BACKGROUND: Currently, studies have focused on the modernization of existing methods of forensic age assessment (bone and skeletal) through the active use of modern methods of medical imaging (e.g., computed tomography) and artificial intelligence for their analysis. This approach enables the creation of new methods for assessing biological age, which is characterized by increased accuracy and reproducibility. AIM: To develop and test an algorithm for predicting the biological age of an individual based on computed tomography analysis of the knee joint using artificial neural networks and computer vision. MATERIALS AND METHODS: This observational retrospective transverse (one time) study analyzed computed tomography scans (334) of the knee joint performed in the Departments of Radiation Diagnostics of the Priorov Central Institute for Trauma and Orthopedics, Vreden National Medical Center for Traumatology and Orthopedics, between 2018 and 2021. The study enrolled persons of both sexes aged 13–45 years. Cases of developmental abnormalities, knee injuries, signs of general connective tissue pathology were excluded. Research methods include the use of intelligent information technologies (a formalized set of mathematical and software solutions). RESULTS: Based on the experiments conducted, an algorithm for assessing age according to the computed tomography scans of the knee joint has been developed. The main components of the developed system are as follows: a preprocessing module, an intelligent computing core, a data analysis module, a three-dimensional reconstruction module, a property extraction module, and a final age assessment module. The essence of the proposed method is the simultaneous use of artificial neural networks and clearly formalized mathematical procedures for calculating the properties of the epiphyseal line. To obtain the results and conduct primary experimental studies that confirmed the feasibility, correctness, and operability of the method, software using the YOLOv5 neural network was developed. The result of the error matrix analysis after training shows a probability of correct recognition of the order of 80%. Verification of experimental studies was performed on 46 cases. At present, the age estimation error is approximately 1 year for children and adolescents. CONCLUSIONS: The experimental results have confirmed the adequacy of the age estimates obtained to the actual age of the individual and, consequently, the applicability of the proposed method in forensic medical institutions. The proposed method is currently implemented as a set of software components with subsequent manual integration of automatically calculated data. The plan was to supplement the database of computed tomography images to increase the training sample and the accuracy of age prediction.

The article presents the experience of artificial intelligence application in research process. The article contains general information about basic concepts of machine learning (clustering and visualization), as well as considers more detaily an experience of clinical testing. The effectiveness of applying Data Analysis methods and means as one of the research stages is demonstrated on the example of a case on processing medical information using algorithms of machine learning: solving the problem on diagnostic value of the proposed indicator (FTF) for determining target age groups. Implementation of such approach of digital transformation improves the operational effectiveness of researches, as well as quality and availability of final technological products being developed - software for solving expert problems.

A relevant and highly demanded modern medicine problem in many of its areas is the timely detection and recognition of pathogenic microorganisms and microbial communities in the patient’s tissues for the speedy prescription and correct use of medicines from mutually exclusive tactics. The transition to a new level in the speed of visualization of the samples’ contents taken and the accuracy of diagnostics is possible because of the use of lanthanide staining in combination with scanning electron microscopy to retrieve a series of high-resolution images with subsequent automatic labelling and classification of microbiological objects. This paper presents the results of using the YOLOv5 neural network model to detect 15 different most common opportunistic classes of bacteria in 380 images. As a result, a 71.5% average accuracy and 69.8% recall were achieved by using the YOLOv5 base model without freezing layers.

THE AIM OF THE STUDY:Was to compare the prevalence and structure of sudden death from cardiovascular diseases (CVD) in the structure of nonviolent death in Moscow Region as a representative of the large region and St. Petersburg as a city of federal importance. Frequency and structure of this parameter, as well as its dynamics were analyzed. The comparison of obtained results with data of Federal State Statistics Service was done. Nonparametric analysis of the initial data has been conducted. Clustering and visualization, based on the following parameters of initial sample, have been done: «death rate from CVD in the structure of nonviolent death», «morbidity rate», «incomes» and «unemployment rate». Correlation dependences between death rate from CVD according to the Form №42 and mentioned parameters of medical and social state of the subject have been determined. Dependences between sudden death rate from CDV and a number of medical and social parameters have been established.

The diagnosis of many diseases is largely possible thanks to MRI(Magnetic Resonance Imaging). This technology allows to study internal organs of the patient: the brain, spine, bones, joints, vessels and etc. The resolution of the MRI image is limited due to various factors: movement of the patient during the scan, the continuous movement of internal organs. The higher the quality of the MRI image, the longer it takes to scan. For more accurate diagnostics it is possible to increase resolution of the yielded images. This is achieved by using SISR(Single Image Super Resolution) algorithms, which allow you to obtain images with increased resolution from a single input image. In this paper the idea of the image super-resolution algorithms is presented, various forms of the problem and solutions to it are provided. The advantages of the SISR algorithms are described. The relevance of this task in the field of medical MRI images is explained. Metrics for comparing image quality PSNR and SSIM are given and described. A dataset for testing is presented. The stage of data preparation is described: the principle of selecting images from a set of datasets, converting data into the required format, compressing images to obtain input data for selected neural network models. The PSNR, SSIM metrics of two neural network models mDCSRN and FAWDN are measured on equally prepared input data. The comparison results are presented in the form of images and averaged data for the entire sample is stored in the table.

