JP7806355B2 - Skin viscoelasticity index estimation method - Google Patents
Skin viscoelasticity index estimation methodInfo
- Publication number
- JP7806355B2 JP7806355B2 JP2022122760A JP2022122760A JP7806355B2 JP 7806355 B2 JP7806355 B2 JP 7806355B2 JP 2022122760 A JP2022122760 A JP 2022122760A JP 2022122760 A JP2022122760 A JP 2022122760A JP 7806355 B2 JP7806355 B2 JP 7806355B2
- Authority
- JP
- Japan
- Prior art keywords
- cell
- image
- stratum corneum
- standard deviation
- brightness
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Landscapes
- Investigating Or Analysing Biological Materials (AREA)
- Measuring And Recording Apparatus For Diagnosis (AREA)
- Image Analysis (AREA)
Description
本発明は、皮膚粘弾性指標の推定方法に関する。 The present invention relates to a method for estimating a skin viscoelasticity index.
皮膚の健康状態は、様々な指標を用いて評価される。例えば、皮膚の「はり」感の有無が健康状態や老化度の評価指標の一つとされる。「はり」は、角層・表皮由来のはりと真皮由来のはりの二つに分けられる。特に真皮由来のはりは、指で皮膚を押すと押し返すような弾力があり、指を離すと速やかに元に戻る状態をいい、物理的には粘弾性ともいう。そして皮膚の示すこの粘弾性は、年齢や紫外線被曝、化学物質への皮膚露出などによっても低下する。このため、皮膚のはりは、皮膚老化の指標ともなり得る。
皮膚を一つの弾性体であるとみなして、皮膚の粘弾性率を測定し、「はり」という官能的な評価方法に代えて用いようとする方法及び測定装置が開発されている。代表的な皮膚粘弾性測定装置としてＣＵＴＯＭＥＴＥＲ・キュートメーター（商品名）（Ｃｏｕｒａｇｅ ＆ Ｋｈａｚａｋａ社製）がある。この装置は、皮膚表面を陰圧状態のプローブ開口部に引き込み、開口部に引き込まれた皮膚長をプリズムで測定し、次いで吸引を解除し、解除したときの変位長（戻り）を同様に測定し、この測定結果をパラメータとして粘弾性率を求めるものである（非特許文献１参照）。この装置を用いた粘弾性率を指標として、皮膚の老化度の評価や、化粧料の効果を評価することが化粧品業界や、美容関連医療分野において広く普及している。
しかし、皮膚粘弾性測定装置による皮膚粘弾性の測定技術は、測定に用いるプローブや装置が高価であり、また使用にあたっては熟練を要する。そのため、測定可能な施設は専門的なサロンや医療機関等に限られ、安価、簡便、迅速に皮膚粘弾性指標を知ることは困難である。
The health of skin is evaluated using a variety of indices. For example, the presence or absence of a feeling of "firmness" in the skin is considered one of the indicators for assessing health and the degree of aging. "Firmness" can be divided into two types: firmness derived from the stratum corneum/epidermis, and firmness derived from the dermis. Dermal firmness, in particular, has the elasticity to push back when pressed with a finger, and quickly returns to its original state when the finger is removed; in physical terms, this is also called viscoelasticity. The viscoelasticity of the skin decreases with age, exposure to ultraviolet rays, and exposure to chemicals. For this reason, skin firmness can also be used as an indicator of skin aging.
Methods and devices have been developed to measure the viscoelastic modulus of skin by treating the skin as a single elastic body, replacing the sensory evaluation method of "firmness." A representative skin viscoelasticity measuring device is the CUTOMETER (product name) (manufactured by Courage & Khazaka). This device draws the skin surface into a probe opening under negative pressure, measures the length of the skin drawn into the opening with a prism, then releases the suction and similarly measures the displacement length (return) upon release, and uses these measurement results as parameters to determine the viscoelastic modulus (see Non-Patent Document 1). Using the viscoelastic modulus measured with this device as an index to evaluate the degree of skin aging and the effectiveness of cosmetics has become widespread in the cosmetics industry and in the field of aesthetic medicine.
However, the technology for measuring skin viscoelasticity using a skin viscoelasticity measuring device requires expensive probes and equipment, and requires skilled operation. As a result, measurements are only available at specialized salons and medical institutions, making it difficult to determine skin viscoelasticity indices cheaply, easily, and quickly.
本発明は、皮膚粘弾性指標を推定する推定方法を提供することを課題とする。 The present invention aims to provide a method for estimating skin viscoelasticity indexes.
本発明は、上記課題を解決するために鋭意検討した結果、角層の観察画像から得られる角層構造を指標として、皮膚粘弾性指標を推定できることを見出してなされたものである。 The present invention was made after extensive research to solve the above-mentioned problems, and it was discovered that skin viscoelasticity indexes can be estimated using the stratum corneum structure obtained from observation images of the stratum corneum as an index.
具体的には、本発明の課題を解決するための手段は以下の通りである。
１．角層画像の１細胞領域を特定し、
これらから算出される角層パラメータ値の１以上と、皮膚粘弾性測定装置により計測される皮膚弾力パラメータとの下記（１）～（５）から選ばれる１以上の相関関係に基づいて、皮膚粘弾性指標を推定することを特徴とする皮膚粘弾性指標推定方法。
（１）長方形の長短辺比、画像内の１細胞輝度平均値の標準偏差、画像内の１細胞輝度中央値の標準偏差からなる群から選択される１以上の角層パラメータを説明変数、Ｒ０：最大吸引高さを目的変数とする相関関係
（２）画像内の１細胞の各画素の輝度の標準偏差の画像内の標準偏差を説明変数、Ｒ２：解除後の最終戻り／最大吸引高さを目的変数とする相関関係
（３）１細胞輝度標準偏差、１細胞輝度最小値、画像内の１細胞の各画素の輝度の標準偏差の画像内の標準偏差、画像内の１細胞輝度最小値の標準偏差からなる群から選択される１以上の角層パラメータを説明変数、Ｒ７：陰圧解除後の瞬時の戻り／最大吸引高さを目的変数とする相関関係
（４）画像内の１細胞輝度最小値の標準偏差、画像内の１細胞の各画素の輝度の標準偏差の画像内の標準偏差からなる群から選択される１以上の角層パラメータを説明変数、Ｒ５：陰圧解除後の瞬時の戻り／吸引時の瞬時の変形を目的変数とする相関関係
（５）１細胞面積、１細胞周囲長、１細胞に外接する長方形面積、画像内の１細胞輝度中央値の標準偏差からなる群から選択される１以上の角層パラメータを説明変数、Ｒ６：吸引時の時間をかけた変形／吸引時の瞬時の変形を目的変数とする相関関係
２．前記１細胞領域が、予め複数枚の学習用角層画像と各学習用角層画像における目視評価による１細胞領域を機械学習させた機械学習モデルを用いて特定されることを特徴とする１．に記載の皮膚粘弾性指標推定方法。
Specifically, the means for solving the problems of the present invention are as follows.
