Aim. To develop and verify a method for diagnosis of peptic ulcer based on neural network analysis of data on patients’ risk factors.Materials and methods. This article presents the results of a study based on materials on risk factors of 488 patients. The data was analyzed using internally developed artificial neural network (Certificate of State Registration of Program for Computers (RU) no. 2017613090).The results of the study. The proposed approach demonstrated the levels of sensitivity of 74.4%, m = 4.3 and specificity of 93.3%, m = 2.46 during clinical testing.The prediction of the age of probable hospitalization ensured the generation of an array of data for which the Mean Absolute Error (MAE) of the prognosis was 1.8 years, m = 0.11 in the training set and 1.9 years, m = 0.15 in the clinical testing set. The maximum of absolute prognosis error in the clinical testing set did not exceed 2.2 at p = 0.95 and 2.3 years at p = 0.99.Conclusion. A new method is proposed for diagnosis of peptic ulcer based on a neural network analysis of data on patients’ risk factors. During clinical testing of the model, this approach demonstrated Area Under the Curve (AUC) levels reaching 0.943. The use of the artificial neural network also made it possible to predict the age of probable hospitalization. The use of the neural network demonstrated additional advantages including: non-invasiveness, the lack of need to prepare the patient for the study and the possibility to obtain results immediately after the onset of the disease without a time delay for sample processing.

Aim. Using multilayer perceptron artificial neural network, to develop a mathematical model for predicting the need for surgical intervention in patients admitted for hepatopancreatoduodenal zone diseases and to assess the capabilities for its clinical application. Methods. The study was performed using the data of 488 patients with peptic ulcer, cholecystitis or pancreatitis, analyzed using multilayer perceptron artificial neural network, trained to distinguish vectors of data on risk factors of patients who did or did not require surgical intervention during current hospitalization. Results. Patients in the training sample who had required surgical intervention during hospitalization were different from patients who had undergone conservative treatment by such characteristics as gender, age, duration of the disease, state on admission, and the structure of risk factors. The acquired data made it possible to train the artificial neural network. The ROC analysis of the mathematical model demonstrated the area under the curve (AUC) equal to 0.880 for the training group (n=385) and 0.739 for the clinical approbation group (n=103). Conclusion. The AUC indicators of the created model can be characterized as very good in terms of predicting the need for surgical treatment in the training group and good for the clinical approbation group: sensitivity and specificity of the model exceed 80% in the training group and are highest in patients with peptic ulcer disease; in the clinical approbation group these parameters were lower as expected, however, remained at the level of 60-70%.

The purpose of the study was to evaluate the possibilities of neuronet differential diagnosis of hepatopancreatoduodenal zone diseases and their potential for the formation of an individualized preventive strategy. The work was performed on the basis of materials of 488 patients with the pathology of the hepatopancreatoduodenal zone. Neural network analysis of information on risk factors (sex, age, eating habits, stress, family status, bad habits) was used. Results. The sensitivity of the differential diagnostic model reaches Se = 84,7, m = 1,66 for peptic ulcer, Se = 81,4, m = 1,8 for pancreatitis and Se = 92,1, m = 1,24 for cholecystitis. Specificity levels equaled respectively to Sp = 91,5, m = 1,29, Sp = 90,1, m = 1,38 and Sp = 82,7, m = 1,75. The article also presents a method of developing an individualized diagnostic strategy using the artificial neural network.

Objective was in the public health study to select and assess practical application possibilities of optimal biosatistical methods for monitoring the functioning of artificial neural networks trained for predicting the quantitative health indicators in patients with hepatopancreatoduodenal zone diseases. Methods. The study was conducted on materials of 385 patients with hepatopancreatoduodenal diseases who underwent in-patient treatment in hospitals in the city of Kursk. There was used the internally developed information system "System of Intellectual Analysis and Diagnosis of Diseases" (Certificate of State Registration of the Program for Computers n. 2017613090). The application provides functionality for creation, configuration, training and practical application of the artificial neural network, multi layer perceptron. Hyperbolic tangent was used as an activation function. Results. There is presented the experience of selection and practical application of mathematical methods for controlling the operation of the artificial neural network in assessing the quantitative health indicators in patients with peptic ulcer, cholecystitis and pancreatitis. There is shown the expediency of the primary interpretation of the output layer neuron OUT value where OUT is an element of R boolean AND OUT is an element of (-1; 1) to the scale and units of measurement of the evaluated health indicator with subsequent statistical processing of the array of obtained values. The optimal mathematical methods include: the calculation of the means (and their errors) for the arrays of empirical and resulting from the neural network operation values, with subsequent comparison of the arrays using the X-2 criterion and determining the significance level a. The next step is to estimate the forecast mean error (ME), the forecast mean square error (MSE), the forecast mean absolute error (MAE), the maximum forecast error for the 99th and 95th normal distribution percentiles, the mean percent error (MPE) and the average absolute percentage error (MPAE). An example of a tabular representation of the analysis data is given. Conclusion. The most convenient and informative mathematical methods for assessing the operation quality of the artificial neural network predicting the quantitative health indicators in patients with hepatopancreatoduodenal zone diseases are various types of forecast errors (mean error, mean absolute error, mean percentage absolute error, etc.). It is expedient to calculate the maximum absolute error of the forecast (for p = 0.05 and p = 0.01), which increases the visibility of the results, as well as the X-2 criterion, that allows the estimation of hypothesis' significance that there are no differences between the arrays of calculated and empirical quantitative indicators.

