JP7426068B2 - Mibyou identification system - Google Patents
Mibyou identification system Download PDFInfo
- Publication number
- JP7426068B2 JP7426068B2 JP2019202778A JP2019202778A JP7426068B2 JP 7426068 B2 JP7426068 B2 JP 7426068B2 JP 2019202778 A JP2019202778 A JP 2019202778A JP 2019202778 A JP2019202778 A JP 2019202778A JP 7426068 B2 JP7426068 B2 JP 7426068B2
- Authority
- JP
- Japan
- Prior art keywords
- data
- sample
- actual measurement
- subject
- measuring
- 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
- 238000005259 measurement Methods 0.000 claims description 49
- 238000013528 artificial neural network Methods 0.000 claims description 28
- 238000011156 evaluation Methods 0.000 claims description 22
- 230000037237 body shape Effects 0.000 claims description 11
- 244000005700 microbiome Species 0.000 claims description 10
- 229940124595 oriental medicine Drugs 0.000 claims description 10
- 201000010099 disease Diseases 0.000 claims description 6
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 claims description 6
- 108090000623 proteins and genes Proteins 0.000 claims description 4
- 238000012360 testing method Methods 0.000 description 16
- 208000024891 symptom Diseases 0.000 description 10
- 108020004414 DNA Proteins 0.000 description 7
- 238000000034 method Methods 0.000 description 7
- 238000010586 diagram Methods 0.000 description 5
- 230000005856 abnormality Effects 0.000 description 4
- 210000004369 blood Anatomy 0.000 description 4
- 239000008280 blood Substances 0.000 description 4
- 230000007812 deficiency Effects 0.000 description 4
- 230000000968 intestinal effect Effects 0.000 description 4
- 230000035772 mutation Effects 0.000 description 4
- 241000894006 Bacteria Species 0.000 description 3
- 206010013786 Dry skin Diseases 0.000 description 3
- 206010037660 Pyrexia Diseases 0.000 description 3
- 239000003814 drug Substances 0.000 description 3
- 230000037336 dry skin Effects 0.000 description 3
- 230000002550 fecal effect Effects 0.000 description 3
- 230000002068 genetic effect Effects 0.000 description 3
- 239000002773 nucleotide Substances 0.000 description 3
- 125000003729 nucleotide group Chemical group 0.000 description 3
- 206010028813 Nausea Diseases 0.000 description 2
- 206010034568 Peripheral coldness Diseases 0.000 description 2
- 206010039792 Seborrhoea Diseases 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 210000001624 hip Anatomy 0.000 description 2
- JVTAAEKCZFNVCJ-UHFFFAOYSA-N lactic acid Chemical compound CC(O)C(O)=O JVTAAEKCZFNVCJ-UHFFFAOYSA-N 0.000 description 2
- 230000008693 nausea Effects 0.000 description 2
- 230000037312 oily skin Effects 0.000 description 2
- 206010000060 Abdominal distension Diseases 0.000 description 1
- 206010003445 Ascites Diseases 0.000 description 1
- 241000186000 Bifidobacterium Species 0.000 description 1
- 208000035985 Body Odor Diseases 0.000 description 1
- 206010006326 Breath odour Diseases 0.000 description 1
- 229920000742 Cotton Polymers 0.000 description 1
- 208000020401 Depressive disease Diseases 0.000 description 1
- 208000003556 Dry Eye Syndromes Diseases 0.000 description 1
- 206010013774 Dry eye Diseases 0.000 description 1
- 206010013789 Dry throat Diseases 0.000 description 1
- 206010015137 Eructation Diseases 0.000 description 1
- 206010022998 Irritability Diseases 0.000 description 1
- 208000037093 Menstruation Disturbances Diseases 0.000 description 1
- 206010027339 Menstruation irregular Diseases 0.000 description 1
- 206010049816 Muscle tightness Diseases 0.000 description 1
- 108091028043 Nucleic acid sequence Proteins 0.000 description 1
- 206010033557 Palpitations Diseases 0.000 description 1
- 208000003251 Pruritus Diseases 0.000 description 1
- 208000037656 Respiratory Sounds Diseases 0.000 description 1
- 206010040904 Skin odour abnormal Diseases 0.000 description 1
- 206010042674 Swelling Diseases 0.000 description 1
- 210000003423 ankle Anatomy 0.000 description 1
- 230000004596 appetite loss Effects 0.000 description 1
- 210000000617 arm Anatomy 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000001580 bacterial effect Effects 0.000 description 1
- 208000027687 belching Diseases 0.000 description 1
- 230000000740 bleeding effect Effects 0.000 description 1
- 244000309466 calf Species 0.000 description 1
- 208000002173 dizziness Diseases 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 206010016256 fatigue Diseases 0.000 description 1
- 206010016766 flatulence Diseases 0.000 description 1
- 210000002683 foot Anatomy 0.000 description 1
- 235000014655 lactic acid Nutrition 0.000 description 1
- 239000004310 lactic acid Substances 0.000 description 1
- 235000021266 loss of appetite Nutrition 0.000 description 1
- 208000019017 loss of appetite Diseases 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 201000003102 mental depression Diseases 0.000 description 1
- 230000000813 microbial effect Effects 0.000 description 1
- 210000002200 mouth mucosa Anatomy 0.000 description 1
- 210000000056 organ Anatomy 0.000 description 1
- 238000002559 palpation Methods 0.000 description 1
- 210000004197 pelvis Anatomy 0.000 description 1
- 102000054765 polymorphisms of proteins Human genes 0.000 description 1
- 239000011148 porous material Substances 0.000 description 1
- 208000024584 respiratory abnormality Diseases 0.000 description 1
- 210000003296 saliva Anatomy 0.000 description 1
- 230000037307 sensitive skin Effects 0.000 description 1
- 230000036548 skin texture Effects 0.000 description 1
- 241000894007 species Species 0.000 description 1
- 230000002269 spontaneous effect Effects 0.000 description 1
- 230000004083 survival effect Effects 0.000 description 1
- 230000035900 sweating Effects 0.000 description 1
- 230000008961 swelling Effects 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- 210000000689 upper leg Anatomy 0.000 description 1
- 210000001835 viscera Anatomy 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Landscapes
- Medical Treatment And Welfare Office Work (AREA)
- Investigating Or Analysing Biological Materials (AREA)
- Measuring Or Testing Involving Enzymes Or Micro-Organisms (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
- Measuring And Recording Apparatus For Diagnosis (AREA)
Description
本発明は、東洋医学的未病の状態を科学的に識別するシステムに関する。 The present invention relates to a system for scientifically identifying a pre-symptomatic state according to Oriental medicine.