1. Identify a single cell area in the stratum corneum image.
A method for estimating a skin viscoelasticity index, characterized by estimating a skin viscoelasticity index based on one or more correlations selected from (1) to (5) below between one or more stratum corneum parameter values calculated from these and skin elasticity parameters measured by a skin viscoelasticity measuring device.
(1) Correlation in which one or more stratum corneum parameters selected from the group consisting of the ratio of the length and width of a rectangle, the standard deviation of the average brightness of one cell in the image, and the standard deviation of the median brightness of one cell in the image are used as explanatory variables, and R0 is the maximum suction height as the objective variable. (2) Correlation in which the standard deviation of the brightness of each pixel of one cell in the image is used as the explanatory variable, and R2 is the final return/maximum suction height after release as the objective variable. (3) Correlation in which one or more stratum corneum parameters selected from the group consisting of the standard deviation of brightness of one cell, the minimum brightness of one cell, the standard deviation of the standard deviation of brightness of each pixel of one cell in the image, and the standard deviation of the minimum brightness of one cell in the image are used as explanatory variables. (4) a correlation in which one or more stratum corneum parameters selected from the group consisting of the standard deviation of the minimum brightness of a single cell in the image and the standard deviation of the brightness of each pixel of a single cell in the image are used as explanatory variables; R5: a correlation in which one or more stratum corneum parameters selected from the group consisting of the area of a single cell, the perimeter of a single cell, the area of a rectangle circumscribing a single cell, and the standard deviation of the median brightness of a single cell in the image are used as explanatory variables; and R6: a correlation in which deformation over time during suction/instantaneous deformation during suction are used as explanatory variables. 2. The method for estimating a skin viscoelasticity index described in 1., wherein the single-cell region is identified using a machine learning model that has previously been trained on a plurality of training stratum corneum images and single-cell regions in each training stratum corneum image through visual evaluation.
本発明の皮膚粘弾性指標推定方法は、角層構造から得られる数値を指標とするものであり、非常に容易に、かつ精度良く、皮膚粘弾性指標を推定することができる。 The skin viscoelasticity index estimation method of the present invention uses a numerical value obtained from the stratum corneum structure as an index, and can estimate the skin viscoelasticity index very easily and accurately.
本発明は、角層画像の１細胞領域を特定し、
これから算出される角層パラメータ値の１以上と、顔の皮膚画像解析カウンセリングシステムによる解析結果との間に存在する相関関係に基づいて、皮膚粘弾性指標を推定する皮膚粘弾性指標推定方法に関する。
The present invention identifies a single cell region in an image of the stratum corneum,
The present invention relates to a skin viscoelasticity index estimation method that estimates a skin viscoelasticity index based on the correlation between one or more of the stratum corneum parameter values calculated from this and the analysis results obtained by a facial skin image analysis counseling system.
本発明の皮膚粘弾性指標推定方法により、以下に示す皮膚粘弾性指標を推定することができる。
Ｒ０＝Ｕｆ、肌の伸び、柔らかさの指標である。機器で吸引後１０秒時点での最終伸長値。
Ｒ２＝Ｕａ／Ｕｆ、粘性変形を含む全体の弾力を示し、皮膚の粘弾性の指標である。吸引から解除後の最終的な皮膚の戻り率。
Ｒ７＝Ｕｒ／Ｕｆ、生物学的な粘弾性（全体弾力に対する急激な退縮の割合）であり、純弾性の指標である。最終伸長値に対する戻り率の初速。
Ｒ５＝Ｕｒ／Ｕｅ、粘性変形を含まない皮膚の正味の弾性を示し、皮膚の瞬発的な回復力の指標である。吸引解除後の戻り率の初速／吸引開始時の伸長値の初速。
Ｒ６＝Ｕｖ／Ｕｅ、弾性伸展に対する粘弾性伸展の割合を示し、皮膚の瞬時の粘弾性の指標である。吸引後の初速を除いた最終的な皮膚の伸長値／皮膚の吸引後の伸長値の初速。
（Ｕｅ：吸引時の瞬時の変形、Ｕｖ：吸引時の時間をかけた変形、Ｕｆ：最大吸引高、Ｕｒ：陰圧解除後の瞬時の戻り、Ｕａ：解除後の最終戻り）
The skin viscoelasticity index estimation method of the present invention can estimate the following skin viscoelasticity indexes.
R0 = Uf, an index of skin stretch and softness. The final stretch value at 10 seconds after suction with the device.
R2 = Ua/Uf, which indicates the overall elasticity including viscous deformation and is an index of the viscoelasticity of the skin. The final skin return rate after release from suction.
R7 = Ur/Uf, biological viscoelasticity (ratio of rapid retraction to total elasticity), a measure of pure elasticity. Initial rate of return to final elongation value.
R5 = Ur/Ue, which indicates the net elasticity of the skin excluding viscous deformation and is an index of the skin's instantaneous recovery force. Initial speed of return rate after suction is released / Initial speed of elongation value at the start of suction.