This review describes the history of development of a new line of chemical reagents that prompts to significantly reevaluate the application of scanning electron microscopy (SEM) in medical and biological studies, particularly in ophthalmology; considers the establishing of SEM as an analytical method; covers the problems in its application associated with the needs of clinical medicine and the complexities of biological sample preparation for electron microscopy. The article also presents in chronological order the technical solutions associated with creating a unique line of reagents for supravital staining. The multitude of technical solutions allows considering SEM as a method of express diagnostics. The review discusses examples of practical application of these methods for solving certain cases in clinical ophthalmology. The niche of SEM is considered among other methods of clinical diagnostics, as well as its future development involving the use of artificial intelligence.

The purpose of the study is to study the prevalence and frequency of sudden death (SD) from cardiovascular diseases (CVD) in the structure of non-violent death, taking into account the socio-economic development of the constituent entities of the Russian Federation. We analyzed the frequency of this indicator, compared it with the data of the Federal State Statistics Service, and determined the overall dynamics. We conducted a non-parametric analysis of the initial data, carried out clustering and visualization based on the following parameters of the initial sample: «CVD mortality in the structure of non-violent death», «morbidity» and «per capita income level». Correlation dependences of the level of mortality from CVD according to the form 42 on the indicated indicators of the socio-economic condition of the subject were determined. Identified subjects of the Russian Federation with an increase in mortality from CVD; established the dependence of the level of VS on CVD and a number of medical and economic indicators.

As a result of a sharp popularity increase of artificial intelligence resource-intensive methods, a serious problem arises in the preliminary data preparation for the convolutional neural networks models effective training. The authors present an approach based on the training dataset iterative updating principle using the YOLO neural network model for areas of interest detection, objects selection, and the original images labeling process automation. The proposed approach was tested with various model configurations for bacteria labeling on images obtained using a scanning electron microscope and, on average, demonstrated ~90% precision on a training dataset increased by 1.75 times over the initial training dataset.

The objective is to study the prevalence and rate of sudden death from cardiovascular diseases (CVD) in the structure of non-violent death in people under 35 in the Tula region as a representative of a small region of the Russian Federation. A comparison with the data of the Federal State Statistics Service was made. The overall trend was determined. A non-parametric analysis of the initial data, clustering and visualization were performed based on the following parameters of the initial sample: «mortality rate from cardiovascular diseases (CVD) in the structure of non-violent death,» «morbidity,» and «number of doctors in the region.» The correlation of mortality from CVD according to Form 42 and the specified indicators of the medical and social status of the region was determined. The relationship between the rate of CVD-related sudden death and specified medical and social indicators was established.

The progress of cognitive impairments at the initial stage is poorly researched and is difficult to be diagnosed. This paper proposes an integrated approach based on the initial data analysis and the subsystem construction to predict the progression of further complications in patients with cognitive impairment symptoms. The combined use of statistical analysis and machine learning methods made it possible to achieve 78% accuracy in diagnosing subtle cognitive impairments in a sample of 526 patients. Visualization methods helped to improve the created model’s interpretability.

Применение методов машинного обучения в медицинской сфере стремительно демонстрирует свою перспективность, но остается сложной темой для исследования ввиду мультидисциплинарности и новизны большинства методов. В данной статье описан сценарий, применимый для обработки медицинских данных, состоящий из решения таких задач как: предварительная обработка с учетом специфики и разнообразности исходных данных, уменьшение размерности, классификация, кластеризация и визуализация объектов. Результат этого подхода может быть использован для интерпретации исходной информации о пациентах для дальнейшего применения ее в клинической практике медицинскими специалистами.