同じ「未病」でも「西洋医学的未病」と「東洋医学的未病」とでは意味が違っており、西洋医学では「自覚症状はないが、検査で異常が確認される状態」を未病と言い、東洋医学では「自覚症状はあるが、検査で異常がない状態」を未病と言っている。
西洋医学では検査で異常が確認できなければ治療は行わない。一方、東洋医学では、未病も治療対象であり、本格的な病気になる前に正常に戻すことを治療原則としている。
東洋医学では、人体を構成し生存を維持する基本要素として「気」「血」「津液」が存在し、それらの不調が病をもたらすと考えられている。未病は、健康と病との中間の状態に当たる。
Even though it is the same word, ``Mibyou'', ``Mibyou according to Western medicine'' and ``Mibyo according to Oriental medicine'' have different meanings.In Western medicine, ``Mibyo'' is defined as ``a condition in which there are no subjective symptoms but abnormalities are confirmed by tests''. In Oriental medicine, ``a condition in which there are subjective symptoms but no abnormalities are detected in examinations'' is called ``non-illness.''
Western medicine does not provide treatment unless an abnormality is confirmed through testing. On the other hand, in Oriental medicine, even pre-symptomatic conditions are treated, and the principle of treatment is to restore normality before a full-scale disease develops.
In Oriental medicine, it is believed that ``Qi'', ``Blood'', and ``Tsui'' are the basic elements that make up the human body and maintain its survival, and that imbalances in these elements lead to disease. Mibyou is a state intermediate between health and disease.
東洋医学の診察では、患者の動作・状態等を観察する「望診」、患者の声・呼吸音等を聞き、口臭・体臭等を嗅ぐ「聞診」、患者の自覚症状・病歴・生活習慣等を確認する「問診」、そして、患者の体に触れて筋肉の緊張度や内臓の状態等を診る「切診」の四診が行われ、それらの情報を総合して「証」、即ち、患者の体質を読取り、その体に合った治療法が採られる。
こうした診察には時間が掛かる。また、多くの経験を積んだ熟練者でなければ適切な証を立てることが難しい。
Oriental medicine examinations include ``visual examination'' to observe the patient's movements and condition, ``auscultation'' to listen to the patient's voice and breathing sounds, and sniff for bad breath and body odor, and the patient's subjective symptoms, medical history, lifestyle habits, etc. Four examinations are carried out: ``interview'' to confirm the patient's body, and ``cut examination'' to touch the patient's body to check the degree of muscle tension and the state of internal organs. The patient's constitution is read and a treatment method suited to the patient's body is adopted.
These examinations take time. Additionally, it is difficult to provide appropriate evidence unless you are an expert with extensive experience.
東洋医学の四診を支援するため、種々の器具が開発さている。下記特許文献１には、望診の内の舌診を容易にする器官画像撮影装置が開示されている。また、下記特許文献２には、切診を支援する触診支援装置が開示されている。 Various instruments have been developed to support the four examinations of Oriental medicine. Patent Document 1 listed below discloses an organ image photographing device that facilitates tongue examination during visual examination. Further, Patent Document 2 listed below discloses a palpation support device that supports cutting.
しかし、望診や切診等を機械的に支援する手段が提供されたとしても、四診の情報を総合して患者の未病状態を読取ることは容易ではない。 However, even if a means to mechanically support physical examination, surgical examination, etc. is provided, it is not easy to understand the patient's pre-symptomatic state by integrating information from four examinations.
本発明は、こうした事情を考慮して創案したものであり、東洋医学的未病（以下、単に「未病」と言う。）の状態を科学的に識別できるシステムを提供することを目的としている。 The present invention was devised in consideration of these circumstances, and aims to provide a system that can scientifically identify the state of Mibyo (hereinafter simply referred to as "Mibyou") according to Oriental medicine. .
本発明は、被験者の未病状態を識別する識別システムであって、被験者及び複数のサンプル提供者の体質に関係するデータを実測する実測手段と、サンプル提供者の未病状態を評価する評価データとこのサンプル提供者の実測手段で実測された実測データとを含むサンプルデータの複数個を保持するデータベースと、データベースに保持されたサンプルデータの実測データと評価データとの対応関係から、被験者の実測手段で実測された実測データに基づいて被験者の未病状態を識別する識別部と、を備える。
The present invention is an identification system for identifying a pre-symptomatic state of a subject, and includes an actual measuring means for actually measuring data related to the physical constitution of the subject and a plurality of sample donors, and evaluation data for evaluating the pre-symptomatic state of the sample donors. A database holds a plurality of pieces of sample data including actual measurement data actually measured by the sample provider's actual measurement means, and the actual measurement of the subject from the correspondence between the actual measurement data and evaluation data of the sample data held in the database. An identification unit that identifies the pre-symptomatic state of the subject based on actual measurement data measured by the means .