R6 = Uv/Ue, which indicates the ratio of viscoelastic stretch to elastic stretch, and is an index of the instantaneous viscoelasticity of the skin. Final skin stretch value excluding the initial speed after suction/initial speed of the stretch value after suction of the skin.
(Ue: instantaneous deformation during suction, Uv: deformation over time during suction, Uf: maximum suction height, Ur: instantaneous return after release of negative pressure, Ua: final return after release)
画像を得るための角層の採取方法は、バイオプシー、テープストリッピング法のいずれであってもよいが、被験者の負担が少ないため、テープストリッピング法が好ましい。テープストリッピング法は、粘着性テープを皮膚に貼り付けた後、剥がして皮膚の表層を回収する方法である。 The stratum corneum can be collected for imaging by either biopsy or tape stripping, but tape stripping is preferred because it places less strain on the subject. Tape stripping involves applying adhesive tape to the skin and then peeling it off to collect the surface layer of the skin.
テープストリッピング法で採取した角層細胞は、透過光観察（微分干渉法、位相差法、暗視野観察法などを含む）または反射光観察により撮像し、細胞の形態を観察する。角層細胞は、無染色の状態で観察することが好ましいが、必要に応じて染色することもできる。
細胞の観察は、細胞観察が可能な顕微鏡を用いて行うことができ、例えば、キーエンス株式会社製のデジタルマイクロスコープＶＨＸ－５００、ＡｎＭｏ Ｅｌｅｃｔｒｏｎｉｃｓ Ｃｏｒｐｏｒａｔｉｏｎ社製のデジタルマイクロスコープＤｉｎｏ－Ｌｉｔｅ等を用いることができる。
観察条件は、細胞の詳細が確認できるものであれば制限されないが、例えば、１．０μｍ／ｐｉｘｅｌ以上の解像度で約２０万画素以上の条件等が挙げられる。
The corneal cells collected by tape stripping are imaged using transmitted light observation (including differential interference contrast, phase contrast, dark field observation, etc.) or reflected light observation to observe the cell morphology. Corneal cells are preferably observed in an unstained state, but can also be stained if necessary.
Cells can be observed using a microscope capable of cell observation, such as a digital microscope VHX-500 manufactured by Keyence Corporation or a digital microscope Dino-Lite manufactured by AnMo Electronics Corporation.
There are no limitations on the observation conditions as long as the details of the cells can be confirmed, but examples include conditions such as a resolution of 1.0 μm/pixel or more and approximately 200,000 pixels or more.
撮像した角層画像を、画像処理システムを用いて画像処理し、角層画像の１細胞領域を特定する。この際、細胞領域面積、重層剥離領域、重層剥離率等を特定することもできる。
１細胞領域とは、角層画像内の個々の角層細胞の領域である。
細胞領域面積とは、角層画像中の角層細胞領域の面積の総和である。
重層剥離面積とは、角層が２層以上重なって剥離した領域の面積である。
重層剥離率とは、「細胞領域面積」に対する「重層剥離面積」の割合（重層剥離面積／細胞領域面積）である。
The captured image of the stratum corneum is processed using an image processing system to identify a single cell region in the stratum corneum image. In this process, the area of the cell region, the layer delamination region, the layer delamination rate, etc. can also be identified.
A single cell region is the region of an individual stratum corneum cell within the stratum corneum image.
The cell region area is the total area of the stratum corneum cell region in the stratum corneum image.
The area of multilayer peeling is the area of the region where two or more layers of the stratum corneum are peeled off in layers.
The layer detachment rate is the ratio of the "layer detached area" to the "cell region area" (layer detached area/cell region area).
そして、この特定した１細胞領域等に基づいて、推定する皮膚粘弾性指標の種類等に応じて、１細胞面積、１細胞周囲長、１細胞に外接する長方形面積、長方形の長短辺比、１細胞輝度標準偏差、１細胞輝度最小値、画像内の１細胞輝度平均値の標準偏差、画像内の１細胞輝度中央値の標準偏差、画像内の１細胞輝度最小値の標準偏差、画像内の１細胞の各画素の輝度の標準偏差の画像内の標準偏差からなる群から選択される１以上を数値化する。角層画像は、カラー画像を用いてもよく、画像処理によりグレースケールにした画像を用いてもよく、両方を用いることもできる。 Then, based on this identified single-cell area, etc., and depending on the type of skin viscoelasticity index to be estimated, one or more of the following is quantified: single-cell area, single-cell perimeter, area of a rectangle circumscribing a single cell, ratio of the length and width of the rectangle, single-cell luminance standard deviation, single-cell luminance minimum value, standard deviation of the average single-cell luminance value in the image, standard deviation of the median single-cell luminance value in the image, standard deviation of the minimum single-cell luminance value in the image, and standard deviation of the luminance of each pixel of a single cell in the image. The stratum corneum image may be a color image, an image converted to grayscale by image processing, or both.
１細胞面積とは、角層画像内の個々の角層細胞の面積を全細胞で平均した値である。
１細胞周囲長とは、角層画像内の個々の角層細胞の周囲長を全細胞で平均した値である。
１細胞に外接する長方形面積とは、角層画像内の個々の角層細胞全体を覆う最小面積の長方形の面積を全細胞で平均した値である。
長方形の長短辺比とは、角層画像内の個々の角層細胞全体を覆う最小面積の長方形の長辺と短辺の比（長辺／短辺）を全細胞で平均した値である。
The area of one cell is the average area of all cells in the stratum corneum image.
The perimeter of one cell is the average value of the perimeters of individual stratum corneum cells in the stratum corneum image over all cells.
The rectangular area circumscribing one cell is the average area of the smallest rectangle that covers the entirety of each individual stratum corneum cell in the stratum corneum image over all cells.
The ratio of the long and short sides of a rectangle is the average ratio of the long side to the short side (long side/short side) of the rectangle with the smallest area that covers the entire individual stratum corneum cell in the stratum corneum image for all cells.