Purpose. To develop an artificial neural network for diagnosing and predicting the development of cholecystitis based on an analysis of data on risk factors, and to explore the possibilities of its application in real clinical practice.Materials and methods. The collection of materials was held in at the hospitals of the city of Kursk and included a survey of 488 patients with hepatopancreatoduodenal diseases. 203 patients were suffering from cholecystitis, in 285 patients the diagnosis of cholecystitis was excluded. Analysis of risk factors’ data (such as sex, age, bad habits, profession, family relationships, etc.) was carried out using an internally developed artificial neural network (multilayer perceptron with hyperbolic tangent as the activation function). The computer program “System of Intellectual Analysis and Diagnosis of Diseases” was registered in accordance with established procedure (Certificate No. 2017613090).Results. The use of neural network analysis of data on risk factors in comparison with the processing of information that forms a clinical picture allows the diagnosis of a potential disease with cholecystitis before the onset of symptoms. The training of the artificial neural network with a quantitative output coding the age of probable hospitalization made it possible to generate an array of values, signifficantly (α ≤ 0.001) not differing from the empirical data. The difference between the mean calculated and mean empirical values was 0.45 for the training set and 1.75 for the clinical approbation group. The mean absolute error was within the range of 1.87–2.07 years.Conclusion. 1. The proposed new approach to the diagnosis and prognosis of cholecystitis has demonstrated its effectiveness, which is confirmed in clinical approbation by the levels of sensitivity (94.44%, m = 2.26) and specificity (80.6%, m = 3.9).2. The error in predicting the age of probable hospitalization of patients with cholecystitis did not exceed 2.29 and 2.38 years for p = 0.95 and p = 0.99, respectively.

Background. In the social and hygienic study, to develop an artificial neural network designed to diagnose pancreatitis and to predict the time of its onset based on an analysis of information about risk factors, and to test the program in clinical practice. Materials and methods. The study was conducted on the materials of 488 patients (including 167 clients with pancreatitis) who underwent inpatient treatment in the city of Kursk for hepatopancreatoduodenal zone diseases. Data processing of information on health risk factors (sex, age, bad habits, stress, professional and family history, previous treatment) was carried out using an internally developed software package – “System of Intellectual Analysis and Diagnosis of Diseases” (Certificate for State Registration No. 2017613090). Results. A new approach to the diagnosis and prediction of pancreatitis based on a neural network analysis of data on risk factors was proposed. The sensitivity and specificity levels of this method equaled to 76.74% (m = 4.16) and 90% (m = 2.96), respectively. The error in predicting the age of probable hospitalization did not exceed 2.87 and 3.02 years (for p = 0.95 and p = 0.99, respectively). At the same time, the system demonstrated additional advantages: non-invasiveness, low requirements for equipment and professional training of health workers, an opportunity to evaluate the result from the time of the onset of the disease. Conclusion. The effectiveness of the proposed approach was confirmed at the stage of clinical approbation with sensitivity and specificity levels corresponding to similar indicators of traditional diagnostic methods – ultrasound, computed tomography and determination of α-amylase and lipase levels.