そして、識別部は、入力層、中間層及び出力層を有するニューラルネットワークを用い、サンプルデータに含まれる実測データを入力層に入力して、そのサンプルデータに含まれる評価データを出力層から出力する学習を行う。そして、学習済みのニューラルネットワークの入力層に被験者の実測データを入力し、被験者の未病状態を表す評価データをニューラルネットワークの出力層から得る。
学習済みニューラルネットワークは、人工知能（ＡＩ）の機械学習を通じて構築することができ、このニューラルネットワークを用いて被験者の未病状態を短時間に識別することができる。
Then, the identification unit uses a neural network having an input layer, a middle layer, and an output layer, inputs the measured data included in the sample data to the input layer, and outputs the evaluation data included in the sample data from the output layer. Learn. Then, the measured data of the subject is input to the input layer of the trained neural network, and evaluation data representing the pre-symptomatic state of the subject is obtained from the output layer of the neural network.
A trained neural network can be constructed through machine learning using artificial intelligence (AI), and can be used to quickly identify the pre-symptomatic state of a subject.
また、本発明の未病識別システムの実測手段には、皮膚の状態を実測する皮膚測定手段、体型を実測する体型計測手段、遺伝子を測定するＤＮＡ測定手段、及び、体内のマイクロバイオームを測定するマイクロバイオーム測定手段の少なくとも一つが含まれる。
これらの実測手段を用いて、被験者及びサンプル提供者の実測データが測定される。
Furthermore, the actual measurement means of the pre-symptomatic identification system of the present invention includes a skin measurement means for actually measuring the condition of the skin, a body shape measurement means for actually measuring the body shape, a DNA measurement means for measuring genes, and a measurement means for measuring the microbiome in the body. At least one microbiome measuring means is included.
Actual measurement data of the subject and the sample provider are measured using these measurement means.
また、本発明の未病識別システムで用いるニューラルネットワークは、多層の中間層を有していても良い。
多層の中間層を有するニューラルネットワークを用いて深層学習（ディープラーニング）を行い、被験者の未病状態を識別することができる。
Moreover, the neural network used in the pre-symptomatic identification system of the present invention may have multiple intermediate layers.
By performing deep learning using a neural network with multiple intermediate layers, it is possible to identify the pre-symptomatic state of a test subject.
本発明の未病識別システムは、被験者の未病状態を、熟練を必要とせずに、簡便に、且つ、高い精度で科学的に識別することができる。 The pre-symptomatic identification system of the present invention can scientifically identify the pre-symptomatic state of a subject easily and with high accuracy without requiring any skill.
図１は、本発明の実施形態に係る未病識別システムの構成を模式的に示している。
このシステムは、システム構築に協力するサンプル提供者から得られたサンプルデータ１０を多数保存するサンプルデータベース９と、サンプルデータを教師データに利用してニューラルネットワーク２１の学習を行い、学習済みのニューラルネットワーク２１を用いて被験者の未病状態を識別する識別部２０とを備えている。
FIG. 1 schematically shows the configuration of a pre-symptomatic identification system according to an embodiment of the present invention.
This system includes a sample database 9 that stores a large number of sample data 10 obtained from sample providers who cooperate with system construction, and a neural network 21 that uses the sample data as training data to train a neural network 21 that has already been trained. The identification unit 20 identifies the pre-symptomatic state of the subject using 21.
サンプルデータ１０は、サンプル提供者の未病状態を評価した未病状態評価データ１１と、実測手段１２０で測定したサンプル提供者の実測データ１２とから成る。
サンプル提供者の未病状態評価データ１１は、サンプル提供者の未病状態を判定した医師の医師評価データ１１１である。
The sample data 10 includes pre-symptomatic state evaluation data 11 that evaluates the pre-symptomatic state of the sample donor, and actual measurement data 12 of the sample donor measured by the actual measuring means 120.
The sample donor's pre-symptomatic state evaluation data 11 is the physician evaluation data 111 of the doctor who determined the sample donor's pre-symptomatic state.
例えば、医師は、サンプル提供者から、普通にしていても自然と汗が出る自汗や、倦怠感、食欲不振、手足の冷え等の症状を読み取ると「気虚」と判定する。精神の抑うつ、げっぷ、おなら、腹部の張り等の症状を読み取ると「気滞」と判定する。イライラ感、呼吸器の異常、吐き気の症状を読み取ると「気逆」と判定する。目のかすみや乾き、動悸、月経不順の症状を読み取ると「血虚」と判定する。発熱、各部の出血、乾燥、皮膚のかゆみの症状を読み取ると「血熱」と判定する。肌や髪の乾燥、喉、鼻、口の渇き、関節の異常の症状を読み取ると「内燥」と判定する。また、めまい、吐き気、腹水が溜る状態、むくみの症状を読み取ると「内湿」と判定する。 For example, a doctor determines that a sample donor has ``qi deficiency'' when he or she detects symptoms such as spontaneous sweating even when doing normal activities, fatigue, loss of appetite, and cold hands and feet. If symptoms such as mental depression, belching, flatulence, and abdominal distension are detected, it is determined that the patient has ``qi stagnation''. If symptoms such as irritability, respiratory abnormalities, and nausea are detected, it is determined that the patient is suffering from ``Ki Gyaku''. When symptoms such as blurred and dry eyes, palpitations, and irregular menstruation are detected, it is determined that the patient has ``blood deficiency.'' When the symptoms of fever, bleeding in various parts, dryness, and itchy skin are read, it is determined to be a "blood fever." If symptoms such as dry skin and hair, dry throat, nose, and mouth, and abnormalities in the joints are detected, it is determined to be ``internal dryness''. In addition, if symptoms such as dizziness, nausea, ascites, and swelling are detected, it is determined that the patient has internal humidity.
