RU2018110753A - A method for predicting the risk of developing skin melanoma based on a neural network analysis of predictors - Google Patents

A method for predicting the risk of developing skin melanoma based on a neural network analysis of predictors Download PDF

Info

Publication number
RU2018110753A
RU2018110753A RU2018110753A RU2018110753A RU2018110753A RU 2018110753 A RU2018110753 A RU 2018110753A RU 2018110753 A RU2018110753 A RU 2018110753A RU 2018110753 A RU2018110753 A RU 2018110753A RU 2018110753 A RU2018110753 A RU 2018110753A
Authority
RU
Russia
Prior art keywords
predictors
risk
neural network
skin
predicting
Prior art date
Application number
RU2018110753A
Other languages
Russian (ru)
Other versions
RU2018110753A3 (en
Inventor
Евгений Юрьевич Неретин
Юрий Леонидович Минаев
Original Assignee
Частное учреждение образовательная организация высшего образования "Медицинский университет "Реавиз"
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Частное учреждение образовательная организация высшего образования "Медицинский университет "Реавиз" filed Critical Частное учреждение образовательная организация высшего образования "Медицинский университет "Реавиз"
Priority to RU2018110753A priority Critical patent/RU2018110753A/en
Publication of RU2018110753A publication Critical patent/RU2018110753A/en
Publication of RU2018110753A3 publication Critical patent/RU2018110753A3/ru

Links

Landscapes

  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Claims (1)

Способ прогнозирования риска развития меланомы кожи на основе нейросетевого анализа предикторов, характеризующийся тем, что обследуют пациентов, определяют значимые факторы - предикторы: место жительства (село или город), социальный статус, наличие онкологических и инфекционных заболеваний в анамнезе, изменение пигментации, появление венчика гиперемии, воспаления, работу в замкнутом помещении более 20 лет, кровоточивость, злокачественные новообразования у членов семьи, возраст на момент исследования, быстрый рост пигментации, беременность, появление дочерних пигментных образований, перенесенное онкозаболевание в прошлом, ожоги кожи в детстве, место жительства, расовую принадлежность, хроническую травматизацию; затем эти предикторы вводят в трехслойную нейронную сеть, предварительно обученную прогнозировать меланому кожи, которая на основе эталонных предикторов согласно таблице 1, содержащейся в описании, проводит автоматическое индивидуальное прогнозирование риска развития меланомы кожи или его отсутствие в последующий период до 3 лет со степенью достоверности порядка 85%.A method for predicting the risk of developing skin melanoma based on a neural network analysis of predictors, characterized by examining patients, determines significant factors - predictors: place of residence (village or city), social status, history of cancer and infectious diseases, change in pigmentation, the appearance of a corolla of hyperemia , inflammation, work in an enclosed space for more than 20 years, bleeding, malignant neoplasms in family members, age at the time of the study, rapid growth of pigmentation, pregnant st, the appearance of pigmented lesions subsidiaries transferred of cancer in the past, skin burns as a child, place of residence, race, chronic trauma; then these predictors are introduced into a three-layer neural network, previously trained to predict skin melanoma, which, based on reference predictors according to table 1 contained in the description, performs automatic individual prediction of the risk of developing melanoma of the skin or its absence in a subsequent period of up to 3 years with a confidence level of about 85 %
RU2018110753A 2018-03-26 2018-03-26 A method for predicting the risk of developing skin melanoma based on a neural network analysis of predictors RU2018110753A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
RU2018110753A RU2018110753A (en) 2018-03-26 2018-03-26 A method for predicting the risk of developing skin melanoma based on a neural network analysis of predictors

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
RU2018110753A RU2018110753A (en) 2018-03-26 2018-03-26 A method for predicting the risk of developing skin melanoma based on a neural network analysis of predictors

Publications (2)

Publication Number Publication Date
RU2018110753A true RU2018110753A (en) 2019-09-26
RU2018110753A3 RU2018110753A3 (en) 2019-09-26

Family

ID=68083900

Family Applications (1)

Application Number Title Priority Date Filing Date
RU2018110753A RU2018110753A (en) 2018-03-26 2018-03-26 A method for predicting the risk of developing skin melanoma based on a neural network analysis of predictors

Country Status (1)

Country Link
RU (1) RU2018110753A (en)

Also Published As

Publication number Publication date
RU2018110753A3 (en) 2019-09-26

Similar Documents

Publication Publication Date Title
MX2018002933A (en) Psma-targeted nir dyes and their uses.
WO2013134786A3 (en) Biomarker compositions and methods
BR0215579A (en) Compound and methods of treating psoriasis, selected disease from the group consisting of leukemia, colon cancer, breast and prostate cancer, autoimmune disease, inflammatory disease, skin condition and renal osteodystrophy
BR0308928A (en) Use of botulinum toxin for the treatment of cardiovascular disease, as well as composition for use in a cardiovascular procedure.
AR050215A1 (en) METHOD FOR DIAGNOSING DISEASES USING COPEPTINE.
MX2023001945A (en) Compositions and methods for screening solid tumors.
WO2018007872A9 (en) Biomarkers for inflammatory bowel disease
WO2013022995A3 (en) Biomarker compositions and methods
EA201691399A1 (en) METHODS AND READY MEDICATION FORMS FOR THE TREATMENT OF ALZHEYMER
BR112016018205A8 (en) methods of treating alzheimer's disease, patient selection, patient identification, prediction, therapeutic efficacy optimization, kit, agent use and in vitro agent use
ATE375811T1 (en) USE OF FLUORESCENT POLYMERS FOR THE TREATMENT OF HUMAN HAIR
ATE400297T1 (en) CTGF AS A TARGET MOLECULE IN THE TREATMENT OF MICROALBUMINURIA IN PATIENTS WITH DIABETIC NEPHROPATHY
EA201891913A1 (en) IMPROVED PREPARATIONS OF ADOLESCENT CELLS OF ADULT LIVER
NO20073175L (en) Antigenic epitopes of interleukin-21, related antibodies and their use in the medical field
WO2017100663A8 (en) Improved methods for treating her2-positive breast cancer
RU2018110753A (en) A method for predicting the risk of developing skin melanoma based on a neural network analysis of predictors
WO2017201448A8 (en) Animal models for psoriasis and screening methods
WO2020176461A3 (en) Methods and compositions for treating neuroendocrine prostate cancer
MX2024000466A (en) Compositions, devices, and methods of crohn's disease sensitivity testing.
RU2014113841A (en) METHOD FOR EARLY DIAGNOSTICS OF ENDOGENOUS INTOXICATION
MX2018015334A (en) Compositions, devices, and methods of gastroesophageal reflux disease sensitivity testing.
Burkov et al. Prediction of thrombotic and stenotic complications after «KemAngioprotez» biological prosthesis implantation by means of mathematical model
EA202092123A1 (en) ANTIBODIES AGAINST CLAUDINE 18.2 AND THEIR APPLICATION
Paramsothy et al. OP019 In faecal microbiota transplantation (FMT) for ulcerative colitis, fusobacterium is associated with lack of remission, while metabolic shifts to starch degradation and short-chain fatty acid production are associated with remission (FOCUS study)
Bloomer et al. 120 PHYLOGENETIC ANALYSIS OF THE Y CHROMOSOME IN ABDOMINAL AORTIC ANEURYSMS: A META-ANALYSIS OF THREE BRITISH COHORTS