１細胞輝度標準偏差とは、角層画像内の個々の角層細胞の細胞領域内の輝度の標準偏差を求めた値である。
１細胞輝度最小値とは、角層画像内の個々の角層細胞の細胞領域内の輝度の最小値を全細胞で平均した値である。
１細胞輝度平均値とは、角層画像内の個々の角層細胞の細胞領域内の平均輝度値を全細胞で平均した値である。
１細胞輝度中央値とは、角層画像内の個々の角層細胞の細胞領域内の輝度値中央値を全細胞で平均した値である。
画像内の１細胞の各画素の輝度の標準偏差の画像内の標準偏差とは、角層画像内の個々の角層細胞の細胞領域内の輝度の標準偏差（１細胞内輝度標準偏差）の角層画像内の全細胞での標準偏差である。
以下、１細胞面積、１細胞周囲長、１細胞に外接する長方形面積、長方形の長短辺比、１細胞輝度標準偏差、１細胞輝度最小値、画像内の１細胞輝度平均値の標準偏差、画像内の１細胞輝度中央値の標準偏差、画像内の１細胞輝度最小値の標準偏差、画像内の１細胞の各画素の輝度の標準偏差の画像内の標準偏差からなる群から選択される１以上を数値化したものを角層パラメータ値ともいう。
The single-cell luminance standard deviation is a value obtained by calculating the standard deviation of luminance within the cell region of an individual stratum corneum cell in an image of the stratum corneum.
The minimum luminance value of one cell is the average of the minimum luminance values within the cell region of each stratum corneum cell in the stratum corneum image over all cells.
The average brightness of one cell is the average brightness of the cell region of each stratum corneum cell in the stratum corneum image, calculated by averaging the average brightness of all cells.
The single-cell median brightness is the average of the median brightness values within the cell region of each individual stratum corneum cell in the stratum corneum image across all cells.
The standard deviation of the brightness of each pixel in a cell in an image is the standard deviation of the brightness within the cell region of each stratum corneum cell in the stratum corneum image (standard deviation of brightness within a single cell) across all cells in the stratum corneum image.
Hereinafter, the stratum corneum parameter value refers to a numerical value of one or more selected from the group consisting of the area of a single cell, the perimeter of a single cell, the area of a rectangle circumscribing a single cell, the ratio of the length to the width of the rectangle, the standard deviation of the luminance of a single cell, the minimum luminance of a single cell, the standard deviation of the average luminance of a single cell in an image, the standard deviation of the median luminance of a single cell in an image, the standard deviation of the minimum luminance of a single cell in an image, and the standard deviation of the luminance of each pixel of a single cell in an image.
画像処理システムによる数値化は、公知の画像処理システムを用いて行うことができ、例えば、上記したデジタルマイクロスコープに付属の画像処理ソフトウェアや、市販の画像処理ソフトウェア等を用いることができる。また、予め複数枚の学習用角層画像と、各学習用角層画像における、目視評価による１細胞領域（１つ１つの細胞の領域）を機械学習させた機械学習モデルを備える画像処理システムを用いることもできる。この機械学習モデルには、目視評価による細胞領域（細胞が存在している領域）、目視評価による重層剥離領域（角層が２層以上重なって剥離した領域）等を学習させることもできる。この機械学習モデルに数値化したい角層画像を入力し、この角層画像の１細胞領域を出力し、機械学習モデルが出力した１細胞領域に基づいて角層パラメータ値（１細胞面積、１細胞周囲長、１細胞に外接する長方形面積、長方形の長短辺比、１細胞輝度標準偏差、１細胞輝度最小値、画像内の１細胞輝度平均値の標準偏差、画像内の１細胞輝度中央値の標準偏差、画像内の１細胞輝度最小値の標準偏差、画像内の１細胞の各画素の輝度の標準偏差の画像内の標準偏差）を算出することで、目視評価に基づく角層パラメータ値と同等の角層パラメータ値を迅速に算出することができる。さらに、機械学習モデルを備える画像処理システムが出力した１細胞領域を、人が目視で修正して角層パラメータ値を求めることもできる。 Digitalization using an image processing system can be performed using a known image processing system, such as the image processing software included with the digital microscope described above or commercially available image processing software. Alternatively, an image processing system equipped with multiple training stratum corneum images and a machine learning model trained on visually evaluated single-cell regions (the regions of individual cells) in each training stratum corneum image can be used. This machine learning model can also be trained on visually evaluated cell regions (regions where cells exist) and visually evaluated multilayer delamination regions (regions where two or more layers of stratum corneum have been delaminate). By inputting the stratum corneum image to be quantified into this machine learning model, outputting the single-cell region of this stratum corneum image, and calculating stratum corneum parameter values (single-cell area, single-cell perimeter, area of the rectangle circumscribing the single cell, rectangle length ratio, single-cell luminance standard deviation, single-cell luminance minimum value, standard deviation of mean single-cell luminance within the image, standard deviation of median single-cell luminance within the image, standard deviation of minimum single-cell luminance within the image, and standard deviation of the luminance standard deviation of each pixel of a single cell within the image) based on the single-cell region output by the machine learning model, it is possible to quickly calculate stratum corneum parameter values equivalent to those based on visual evaluation. Furthermore, the single-cell region output by an image processing system equipped with a machine learning model can also be visually corrected by a human to determine stratum corneum parameter values.