サンプル提供者の実測データ１２を測定する実測手段１２０は、サンプル提供者の肌の状態を測定する皮膚測定手段１２１や、サンプル提供者の体型を測定する体型計測手段１２２、サンプル提供者の遺伝子を測定するＤＮＡ測定手段１２３、サンプル提供者の体内に常在する微生物叢を測定するマイクロバイオーム測定手段１２４等である。 The actual measuring means 120 for measuring the actual measured data 12 of the sample donor includes a skin measuring means 121 for measuring the skin condition of the sample donor, a body shape measuring means 122 for measuring the body shape of the sample donor, and a measuring means 120 for measuring the sample donor's genes. These include a DNA measuring means 123 for measuring, a microbiome measuring means 124 for measuring the microbial flora resident in the body of the sample donor, and the like.
サンプル提供者の実測データ１２を得るための実測手段１２０は、被験者の実測データ２２を得るときにも使用される。
皮膚測定手段１２１は、脂性肌／乾燥肌の関係を表すスコア（脂性肌の程度が高いほど値が大きく、乾燥肌の程度が高いほど値が小さい）や、耐性肌／敏感肌の関係を表すスコア、色素性肌／非色素性肌の関係を表すスコア、シワ肌／ハリ肌の関係を表すスコア、スポット（しみ、斑点、ほくろ）のスコア、肌のキメの状態、毛穴の大きさ、ＵＶスポットの発生状態、赤いシミの発生状態等を測定する。
皮膚測定手段１２１として、例えばＶＩＳＩＯ（皮膚計測機）などを使用する。
The actual measurement means 120 for obtaining actual measurement data 12 of the sample provider is also used when obtaining actual measurement data 22 of the subject.
The skin measuring means 121 is a score representing the relationship between oily skin/dry skin (the higher the degree of oily skin, the larger the value, and the higher the degree of dry skin, the lower the value), and the relationship between resistant skin/sensitive skin. score, score representing the relationship between pigmented skin/non-pigmented skin, score representing the relationship between wrinkled skin/firm skin, score for spots (spots, spots, moles), skin texture condition, pore size, UV Measure the appearance of spots, red stains, etc.
As the skin measuring means 121, for example, VISIO (skin measuring instrument) is used.
体型計測手段１２２は、身長、バスト、二の腕、ウエスト、ヒップ、太もも、ふくらはぎ、足首、肩幅、骨盤等の各部位の寸法を測定する。
体型計測手段１２２として、例えば図４に示す３次元測定装置を使用する。図４（ａ）の装置は、垂直線上に固定された４台の距離センサ付きカメラＳＤの前で人物Ｓを乗せたステージＴが回転するように構成されている。４台のカメラＳＤは、ステージＴが所定角度回転するごとに人物Ｓを撮影する。ステージＴが１回転する間にカメラＳＤで撮影された複数の画像から被写体の複数の特徴点が抽出され、特徴点の位置関係が計算されて、人物Ｓの各種の寸法が測定される。図４（ｂ）の装置は、人物Ｓが載る固定ステージＴの周りを４台の距離センサ付きカメラＳＤが回転する。また、図４（ｃ）の装置は、垂直線上に固定された４台の距離センサ付きカメラＳＤを１２０度の角度を空けて３列配置し、人物を回転させなくても人物の全周が撮影できるように構成されている。
The body shape measuring means 122 measures the dimensions of various parts such as height, bust, upper arms, waist, hips, thighs, calves, ankles, shoulder width, and pelvis.
As the body shape measuring means 122, for example, a three-dimensional measuring device shown in FIG. 4 is used. The apparatus shown in FIG. 4A is configured such that a stage T on which a person S is placed rotates in front of four cameras SD with distance sensors fixed on a vertical line. The four cameras SD photograph the person S every time the stage T rotates by a predetermined angle. A plurality of feature points of the subject are extracted from a plurality of images taken by the camera SD during one rotation of the stage T, the positional relationships of the feature points are calculated, and various dimensions of the person S are measured. In the apparatus shown in FIG. 4(b), four cameras SD with distance sensors rotate around a fixed stage T on which a person S is placed. In addition, the device shown in Fig. 4(c) has four cameras SD with distance sensors fixed on a vertical line arranged in three rows at an angle of 120 degrees, so that the entire circumference of the person can be measured without rotating the person. It is configured so that it can be photographed.
