estimation accuracy of the estimation model can be maintained almost similarly to Example 1.

In Example 1, an example was shown in which all (6950) absorbances in the wavenumber region from 600 cm ^-1 to 4000 cm ^-1 (wavenumber resolution 0.5 cm ^-1 ) output from an infrared spectrometer are used as the infrared absorption spectrum information. In other words, the absorbances included in the infrared absorption spectrum information used in Example 1 are intensity values at a constant wavenumber interval (wavenumber resolution 0.5 cm ^-1 ).
In Example 3, infrared absorption spectrum information was used that included 25 absorbances in the wavenumber region from 600 cm ⁻¹ to 4000 cm ⁻¹ with a constant wavenumber interval that was larger than that of Example 1. That is, the number of absorbances in the infrared absorption spectrum information used in Example 3 was 1/278 of that of Example 1.

In Example 3, an estimation model was constructed in the same manner as in Examples 1 and 2, and the correlation between the target index value (estimated index value) obtained by inputting the infrared absorption spectrum information of the validation data set (30% of the data set) into the estimation model and the known target index value of that data set was verified. Then, the correlation verified in Example 1 was compared with the correlation verified in Example 3.

As a result, for the estimation model (estimation equation) with the stratum corneum moisture content (Skicon-conductance) as the target index value, the correlation coefficient obtained in Example 1 was 0.491 and the P value was 2.49E-08 (see FIG. 6 ), and the correlation coefficient obtained in Example 3 was 0.522 and the P value was 2.26E-09.
Regarding the estimation model (estimation equation) in which the stratum corneum moisture content (Corneometer-capacitance) was used as the target index value, the correlation coefficient obtained in Example 1 was 0.577 and the P value was 2.15E-11 (see FIG. 6 ), and the correlation coefficient obtained in Example 3 was 0.574 and the P value was 3.12E-11.
Regarding the estimation model (estimation equation) with transepidermal water loss (TEWL) as the target index value, the correlation coefficient obtained in Example 1 was 0.460 and the P value was 2.91E-07 (see FIG. 6 ), and the correlation coefficient obtained in Example 3 was 0.417 and the P value was 4.35E-06.
Regarding the estimation model (estimation formula) in which the ratio of the amount of ceramide NP to the amount of ceramide NS (NP/NS) was used as the target index value, the correlation coefficient obtained in Example 1 was 0.607 and the P value was 1.00E-12 (see FIG. 5 ), and the correlation coefficient obtained in Example 3 was 0.545 and the P value was 4.38E-10.
Regarding the estimation model (estimation formula) in which the stratum corneum delamination rate was used as the target index value, the correlation coefficient obtained in Example 1 was 0.453 and the P value was 3.25E-07 (see FIG. 4 ), and the correlation coefficient obtained in Example 3 was 0.542 and the P value was 3.37E-10.
Therefore, even when infrared absorption spectrum information formed from absorbances having wavenumber intervals and numbers different from those in Example 1 is used as in Example 3, it was found that the estimation accuracy of the estimation model can be maintained almost similarly to Example 1.

[Comparative Example]
Finally, the method disclosed in the above-mentioned Patent Document 2 will be presented as a comparative example, and the results of comparing the estimation accuracy between this comparative example and Example 1 will be described.
As described above, the comparative example is a method in which the peak area ratio β/α is calculated and the calculated ratio β/α value is used as an index to evaluate the properties of the stratum corneum of the skin for purposes other than medical treatment. The method of this comparative example estimated transepidermal water loss (TEWL), and the estimation accuracy was calculated by comparing the obtained estimated value with the actual measured value measured as described above.
As a result, the estimation accuracy of the comparative example was 0.274 in correlation coefficient (R value) and 3.30E-03 in P value, whereas the estimation accuracy of the method of Example 1 was 0.460 in correlation coefficient (R value) and 2.91E-07 in P value, as shown in FIG.
Therefore, Example 1 enables estimation of stratum corneum characteristics with higher accuracy than the Comparative Example, demonstrating the superiority of Example 1.