この数値化した角層パラメータ値を指標として皮膚粘弾性指標を推定する方法は特に制限されず、公知の方法を用いることができる。例えば、角層パラメータ値を単回帰分析してもよく、重回帰分析してもよい。重回帰分析は、線形重回帰分析でもよく非線形重回帰分析でもよい。また、１細胞面積、１細胞周囲長、１細胞に外接する長方形面積、長方形の長短辺比、１細胞輝度標準偏差、１細胞輝度最小値、画像内の１細胞輝度平均値の標準偏差、画像内の１細胞輝度中央値の標準偏差、画像内の１細胞輝度最小値の標準偏差、画像内の１細胞の各画素の輝度の標準偏差の画像内の標準偏差以外の他のパラメータを重回帰分析の説明変数に含めてもよい。 The method for estimating the skin viscoelasticity index using these quantified stratum corneum parameter values as indices is not particularly limited, and known methods can be used. For example, the stratum corneum parameter values may be subjected to simple regression analysis or multiple regression analysis. The multiple regression analysis may be linear multiple regression analysis or nonlinear multiple regression analysis. Furthermore, parameters other than the single cell area, single cell perimeter, rectangular area circumscribing a single cell, rectangular side ratio, single cell brightness standard deviation, single cell brightness minimum value, standard deviation of single cell brightness average value in the image, standard deviation of single cell brightness median value in the image, standard deviation of single cell brightness minimum value in the image, and standard deviation of brightness of each pixel of a single cell in the image may be included as explanatory variables in the multiple regression analysis.
＜機械学習用サンプル＞
女性４１２名（１８～８７才、平均４６．８才）について、テープストリッピング法により、顔から角層細胞を採取した。
採取した角層細胞は、無染色の状態で、デジタルマイクロスコープ（キーエンス株式会社、ＶＨＸ－５０００）を用いて、８ｂｉｔ（ＲＧＢカラー）、０．４１μｍ／ｐｉｘｅｌの解像度、１６００×１２００ｐｉｘｅｌの画素数で、透過光による撮影を行った。撮影は各サンプル２～３視野ずつ撮影を行い機械学習として用いた。
<Machine learning sample>
Corneal cells were collected from the faces of 412 women (aged 18-87 years, mean age 46.8 years) using the tape stripping method.
The collected stratum corneum cells were photographed in an unstained state using a digital microscope (Keyence Corporation, VHX-5000) with 8-bit (RGB color), 0.41 μm/pixel resolution, and 1600 x 1200 pixels using transmitted light. Two to three fields of view were photographed for each sample, and these images were used for machine learning.
画像アノテーション用ソフトウェア Ｌａｂｅｌｍｅを用いて、細胞領域、重層剥離領域、１細胞領域を目視により特定し、学習用角層画像を得た。
各学習用角層画像と、目視評価による細胞領域、重層剥離領域、１細胞領域を機械学習させ、細胞領域、重層剥離領域、１細胞領域を出力可能な機械学習モデルを得た。
Using image annotation software Labelme, cell regions, delaminated regions, and single-cell regions were visually identified to obtain learning images of the stratum corneum.
Each training stratum corneum image and the cell region, multilayer peeled region, and single-cell region determined by visual evaluation were subjected to machine learning, and a machine learning model capable of outputting the cell region, multilayer peeled region, and single-cell region was obtained.
＜サンプル＞
女性２２７名（２８～８７才、平均４７．２才）について、テープストリッピング法により、顔から角層細胞を採取した。一人当たり、３～５枚のサンプルを採取し、合計９６６枚のサンプルを得た。
試料調製方法は、機械学習用サンプルと同じである。
サンプル画像を、機械学習モデルを備えた画像処理装置により解析し、各角層画像の１細胞面積、１細胞周囲長、１細胞に外接する長方形面積、長方形の長短辺比、１細胞輝度標準偏差、１細胞輝度最小値、画像内の１細胞輝度平均値の標準偏差、画像内の１細胞輝度中央値の標準偏差、画像内の１細胞輝度最小値の標準偏差、画像内の１細胞の各画素の輝度の標準偏差の画像内の標準偏差を数値化し、角層パラメータ値を求めた。
さらに、真円度、近似する４－６角形、１細胞輝度平均値、１細胞輝度中央値、１細胞輝度最大値、画像内の１細胞輝度最大値の標準偏差、画像内の１細胞面積の標準偏差、画像内の１細胞周囲長の標準偏差、画像内の１細胞真円度の標準偏差、画像内の近似する４－６角形の標準偏差についても数値化し、角層パラメータ値を求めた。
真円度とは、個々の角層細胞の領域の正円度合い（４π×（面積）／（周長の２乗）により算出される。）の１画像における全細胞で平均した値である。
近似する４－６角形とは、個々の角層細胞形状の重心を中心として、角層細胞領域内に正４－６角形を配置し、正４－６角形からはみ出した領域の面積が最小となるように角度と大きさを調整し、はみ出した領域の面積が最小となった際の、はみ出した領域の面積を角層細胞面積で割り返し、正ｎ角形（ｎは４～６）の中でこの値が最小となるｎを近似するｎ角形とし、角層画像内の個々の角層細胞のｎを全細胞で平均した値である。
１細胞輝度最大値とは、角層画像内の個々の角層細胞の細胞領域内の輝度の最大値を全細胞で平均した値である。
<Sample>
Corneal cells were collected from the faces of 227 women (aged 28-87 years, average age 47.2 years) using the tape stripping method. Three to five samples were collected per person, for a total of 966 samples.
The sample preparation method was the same as for the machine learning samples.
The sample images were analyzed using an image processing device equipped with a machine learning model, and the following values were quantified for each stratum corneum image: area per cell, perimeter per cell, area of the rectangle circumscribing one cell, ratio of the rectangular sides, standard deviation of luminance per cell, minimum luminance per cell, standard deviation of average luminance per cell within the image, standard deviation of median luminance per cell within the image, standard deviation of minimum luminance per cell within the image, and standard deviation of luminance of each pixel of one cell within the image, to determine stratum corneum parameter values.
Furthermore, the circularity, approximate tetragon, average brightness per cell, median brightness per cell, maximum brightness per cell, standard deviation of maximum brightness per cell in the image, standard deviation of area per cell in the image, standard deviation of perimeter per cell in the image, standard deviation of circularity per cell in the image, and standard deviation of approximate tetragon in the image were also quantified to determine stratum corneum parameter values.
The circularity is the average value of the circularity of each individual corneocyte region (calculated by 4π×(area)/(square of perimeter)) for all cells in one image.