ＤＮＡ測定手段１２３は、唾液、血液、綿棒で採取した口腔粘膜等のサンプルから遺伝子情報を取得する。近年、ＤＮＡの解析技術が急速に進歩したことにより、低価格で遺伝子検査を引受ける検査機関が増えている。こうした検査機関にサンプルを送ることで遺伝子情報が得られる。検査機関は、サンプルから一塩基多型（ＳＮＰ：塩基配列中の一塩基が変異しているヒトの割合が一定率以上である塩基配列）における変異の有無を解析し、解析結果を送り返してくる。 The DNA measuring means 123 acquires genetic information from samples such as saliva, blood, and oral mucosa collected with a cotton swab. In recent years, with rapid advances in DNA analysis technology, an increasing number of testing institutions are accepting genetic tests at low prices. Genetic information can be obtained by sending samples to these testing institutions. The testing agency analyzes the sample for the presence or absence of mutations in single nucleotide polymorphisms (SNPs: nucleotide sequences in which the proportion of humans with a single nucleotide mutation in the nucleotide sequence exceeds a certain level) and sends back the analysis results. .
マイクロバイオーム測定手段１２４は、糞便検体から腸内細菌叢の情報を取得する。マイクロバイオームの検査機関も最近増えており、こうした検査機関に糞便検体を送ることで腸内細菌叢の情報が得られる。検査機関は、糞便検体を検査して腸内細菌叢の菌種数や菌種の割合、また、ビフィズス菌、乳酸菌、痩せ菌等の腸内細菌がどの程度住み着いているかの情報等を伝えてくれる。 The microbiome measuring means 124 acquires information on intestinal flora from the fecal sample. The number of microbiome testing institutions has increased recently, and by sending fecal samples to these testing institutions, information on the intestinal flora can be obtained. Testing institutions test fecal samples and provide information on the number and proportion of bacterial species in the intestinal flora, as well as the extent to which intestinal bacteria such as bifidobacteria, lactic acid bacteria, and bacterium are present. I'll give it to you.
識別部２０は、ニューラルネットワーク２１を用いて被験者の未病状態を識別する。
ニューラルネットワーク２１は、図２に示すように、入力層、中間層、及び、出力層を有している。
ニューラルネットワークの入力層の各ノードに値ｘｉ（ｉ＝１，２，３，４・・・）を与えると、中間層のノードｊ（ｊ＝１，２，３，４・・・）には、入力層の各ノードの値ｘｉに重みｗ（１）ｊｉを掛けた値の加算値ｖｊが入力する（ｖｊ＝Σｘi・ｗ(1)ji）。そして、出力層のノードｋ（ｋ＝１，２、３・・・）には、中間層の各ノードの値ｖｊに重みｗ(２)ｋｊを掛けた値の加算値ｙｋが入力する（ｙｋ＝Σｖｊ・ｗ(２)ｋｊ）。
識別部２０は、サンプルデータを教師データに用いて、サンプルデータの実測データ１２を入力層の各ノードに与えたとき、出力層から医師評価データ１１１が出力されるように、重みｗ（１）ｊｉ及びｗ(２)ｋｊの値を調整する。
The identification unit 20 uses the neural network 21 to identify the pre-symptomatic state of the subject.
As shown in FIG. 2, the neural network 21 has an input layer, a middle layer, and an output layer.
When a value xi (i=1, 2, 3, 4...) is given to each node in the input layer of the neural network, the node j (j=1, 2, 3, 4...) in the middle layer is given the value xi (i=1, 2, 3, 4...) , the sum vj of the value xi of each node in the input layer multiplied by the weight w(1)ji is input (vj=Σxi·w(1)ji). Then, the sum value yk of the value vj of each node in the intermediate layer multiplied by the weight w(2)kj is input to the node k (k=1, 2, 3...) of the output layer (yk =Σvj・w(2)kj).
The identification unit 20 uses the sample data as teacher data and sets the weight w(1) so that when the actual measurement data 12 of the sample data is given to each node of the input layer, the doctor evaluation data 111 is output from the output layer. Adjust the values of ji and w(2)kj.
この調整をやり易くするため、実測データ１２に対して、最大値ｘｍａｘを１（又は１より僅かに小さい値）、最小値ｘｍｉｎを０（又は０より僅かに大きい値）とする正規化を行う。この正規化した値Ｙは、実測データをＸ、実測データの最大値をＸｍａｘ、実測データの最小値をＸｍｉｎとするとき、次式により求めることができる。
Ｙ＝（Ｘ－Ｘｍｉｎ）／（Ｘｍａｘ－Ｘｍｉｎ）
In order to make this adjustment easier, the measured data 12 is normalized so that the maximum value xmax is 1 (or a value slightly smaller than 1) and the minimum value xmin is 0 (or a value slightly larger than 0). . This normalized value Y can be determined by the following equation, where the actual measurement data is X, the maximum value of the actual measurement data is Xmax, and the minimum value of the actual measurement data is Xmin.
Y=(X-Xmin)/(Xmax-Xmin)
皮膚測定手段１２１で計測した各実測データは、数値化して正規化し、ニューラルネットワーク２１の入力層の個々のノードに入力する。数値化できない（有るか無いかを検出している）実測データは、有る場合にｘｍａｘを、無い場合にｘｍｉｎをノードに入力する。
体型計測手段１２２で計測した各実測データは、そのまま正規化しても良いが、各実測データと一つの値（例えば身長）との比を計算し、その比の値を正規化して入力層の個々のノードに入力するようにしても良い。
ＤＮＡ測定手段１２３の測定結果は、検査した各遺伝子について変異がある場合にｘｍａｘを、無い場合にｘｍｉｎをノードに入力する。
マイクロバイオーム測定手段１２４の測定結果も同様に、数値化できるデータは正規化してノードに入力し、数値化できないデータは、ｘｍａｘまたはｘｍｉｎをノードに入力する。
Each actual measurement data measured by the skin measuring means 121 is digitized and normalized, and input to each node of the input layer of the neural network 21. For actual measurement data that cannot be quantified (presence or absence is detected), input xmax into the node if it exists, and xmin if it does not exist.