10 Information processing device (skin index estimation device)
11 CPU
12 Memory 13 Input/Output I/F
14 Communication unit 15 Display device 16 Input device

Claims (8)

One or more processors
An acquisition step of acquiring infrared absorption spectrum information regarding the stratum corneum peeled from the skin of the subject;
an estimation step of estimating the target index value of the subject by inputting the acquired infrared absorption spectrum information into an estimation model obtained by using a plurality of data sets of infrared absorption spectrum information on the stratum corneum peeled from the skin of a plurality of sample donors and known target index values;
Run
The infrared absorption spectrum information includes at least a plurality of intensity values in a first wave number region of 600 cm ⁻¹ or more and 1800 cm ⁻¹ or less and a plurality of intensity values in a second wave number region of 2800 ^{cm −1} or more and 3700 cm ⁻¹ or less;
The target index value is an index value related to a stratum corneum characteristic or a skin surface characteristic.
Skin index estimation method. One or more processors
An acquisition step of acquiring infrared absorption spectrum information regarding the stratum corneum peeled from the skin of the subject;
an estimation step of estimating the target index value of the subject by inputting the acquired infrared absorption spectrum information into an estimation model obtained by using a plurality of data sets of infrared absorption spectrum information on the stratum corneum peeled from the skin of a plurality of sample donors and known target index values;
Run
The infrared absorption spectrum information includes at least a plurality of intensity values in a peak region and a non-peak region of a first wavenumber region of 600 cm ⁻¹ or more and ¹⁸⁰⁰ cm −1 or less , and a plurality of intensity values in a peak region and a non-peak region of a second wavenumber region of 2800 cm −1 or more and 3700 cm −1 or ^{less ,}
The target index value is an index value related to a stratum corneum characteristic or a skin surface characteristic.
Skin index estimation method. One or more processors
An acquisition step of acquiring infrared absorption spectrum information regarding the stratum corneum peeled from the skin of the subject;
an estimation step of estimating the target index value of the subject by inputting the acquired infrared absorption spectrum information into an estimation model obtained by using a plurality of data sets of infrared absorption spectrum information on the stratum corneum peeled from the skin of a plurality of sample donors and known target index values;
Run
The infrared absorption spectrum information includes at least a plurality of intensity values sampled at regular wavenumber intervals from a first wavenumber region of 600 cm ⁻¹ to 1800 cm ⁻¹ and at regular wavenumber intervals from a second wavenumber region of 2800 cm ⁻¹ to 3700 cm ⁻¹ ,
The target index value is an index value related to a stratum corneum characteristic or a skin surface characteristic.
Skin index estimation method. The target index value is a stratum corneum delamination rate, a stratum corneum cell area, a stratum corneum cell length, or a stratum corneum cell circularity rate.
The skin index estimation method according to claim 1 . The target index value is the total amount of intercellular lipids, the amount of ceramide, the amount of cholesterol, or the amount of free fatty acid, the amount of each ceramide molecular species in the intercellular lipids, or the ratio of the amount of each ceramide molecular species in the intercellular lipids.
The skin index estimation method according to claim 1 . The target index value is the stratum corneum moisture content or the transepidermal water loss.
The skin index estimation method according to claim 1 . The target index value is an index value of the scale state, the smoothness or the wrinkle state of the skin obtained by measuring the surface shape of the skin, or an index value of the roughness of the skin surface, the depth of wrinkles, the number of wrinkles, the volume of wrinkles, the area of wrinkles or the wrinkle area ratio obtained by measuring a replica that reproduces the surface shape of the skin.
The skin index estimation method according to claim 1 . A skin index estimation device including at least the one or more processors and a memory,
The estimation model is available,
The memory stores a computer program;
The skin index estimation device is capable of executing the skin index estimation method according to claim 1 by the one or more processors executing the computer program .

Images