The approximate tetrahexagon is a regular tetrahexagon placed within the stratum corneum cell area, with the center of gravity of the shape of each stratum corneum cell at the center, and the angle and size are adjusted so that the area of the area protruding from the regular tetrahexagon is minimized. When the area of the protruding area is minimized, the area of the protruding area is divided by the area of the stratum corneum cell. The n that gives the smallest value among regular n-gons (n is 4 to 6) is defined as the approximate n-gon, and the n of each stratum corneum cell in the stratum corneum image is averaged across all cells.
The maximum brightness value of one cell is the average of the maximum brightness values within the cell region of each stratum corneum cell in the stratum corneum image over all cells.
なお、９６６枚のサンプルについて、目視評価による１細胞領域、細胞領域、重層剥離領域を、それぞれ機械学習モデルを備えた画像処理装置により解析した１細胞領域、細胞領域、重層剥離領域と比較した。
結果、目視評価による１細胞領域面積／機械学習モデルを備えた画像処理装置により解析した１細胞領域面＝１１７％、目視評価による細胞領域面積／機械学習モデルを備えた画像処理装置により解析した細胞領域面積＝９５％、目視評価による重層剥離領域面積／機械学習モデルを備えた画像処理装置により解析した重層剥離領域面積＝８６％であり、目視評価による１細胞領域、細胞領域、重層剥離領域と、機械学習モデルを備えた画像処理装置により解析した１細胞領域、細胞領域、重層剥離領域は、同等であった。
For the 966 samples, the single-cell regions, cell regions, and multilayer-exfoliated regions evaluated visually were compared with the single-cell regions, cell regions, and multilayer-exfoliated regions analyzed using an image processing device equipped with a machine learning model.
As a result, the area of the single-cell region by visual evaluation/the area of the single-cell region analyzed by the image processing device equipped with the machine learning model = 117%, the area of the cell region by visual evaluation/the area of the cell region analyzed by the image processing device equipped with the machine learning model = 95%, and the area of the multilayered peeled region by visual evaluation/the area of the multilayered peeled region analyzed by the image processing device equipped with the machine learning model = 86%, and the single-cell region, cell region, and multilayered peeled region by visual evaluation were equivalent to the single-cell region, cell region, and multilayered peeled region analyzed by the image processing device equipped with the machine learning model.
＜皮膚粘弾性指標の機械学習モデルを備えた画像処理装置により解析した角層構造による回帰分析＞
皮膚粘弾性指標測定装置名：Ｃｕｔｏｍｅｔｅｒ（登録商標）（Ｃｏｕｒａｇｅ ＋ Ｋｈａｚａｋａ Ｅｌｅｃｔｒｏｎｉｃ ＧｍｂＨ社（Ｃｏｌｏｇｎｅ，Ｇｅｒｍａｎｙ）製）
被験者：３３５名（２０～９２才、平均４４．３才）
測定部位、手順
被験者は、洗顔後１０分間安静にした後、Ｃｕｔｏｍｅｔｅｒ（登録商標）により頬部の皮膚粘弾性が測定され、頬部の角層がテープストリッピング法により採取された。
採取された角層は、デジタルマイクロスコープＤｉｎｏ－Ｌｉｔｅ（ＡｎＭｏ Ｅｌｅｃｔｒｏｎｉｃｓ Ｃｏｒｐｏｒａｔｉｏｎ社）を用いて８ｂｉｔ ＲＧＢ，０．４１μｍ／ｐｉｘｅｌ，１６００×１２００ｐｉｘｅｌの条件で撮像され、ＡＩおよび２値化による細胞形状の認識を行い、各角層パラメータ値を算出した。各角層パラメータ値を説明変数、皮膚粘弾性のＲ値を目的変数として、回帰分析を行った。解析にはＪＭＰ（登録商標）１６．２．０（ＳＡＳ Ｉｎｓｔｉｔｕｔｅ Ｉｎｃ．，ＮＣ，ＵＳＡ）を用いた。
皮膚粘弾性指標である皮膚粘弾性のＲ値は、Ｒ０、Ｒ２、Ｒ７、Ｒ５、Ｒ６である。
<Regression analysis based on stratum corneum structure analyzed using an image processing device equipped with a machine learning model of skin viscoelasticity index>
Name of skin viscoelasticity index measuring device: Cutometer (registered trademark) (manufactured by Courage + Khazaka Electronic GmbH (Cologne, Germany))
Subjects: 335 people (ages 20-92, average age 44.3)
Measurement Site and Procedure After washing their faces, the subjects rested for 10 minutes, after which the skin viscoelasticity of the cheeks was measured using Cutometer (registered trademark), and the stratum corneum of the cheeks was sampled using the tape stripping method.
The collected stratum corneum samples were imaged using a Dino-Lite digital microscope (AnMo Electronics Corporation) at 8-bit RGB resolution, 0.41 μm/pixel, and 1600 × 1200 pixels. Cell shape was recognized using AI and binarization, and stratum corneum parameter values were calculated. Regression analysis was performed using each stratum corneum parameter as the explanatory variable and the skin viscoelasticity R value as the objective variable. JMP (registered trademark) 16.2.0 (SAS Institute Inc., NC, USA) was used for the analysis.
The R values of skin viscoelasticity, which are skin viscoelasticity indexes, are R0, R2, R7, R5, and R6.
「単回帰分析」
各角層パラメータ値と、皮膚粘弾性指標とを単回帰分析した。相関が認められたものについて、相関係数ｒ、ｐ値、単回帰直線の式（Ｙ＝ａＸ＋ｂ、Ｘが各角層パラメータ値、Ｙが皮膚粘弾性指標）を表１に示す。
Simple regression analysis was performed on each stratum corneum parameter value and the skin viscoelasticity index. For those where a correlation was observed, the correlation coefficient r, p value, and the equation of the simple regression line (Y = aX + b, where X is each stratum corneum parameter value and Y is the skin viscoelasticity index) are shown in Table 1.