This article considers the possibility of creating an algorithm for universal image preprocessing that would simplify its segmentation and is suitable for any type of medical or biological study methods. The authors propose a number of equally applicable functions that describe the changes in the parameters of a raster image along the normal to the border of an arbitrary biological object and at the same time doesn’t depend on the instrumental method of image acquisition. The work also determines the absence of any parametric adaptation as the boundary condition for reducing the subjective contribution to the final scientific and practical result. Methods. Based on the biological foundations, the following three simple mathematical expressions were chosen for changing brightness on the border of a biological object: All the functions participating in calculations were not scaled, their size corresponded to the coordinates expressed as raster image dots. The Pearson’s correlation without reliability assessment was used to describe the probabilities of the object’s border being present in the coordinate x=0, and to assess its affinity with properties of the proposed synthetic functions. The basis of functions was tested using digital images of three fundamentally-different objects obtained with different physical detectors: sagittal section of the MRI scan of human head capturing the hippocampus area (Siemens Magnetom Skyra, Germany); optical section of eye cornea (Pentacam, Oculus, Japan); image of transverse histological section of human skin stained with hematoxylin and eosin obtained with microscope (Olympus BX65, Olympus, Japan). In addition, practical evaluation of the applicability of the developed approach was done by processing 100 two-dimensional images of different arbitrarily selected brain MRI sections. A field of probability of the presence of “biological border” in each dot of the image was built for each image. Results. The best affinity — with significant outperformance — to all kinds of digital images of biological objects’ borders was found in the compound function described by the formula. Its affinity to object’s borders on the MRI images was 0.87–0.99; for the borders of biological objects on the images obtained with optical methods it was lower than 0.99 only in two out of seven tests (0.95 and 0.97). Discussion. The high affinity of the proposed compound function may be associated with the fact that the three members of this synthetic expression correspond to three individual additive physical phenomena manifesting its biological basis. The first member is responsible for general biological contrast differentiation — the delineation of tissues appears sharpest at the borders of the organs. The second member describes the conventional membrane present in all healthy organized biological structures, including epithelial lining of the esophagus, muscle fascia, developed mucous membrane etc. The third member of the proposed function characterizes the blurred junction of arbitrarily-uniform structures. The proposed mathematical apparatus describes the formation of scalar field of probabilities for selected borders of biological structures and can easily be transformed into vector representation. Additional advantage of the developed and tested mathematical apparatus is the fact that the formulated approach can be used for determining the orientation of brightness increment gradient. This, in turn, allows distinguishing additional properties of biological objects based on calculation of the direction of development and growth of the tissue and structural anisotropy.

The paper presents the results of a comparative epidemiological study of thermal injury and fatal hypothermia based on the results of consolidated reports of the Bureau of Forensic Medicine of the Moscow Department of Health for 2017-2019. It was analyzed the archival material for three calendar years. It was used a continuous retrospective research method with an assessment of the general aggregate of the death incidence from thermal injury and hypothermia in Moscow in 2017-2019 by means of nonparametric statistical methods. Deaths from thermal injury and hypothermia are the most often accidents. The incidence of this type of death is characterized by ups and downs depending on the season. May, January and December are the most dangerous for burns. A similar pattern, with the exception of May, was noted for hypothermia. It was found that men aged 50-70 and women of 70-90 years old die most often. It was determined that most social characteristics (education, being married, etc.) alter the average age when deaths occur. The results obtained provide statistical justification for further more thorough study of thermal injury and hypothermia.

Проводится анализ уязвимостей нейросетевого алгоритма симметричного шифрования по отношению к различного рода атакам. Особое внимание уделено внутреннему представлению информации нейросетью, которое вносит определенную структурированность, что отрицательно сказывается на криптостойкости. Предложен ряд способов улучшения криптостойкости, выявлены их достоинства и недостатки.

.

The analysis of key stages, implementation features and functioning principles of the neural networks, including deep neural networks, has been carried out. The problems of choosing the number of hidden elements, methods for the internal topology selection and setting parameters are considered. It is shown that in the training and validation process it is possible to control the capacity of a neural network and evaluate the qualitative characteristics of the constructed model. The issues of construction processes automation and hyperparameters optimization of the neural network structures are considered depending on the user's tasks and the available source data. A number of approaches based on the use of probabilistic programming, evolutionary algorithms, and recurrent neural networks are presented.

There are several numbers of diseases and conditions which contribute the progress of cognitive impairment and deterioration of cognitive functions compared to an individual norm. Currently, the causes and progression of cognitive impairment are poorly researched. This paper presents a system analysis of data including 152 patients and the results of their examination. Various machine learning models were used in the research, including classification and clustering algorithms. Methods of data preprocessing and dimension reduction were used for visualization. The result models have demonstrated the possibility of applying the latest artificial intelligence approaches for data analysis in the medical field. It was possible to achieve an accuracy of the cognitive impairment diagnosis in the initial stages of 75% despite the small data set.

ОБЪЕМНЫЕ ХАРАКТЕРИСТИКИ ГИППОКАМПОВ ПО ДАННЫМ МР-ВОЛЮМЕТРИИ У ПАЦИЕНТОВ С БОЛЕЗНЬЮ АЛЬЦГЕЙМЕРА

This paper demonstrates the need for an integrated approach to improve the accuracy of corpse age calculation based on measurements of a heterogeneous nature and conditions of partially distorted or incomplete data, using modern information intelligent technologies. A generalized algorithm developed on the basis of various mathematical methods of primary data processing and the extraction of different structures is presented, in addition to the reliability of the information.