Each actual measurement data measured by the body shape measurement means 122 may be normalized as is, but the ratio between each actual measurement data and one value (for example, height) is calculated, and the value of the ratio is normalized to determine the individual values of the input layer. You may also input it to the node.
As for the measurement results of the DNA measuring means 123, xmax is inputted into the node if there is a mutation in each gene tested, and xmin is inputted if there is no mutation.
Similarly, regarding the measurement results of the microbiome measuring means 124, data that can be quantified is normalized and input to the node, and data that cannot be quantified is inputted as xmax or xmin to the node.
また、出力層から出力される医師評価データ１１１にはラベルを付し、ラベルの情報が出力層のノードから出力されるようにする。例えば、「気虚」のラベルを「０００」、「気滞」のラベルを「００１」、「気逆」のラベルを「０１０」、「血虚」のラベルを「０１１」、「血熱」のラベルを「１００」、「内燥」のラベルを「１０１」、また「内湿」のラベルを「１１０」に設定し、出力層の１番目のノードからラベルの１桁目の数字が出力され、２番目のノードからラベルの２桁目の数字が出力され、３番目のノードからラベルの３桁目の数字が出力されるようにすることで出力層からの医師評価データの出力が可能になる。 Further, a label is attached to the doctor evaluation data 111 outputted from the output layer, so that the label information is outputted from the node of the output layer. For example, the label for "qi deficiency" is "000", the label for "qi stagnation" is "001", the label for "qi reversal" is "010", the label for "blood deficiency" is "011", and "blood fever". Set the label to "100", the label for "Inner dryness" to "101", and the label for "Inner humidity" to "110", and the first digit of the label is output from the first node of the output layer. The second node outputs the second digit of the label, and the third node outputs the third digit of the label, making it possible to output doctor evaluation data from the output layer. become.
識別部２０は、ニューラルネットワーク２１の入力層のノードに各サンプル提供者の実測データ１２を入力したとき、出力層のノードから、そのサンプル提供者に対する医師評価データ１１１が出力されるように重みｗ（１）ｊｉ及びｗ(２)ｋｊの値を設定する。
この操作を効率化するため、ニューラルネットワークでは「誤差逆伝搬法（バックプロパゲーション）」と言う方法が広く行われている。この方法では、出力側から順番に答えとどのくらい違っているか比較して誤差を算出し、誤差を埋めるようにして重み付けの値を導き出す。
The identification unit 20 sets the weight w so that when the measured data 12 of each sample provider is input to the nodes of the input layer of the neural network 21, the doctor evaluation data 111 for that sample provider is output from the node of the output layer. (1) Set the values of ji and w(2) kj.
To make this operation more efficient, a method called ``error backpropagation'' is widely used in neural networks. In this method, the error is calculated by comparing the difference from the answer in order from the output side, and the weighting value is derived by filling in the error.
こうした学習を経て、ニューラルネットワーク２１の入力層のノードに各サンプル提供者の実測データ１２を入力したとき、出力層のノードから、そのサンプル提供者に対する医師評価データ１１１が出力されるようになった段階で、識別部２０は、入力層のノードに入力された実測データ１２の内、中間層の全てのノードとの間の重みｗ（１）ｊｉが０の実測データを探し、該当する実測データを未病識別システムから除外する。
こうすることで、入力層の不必要なノードを整理することができる。
Through such learning, when the measured data 12 of each sample provider is input to the input layer node of the neural network 21, the doctor evaluation data 111 for that sample provider is output from the output layer node. In the step, the identification unit 20 searches for actual measured data whose weight w(1)ji between all the nodes of the intermediate layer is 0 among the actual measured data 12 input to the nodes of the input layer, and selects the corresponding actual measured data. be excluded from the Mibyou identification system.
By doing this, unnecessary nodes in the input layer can be sorted out.
識別部２０は、実測手段１２０で実測された被験者の実測データ２２が入力すると、それを学習済みのニューラルネットワーク２１の入力層の各ノードに入力する。そうすると、学習済みのニューラルネットワーク２１の出力層からは、被験者の未病状態を表す医師評価データが出力される。 When the actual measurement data 22 of the subject measured by the actual measurement means 120 is input, the identification unit 20 inputs it to each node of the input layer of the trained neural network 21 . Then, the output layer of the trained neural network 21 outputs doctor evaluation data representing the pre-symptomatic state of the subject.
図３のフロー図は、この未病識別システムを実現する手順を示している。
先ず、医師評価データと実測データを含むサンプルデータを集める（Ｓｔ．１）。
次いで、サンプルデータを用いてニューラルネットワークの学習を行う（Ｓｔ．２）。
次いで、学習済みニューラルネットワークを用いて、被験者の実測データから被験者の未病状態を表す評価データを得る（Ｓｔ．３）。
Ｓｔ．３で得られた被験者の評価データ及び被験者の実測データは、以降のサンプルデータに含めることができる。そうしたサンプルデータが増えることにより、このシステムを維持するための医師の負担が減少する。
The flow diagram of FIG. 3 shows the procedure for realizing this pre-symptomatic identification system.
First, sample data including doctor evaluation data and actual measurement data is collected (St. 1).
Next, the neural network is trained using the sample data (St. 2).
Next, using the trained neural network, evaluation data representing the pre-symptomatic state of the subject is obtained from the measured data of the subject (St. 3).