表１に示すように、本発明の（１）～（５）に記載された各角層パラメータ値は、（１）～（５）のそれぞれに記載された皮膚粘弾性指標と相関（｜ｒ｜≧０．２）を有していた。 As shown in Table 1, the stratum corneum parameter values described in (1) to (5) of the present invention correlated (|r| ≧ 0.2) with the skin viscoelasticity indexes described in (1) to (5), respectively.
「重回帰分析」
各皮膚粘弾性指標として、相関が認めらなかった角層パラメータを含めた２０個の角層パラメータ値を説明変数として、重回帰分析した。重回帰直線の式（Ｙ＝ａ１Ｘ１＋ａ２Ｘ２＋・・・＋ａ１４Ｘ１４＋ｂ、Ｘｎが各角層パラメータ値、Ｙがマーカータンパク量）を表２、３に示す。
"Multiple regression analysis"
Multiple regression analysis was performed using 20 stratum corneum parameter values, including those for which no correlation was observed, as explanatory variables for each skin viscoelasticity index. The equation of the multiple regression line (Y = a1X1 + a2X2 + ... + a14X14 + b, where Xn is each stratum corneum parameter value and Y is the amount of marker protein) is shown in Tables 2 and 3.
表２、３に示す通り、本発明の（１）～（５）に記載の各角層パラメータ値と、他のパラメータ値を組み合わせることにより、皮膚粘弾性指標とより高い相関が認められた。 As shown in Tables 2 and 3, a higher correlation with the skin viscoelasticity index was observed when the stratum corneum parameter values described in (1) to (5) of the present invention were combined with other parameter values.
Claims (2)
これらから算出される角層パラメータ値の１以上と、皮膚粘弾性測定装置により計測される皮膚弾力パラメータとの下記（１）～（５）から選ばれる１以上の相関関係に基づいて、皮膚粘弾性指標を推定することを特徴とする皮膚粘弾性指標推定方法。
（１）長方形の長短辺比、画像内の１細胞輝度平均値の標準偏差、画像内の１細胞輝度中央値の標準偏差からなる群から選択される１以上の角層パラメータを説明変数、Ｒ０：最大吸引高さを目的変数とする相関関係
（２）画像内の１細胞の各画素の輝度の標準偏差の画像内の標準偏差を説明変数、Ｒ２：解除後の最終戻り／最大吸引高さを目的変数とする相関関係
（３）１細胞輝度標準偏差、１細胞輝度最小値、画像内の１細胞の各画素の輝度の標準偏差の画像内の標準偏差、画像内の１細胞輝度最小値の標準偏差からなる群から選択される１以上の角層パラメータを説明変数、Ｒ７：陰圧解除後の瞬時の戻り／最大吸引高さを目的変数とする相関関係
（４）画像内の１細胞輝度最小値の標準偏差、画像内の１細胞の各画素の輝度の標準偏差の画像内の標準偏差からなる群から選択される１以上の角層パラメータを説明変数、Ｒ５：陰圧解除後の瞬時の戻り／吸引時の瞬時の変形を目的変数とする相関関係
（５）１細胞面積、１細胞周囲長、１細胞に外接する長方形面積、画像内の１細胞輝度中央値の標準偏差からなる群から選択される１以上の角層パラメータを説明変数、Ｒ６：吸引時の時間をかけた変形／吸引時の瞬時の変形を目的変数とする相関関係 Identify a single cell area in the stratum corneum image,
A method for estimating a skin viscoelasticity index, characterized by estimating a skin viscoelasticity index based on one or more correlations selected from (1) to (5) below between one or more stratum corneum parameter values calculated from these and skin elasticity parameters measured by a skin viscoelasticity measuring device.
(1) Correlation in which one or more stratum corneum parameters selected from the group consisting of the ratio of the length and width of a rectangle, the standard deviation of the average brightness of a single cell in the image, and the standard deviation of the median brightness of a single cell in the image are explanatory variables, and R0: maximum suction height is the objective variable. (2) Correlation in which the standard deviation of the brightness of each pixel of a single cell in the image is the explanatory variable, and R2: final return/maximum suction height after release is the objective variable. (3) Correlation in which one or more stratum corneum parameters selected from the group consisting of the standard deviation of brightness of a single cell, the minimum brightness of a single cell, the standard deviation of the standard deviation of brightness of each pixel of a single cell in the image, and the standard deviation of the minimum brightness of a single cell in the image are explanatory variables. R7: correlation with instantaneous return after release of negative pressure/maximum suction height as the objective variable (4) one or more stratum corneum parameters selected from the group consisting of standard deviation of the minimum brightness value of one cell in the image and standard deviation of the brightness of each pixel of one cell in the image as the image standard deviation as the explanatory variable, R5: correlation with instantaneous return after release of negative pressure/instantaneous deformation during suction as the objective variable (5) one or more stratum corneum parameters selected from the group consisting of single cell area, single cell perimeter, rectangular area circumscribing one cell, and standard deviation of the median brightness value of one cell in the image as the explanatory variable, R6: correlation with deformation over time during suction/instantaneous deformation during suction as the objective variable
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2022122760A JP7806355B2 (en) | 2022-08-01 | 2022-08-01 | Skin viscoelasticity index estimation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2022122760A JP7806355B2 (en) | 2022-08-01 | 2022-08-01 | Skin viscoelasticity index estimation method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| JP2024019957A JP2024019957A (en) | 2024-02-14 |
| JP7806355B2 true JP7806355B2 (en) | 2026-01-27 |
Family
ID=89853860
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| JP2022122760A Active JP7806355B2 (en) | 2022-08-01 | 2022-08-01 | Skin viscoelasticity index estimation method |
Country Status (1)
| Country | Link |
|---|---|
| JP (1) | JP7806355B2 (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2005348991A (en) | 2004-06-11 | 2005-12-22 | Pola Chem Ind Inc | Skin discriminating method, method for selecting cosmetic by the discriminating method, and method for presenting cosmetic using the methods |
| JP2006095223A (en) | 2004-09-30 | 2006-04-13 | Pola Chem Ind Inc | Horny layer discrimination |
| JP2007263655A (en) | 2006-03-28 | 2007-10-11 | Pola Chem Ind Inc | Skin judgement method by means of congo red |
| JP2007309805A (en) | 2006-05-19 | 2007-11-29 | Pola Chem Ind Inc | Automatic differentiating device for corneocyte |