Цель. В социально-гигиеническом исследовании разработать искусственную нейронную сеть, предназначенную для диагностики панкреатита и прогнозирования времени его наступления на основе анализа сведений о факторах риска, а также провести апробацию программы в клинической практике.Материалы и методы. Исследование проведено по материалам 488 больных (из них 167 с панкреатитом), проходивших стационарное лечение в городе Курске по поводу заболеваний гепатопанкреатодуоденальной зоны. Обработка информации о факторах риска здоровью (возрастно-половой принадлежности, вредных привычках, стрессах, профессиональном и семейном анамнезе, ранее проводимом лечении) производилась с применением программного комплекса собственной разработки – «Системы интеллектуального анализа и диагностики заболеваний» (свидетельство № 2017613090).Результаты. Предложен новый подход к диагностике и прогнозированию панкреатита на основе нейросетевого анализа данных о факторах риска. Показатели чувствительности и специфичности такого метода, находились на уровне (76,74%, m = 4,16) и (90%, m = 2,96), соответственно. Ошибка прогноза возраста вероятной госпитализации не превышала 2,87 и 3,02 года (p = 0,95 и p = 0,99, соответственно). При этом система демонстрировала дополнительные преимущества: неинвазивность, низкие требования к оборудованию и профессиональной подготовке медработника, возможность оценивать результат с момента возникновения заболевания.Заключение. Эффективность предложенного подхода подтверждена на этапе клинической апробации уровнями чувствительности и специфичности, соответствующими аналогичным показателям традиционных диагностических методов – ультразвукового исследования, компьютерной томографии и определения уровней α-амилазы и липазы.

Aim. To develop a set of information methods to improve the quality of neural network diagnosis of diseases of hepatopancreatoduodenal zone. Methods. The study involved 385 patients with peptic ulcer, cholecystitis and pancreatitis undergoing in-patient treatment in medical organizations of the city of Kursk. For data mining internally developed software «System of Intellectual Analysis and Diagnosis of Diseases» was used which is an environment for the creation, adjustment, training and practical clinical application of an artificial neural network, such as a multilayer perceptron with an activation function - hyperbolic tangent. Results. Hyperbolic tangent (activation function) of the output layer’s neuron takes the value OUT ∈ ℝ ∧ OUT ∈ (-1; 1) which requires an interpretation. For logic network gates, for example, presence/absence of a disease, it can be performed by comparison with an arbitrarily assigned threshold yB ∈ (0; 1). In this approach, the values are interpreted as false (if y ≤-yB), undefined if y ∈ (-yB; yB), or true (if y ≥yB). Network operation control includes calculation of sensitivity, specificity, false positive and false negative results, for which the comparison of arrays of pairs of calculated and empirical values is carried out. In case of artificial neural network use for diagnosing diseases of hepatopancreatoduodenal zone, the optimal mode was achieved assigning yB≈0.3 as a threshold of the output neuron activation function. Conclusion. Assessing the quality of the ability of artificial neural network with logic outputs to diagnose hepatopancreatoduodenal zone diseases, as well as its controlled setting, is most effective by evaluation of sensitivity, specificity, frequency of false positive and false negative results at the threshold value yB≈0.3; the demonstrated sensitivity (83-94.7%) and specificity (83-97.8%) levels are comparable to the traditionally used diagnostic methods.

Aim. To study the impact of risk factors on development of pancreatitis in comparison with other diseases of the hepatopancreatoduodenal area and to evaluate the effectiveness of physical training as a preventive approach. Material and methods. The study involved 488 people who were treated for diseases of hepatopancreatoduodenal area. Results. The following factors significantly (p < 0.001) determining the direction of pathogenesis along the path of pancreatitis or other diseases should be considered most important: patient's dietary intervention and alcohol consumption by patient's family members ( r(Q) = 0.39-0.42 and 0.28-0.67, respectively), as well as gender ( r(Q) = 0.40-0.51, p < 0.001) and employment ( r(Q) = 0.24-0.26, p < 0.05). The strongest correlations between quantitative risk factors and the formation of pancreatitis in contrast to cholecystitis were observed for age (vertical bar r(bs)vertical bar = 0.23, p < 0.001), alcohol abuse (vertical bar r(bs)vertical bar = 0.17, p < 0.001), age at starting smoking (vertical bar r(bs)vertical bar = 0.29, p < 0.001), and diet costs (vertical bar r(bs)vertical bar = 0.21, p < 0.001). Regression analysis revealed a special preventive effect of 'sports expenses' factor rate on the development of bad habits. The optimal level of such costs is 7-10 %. Conclusion. The obtained results can be used to develop the personalized strategy for prevention of pancreatitis and allow recommending to patients to maintain the level of expenses on physical training at the level of 7-10 % of their income.