St. The test subject's evaluation data and the test subject's actual measurement data obtained in step 3 can be included in subsequent sample data. The increase in such sample data will reduce the burden on physicians to maintain this system.
ここでは、ニューラルネットワークの中間層が１層の場合について説明したが、ニューラルネットワークの中間層は多層であっても良い。多層の中間層を有するニューラルネットワークを用いて深層学習（ディープラーニング）を行うことにより、高精度の未病状態の識別が可能になる。
また、ここでは、実測手段として、皮膚測定手段、体型計測手段、ＤＮＡ測定手段、及び、マイクロバイオーム測定手段を挙げたが、そのすべてを使わなくても良いし、また、その他の実測手段を使っても良い。
また、ここでは、ニューラルネットワークを用いて被験者の未病状態を識別する場合について説明したが、サンプルデータの実測データと被験者の実測データとの対応関係を別の手法で求めて被験者の未病状態を識別するようにしても良い。
Although the case where the neural network has one intermediate layer has been described here, the neural network may have multiple intermediate layers. By performing deep learning using a neural network with multiple intermediate layers, it becomes possible to identify pre-symptomatic states with high accuracy.
In addition, although skin measurement means, body shape measurement means, DNA measurement means, and microbiome measurement means are listed here as actual measurement means, it is not necessary to use all of them, and other measurement means may also be used. It's okay.
In addition, here we have explained the case where a neural network is used to identify the test subject's pre-symptomatic state, but the correspondence between the actual measured data of the sample data and the test subject's actual measured data is determined using another method, and the test subject's pre-symptomatic state is determined using a different method. It may also be possible to identify.
本発明の未病識別システムは、未病状態の識別に熟練を必要としないため、各種医療機関等で広く利用することができる。 The pre-symptomatic identification system of the present invention does not require any skill to identify pre-symptomatic states, so it can be widely used in various medical institutions.
９ サンプルデータベース
１０ サンプルデータ
１１ サンプル提供者の未病状態評価データ
１２ サンプル提供者の実測データ
２０ 識別部
２１ ニューラルネットワーク
２２ 被験者の実測データ
１１１ 医師評価データ
１２０ 実測手段
１２１ 皮膚測定手段
１２２ 体型計測手段
１２３ ＤＮＡ測定手段
１２４ マイクロバイオーム測定手段
9 Sample database 10 Sample data 11 Sample donor's pre-symptomatic state evaluation data 12 Sample donor's actual measurement data 20 Identification unit 21 Neural network 22 Actual measurement data of subject 111 Physician evaluation data 120 Actual measurement means 121 Skin measurement means 122 Body shape measurement means 123 DNA measuring means 124 Microbiome measuring means
Claims (3)
前記被験者及び複数のサンプル提供者の体質に関係するデータを実測する実測手段と、
前記サンプル提供者の前記未病状態を評価する評価データと該サンプル提供者の前記実測手段で実測された実測データとを含むサンプルデータの複数個を保持するデータベースと、
前記サンプルデータに含まれる前記実測データと前記評価データとの対応関係から、前記被験者の前記実測手段で実測された実測データに基づいて前記被験者の東洋医学的未病状態を識別する識別部と、
を備え、
前記識別部は、入力層、中間層及び出力層を有するニューラルネットワークを用いて、前記サンプルデータに含まれる前記実測データを前記入力層に入力し、該サンプルデータに含まれる前記評価データを前記出力層から出力する学習を行い、学習済みの前記ニューラルネットワークの前記入力層に前記被験者の実測データを入力して、該被験者の前記未病状態を表す前記評価データを前記出力層から得る、東洋医学的未病識別システム。 An identification system for identifying an oriental medical pre-symptomatic state of a subject,
actual measuring means for actually measuring data related to the physical constitution of the subject and the plurality of sample providers;
a database holding a plurality of sample data including evaluation data for evaluating the pre-symptomatic state of the sample donor and actual measurement data measured by the actual measurement means of the sample donor;
an identification unit that identifies the oriental medicine pre-symptomatic state of the subject based on the actual measurement data of the subject measured by the actual measurement means from the correspondence between the actual measurement data and the evaluation data included in the sample data;
Equipped with
The identification unit uses a neural network having an input layer, a middle layer, and an output layer to input the measured data included in the sample data to the input layer, and output the evaluation data included in the sample data. Oriental medicine that performs learning to output from a layer, inputs actual measurement data of the subject into the input layer of the trained neural network, and obtains the evaluation data representing the pre-symptomatic state of the subject from the output layer. Target disease identification system.
前記実測手段には、皮膚の状態を実測する皮膚測定手段、体型を実測する体型計測手段、遺伝子を測定するＤＮＡ測定手段、及び、体内のマイクロバイオームを測定するマイクロバイオーム測定手段の少なくとも一つが含まれる東洋医学的未病識別システム。 The Oriental medical pre-symptomatic identification system according to claim 1 ,
The actual measuring means includes at least one of a skin measuring means for actually measuring the condition of the skin, a body shape measuring means for actually measuring the body shape, a DNA measuring means for measuring genes, and a microbiome measuring means for measuring the microbiome in the body. Oriental medical disease identification system.