| CN105699654A (en) | 2014-12-15 | 2016-06-22 | 株式会社芳珂 | Purpose for expression amount of NGAL of skin cuticula |
| JP2022018132A (en) | 2020-07-15 | 2022-01-27 | 株式会社アイム | Support method and support program for supporting evaluation of skin condition, and learned model generating method |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH09308619A (en) * | 1996-05-23 | 1997-12-02 | Kao Corp | Method and device for skin surface analysis |
| JP4460988B2 (en) * | 2004-09-30 | 2010-05-12 | ポーラ化成工業株式会社 | How to differentiate stratum corneum |
| JP6598049B2 (en) * | 2012-04-09 | 2019-10-30 | ポーラ化成工業株式会社 | Skin condition discrimination method based on nipple structure analysis |
| JP7509492B2 (en) * | 2020-08-12 | 2024-07-02 | ポーラ化成工業株式会社 | Output method, output system, and output program |
-
2022
- 2022-08-01 JP JP2022122760A patent/JP7806355B2/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2005348991A (en) | 2004-06-11 | 2005-12-22 | Pola Chem Ind Inc | Skin discriminating method, method for selecting cosmetic by the discriminating method, and method for presenting cosmetic using the methods |
| JP2006095223A (en) | 2004-09-30 | 2006-04-13 | Pola Chem Ind Inc | Horny layer discrimination |
| JP2007263655A (en) | 2006-03-28 | 2007-10-11 | Pola Chem Ind Inc | Skin judgement method by means of congo red |
| JP2007309805A (en) | 2006-05-19 | 2007-11-29 | Pola Chem Ind Inc | Automatic differentiating device for corneocyte |
| CN105699654A (en) | 2014-12-15 | 2016-06-22 | 株式会社芳珂 | Purpose for expression amount of NGAL of skin cuticula |
| JP2016114591A (en) | 2014-12-15 | 2016-06-23 | 株式会社ファンケル | Skin viscoelasticity marker and use of the same |
| JP2022018132A (en) | 2020-07-15 | 2022-01-27 | 株式会社アイム | Support method and support program for supporting evaluation of skin condition, and learned model generating method |
Also Published As
| Publication number | Publication date |
|---|---|
| JP2024019957A (en) | 2024-02-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP5263991B2 (en) | Skin texture and / or wrinkle discrimination method and discrimination device, skin discrimination program, and method for selecting an external preparation for skin | |
| JP6725179B2 (en) | Discrimination method of skin condition based on nipple structure analysis | |
| JP2007252891A (en) | Estimation method of evaluation value by visual recognition of beauty of skin | |
| JP7639201B2 (en) | Method for predicting hair and facial characteristics using scalp stratum corneum cell characteristics | |
| JP2018099184A (en) | Diagnostic method of skin condition and screening method, based on bacteria index | |
| JP7259169B2 (en) | Differentiation method of skin condition | |
| JP2012161371A (en) | Skin texture evaluation method by checking viscoelasticity on skin surface | |
| JP2015198909A (en) | Skin classification method, cosmetic recommendation method and skin classification card | |
| JP6471559B2 (en) | Diagnostic device, image processing method, image processing system, and program for the diagnostic device | |
| JP7806355B2 (en) | Skin viscoelasticity index estimation method | |
| CN115516503A (en) | Method for identifying dendritic pores | |
| Vergilio et al. | Enhancing skin aging parameter assessment in clinical trials: AI-Driven analysis of ultrasound images | |
| Kerr et al. | Measurement of capillary refill time (CRT) in healthy subjects using a robotic hand | |
| JP7852179B2 (en) | Skin physiological index estimation method | |
| JP6207926B2 (en) | Body function estimation method | |
| JP2019216652A (en) | Diagnosis method of skin condition based on bacterial amount of specific environment-derived bacteria correlated with specific skin condition, estimation method of diversity of indigenous skin flora, estimation method of bacteria amount of specific environment-derived bacteria, method for analyzing attributes of bacteria that are correlated with specific skin condition, and method for screening substance having skin condition improving effect or cosmetic method | |
| Visscher | Imaging skin: past, present and future perspectives | |
| JP6492297B2 (en) | Method for estimating skin internal structure | |
| JP7442960B2 (en) | How to differentiate skin conditions | |
| JP7853060B2 (en) | Method for estimating skin condition features | |
| JP5096988B2 (en) | Method for measuring the delamination state of the stratum corneum | |
| CN114052666A (en) | A method for assessing the age of facial skin and its application in cosmetic efficacy testing | |
| JP7749366B2 (en) | How to evaluate chapped lips | |
| JP2010088654A (en) | Method of discrimination among skin type by skin physiological index | |
| Hani et al. | Digital Image Analysis of Chronic Ulcers Tissues |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| A621 | Written request for application examination |
Free format text: JAPANESE INTERMEDIATE CODE: A621 Effective date: 20250314 |
|
| A977 | Report on retrieval |
Free format text: JAPANESE INTERMEDIATE CODE: A971007 Effective date: 20251118 |
|
| TRDD | Decision of grant or rejection written | ||
| A01 | Written decision to grant a patent or to grant a registration (utility model) |
Free format text: JAPANESE INTERMEDIATE CODE: A01 Effective date: 20251216 |
|
| A61 | First payment of annual fees (during grant procedure) |
Free format text: JAPANESE INTERMEDIATE CODE: A61 Effective date: 20251217 |
|
| R150 | Certificate of patent or registration of utility model |
Ref document number: 7806355 Country of ref document: JP Free format text: JAPANESE INTERMEDIATE CODE: R150 |