前記ニューラルネットワークが、多層の前記中間層を有している東洋医学的未病識別システム。
The Oriental medical pre-symptomatic identification system according to claim 1 ,
An oriental medical disease identification system, wherein the neural network has multiple intermediate layers.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2019202778A JP7426068B2 (en) | 2019-11-07 | 2019-11-07 | Mibyou identification system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2019202778A JP7426068B2 (en) | 2019-11-07 | 2019-11-07 | Mibyou identification system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| JP2021077048A JP2021077048A (en) | 2021-05-20 |
| JP7426068B2 true JP7426068B2 (en) | 2024-02-01 |
Family
ID=75899860
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| JP2019202778A Active JP7426068B2 (en) | 2019-11-07 | 2019-11-07 | Mibyou identification system |
Country Status (1)
| Country | Link |
|---|---|
| JP (1) | JP7426068B2 (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2012133501A (en) | 2010-12-21 | 2012-07-12 | Akio Sano | Ahead-sick determination index and ahead-sick determination method |
| JP2016143171A (en) | 2015-01-30 | 2016-08-08 | Toto株式会社 | Biological information measurement system |
| WO2019022085A1 (en) | 2017-07-24 | 2019-01-31 | アクシオンリサーチ株式会社 | Assistance system for estimating internal state of system-of-interest |
| JP2019154870A (en) | 2018-03-14 | 2019-09-19 | 株式会社Ｂｈｙ | Diathesis identification system and generation method of database used therefor |
-
2019
- 2019-11-07 JP JP2019202778A patent/JP7426068B2/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2012133501A (en) | 2010-12-21 | 2012-07-12 | Akio Sano | Ahead-sick determination index and ahead-sick determination method |
| JP2016143171A (en) | 2015-01-30 | 2016-08-08 | Toto株式会社 | Biological information measurement system |
| WO2019022085A1 (en) | 2017-07-24 | 2019-01-31 | アクシオンリサーチ株式会社 | Assistance system for estimating internal state of system-of-interest |
| JP2019154870A (en) | 2018-03-14 | 2019-09-19 | 株式会社Ｂｈｙ | Diathesis identification system and generation method of database used therefor |
Also Published As
| Publication number | Publication date |
|---|---|
| JP2021077048A (en) | 2021-05-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Pool et al. | The interexaminer reproducibility of physical examination of the cervical spine | |
| JP5400217B2 (en) | Computer-aided diagnosis system for determining skin composition based on the principles of traditional Chinese medicine (TCM) | |
| US6090044A (en) | System for diagnosing medical conditions using a neural network | |
| Rodrigues-Bigaton et al. | Factor and Rasch analysis of the Fonseca anamnestic index for the diagnosis of myogenous temporomandibular disorder | |
| CN109242849A (en) | Medical image processing method, device, system and storage medium | |
| CN106175726A (en) | Human health status recognition methods based on BP neutral net | |
| JP6914896B2 (en) | Health condition diagnosis system | |
| Vegelin et al. | Influence of knowledge, training and experience of observers on the reliability of anthropometric measurements in children | |
| CN109978873A (en) | A kind of intelligent physical examination system and method based on Chinese medicine image big data | |
| Kopkow et al. | Interrater reliability of the modified scapular assistance test with and without handheld weights | |
| Markova et al. | Effect of a web-based curriculum on primary care practice: basic skin cancer triage trial | |
| Ling et al. | Gait and function in class III obesity | |
| JP7426068B2 (en) | Mibyou identification system | |
| Kvåle et al. | Discriminative validity of the Global Physiotherapy Examination-52 in patients with long-lasting musculoskeletal pain versus healthy persons | |
| Honda et al. | Modeling pathways to affective barriers on colorectal cancer screening among Japanese Americans | |
| Zellner et al. | Differences in respiratory muscle strength measures in well-nourished and malnourished hospitalized patients | |
| CN109065161A (en) | A kind of shared medical detection system and shared medical detection method | |
| Adams et al. | Accuracy of blood pressure measurement and anthropometry among volunteer observers in a large community survey | |
| Natt et al. | The lost art of the history and physical | |
| Voloshan et al. | Development and analysis of diagnostic criteria for creation of an automated computer software for predicting the course and individualizing the treatment of patients with odontogenic maxillary sinusitis | |
| RU2422088C1 (en) | Method of determining person's functional-morphological type | |
| Cervera-Cano et al. | Real-time ultrasound evaluation of CORE muscle activity in a simultaneous contraction in subjects with non-specific low back pain and without low-back pain. Protocol of an observational case-control study | |
| CN102163253A (en) | Health promotion estimating system | |
| UA139751U (en) | METHOD OF COMPREHENSIVE DIAGNOSIS OF THE CONDITION OF THE HUMAN ORGANISM | |
| Sreeraj | Evaluation of Naadi Perception in the Diagnosis of Regional Pathological Conditions |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| A621 | Written request for application examination |
Free format text: JAPANESE INTERMEDIATE CODE: A621 Effective date: 20220719 |
|
| A977 | Report on retrieval |
Free format text: JAPANESE INTERMEDIATE CODE: A971007 Effective date: 20230731 |
|
| A131 | Notification of reasons for refusal |
Free format text: JAPANESE INTERMEDIATE CODE: A131 Effective date: 20230808 |
|
| A521 | Request for written amendment filed |
Free format text: JAPANESE INTERMEDIATE CODE: A523 Effective date: 20231008 |
|
| 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: 20240104 |
|
| A61 | First payment of annual fees (during grant procedure) |
Free format text: JAPANESE INTERMEDIATE CODE: A61 Effective date: 20240115 |
|
| R150 | Certificate of patent or registration of utility model |
Ref document number: 7426068 Country of ref document: JP Free format text: JAPANESE INTERMEDIATE CODE: R150 |