BIG DATA IN MEDICINE

BIG DATA IN MEDICINE

Ezozbek Tokhirov

Ph.D., Associate professor, Tashkent State Transport University,

Uzbekistan, Tashkent

ABSTRACT

In this report there are applications in the field of big data located in the medical industry, as well as access to the main destinations in the medical field. As a result, a comparison of the best world practices for medical services through the US, Europe and other developed countries.

Keywords: Big data, Medicine, Analysis.

INTRODUCTION

Big Data is not only a large amount of information, but also the technology of its storage, and analytical processing methods.

Huge amounts of information require processing and analysis. Of course, we can select a certain amount of information, process it, and calculate the result as a reflection of the regularity of the entire volume of information by calculating the forecast, probability and reliability of such a coincidence

But in the 21st century, information in libraries, hospitals, various research institutes, laboratories have already been digitized. Should this be neglected?

In addition, the above approach does not exclude errors, which sometimes can turn a mathematical forecast into guessing.

Therefore, now random sampling of various variables very quickly tends to a full set of their values, which often need to be processed very quickly.

And, if we touch on one of the most important issues in our life - health, then in the case of medical treatment of large data, this is literally a matter of life and death.

The results of processing «large data» should be obtained as quickly as possible. This will make it possible to turn the analyst from a tool that answers the question «who is to blame?», Typical of traditional analytical systems, into a tool for getting answers «what to do?». The specialist in this case from the doctor of the pathologist becomes a therapist. Speed of data access and speed of their processing is an important criterion for the quality of technologies included in Big Data.

Earlier, it was technically impossible to take into account a multitude of indicators of the work of medical institutions, therefore, funds were channelled to the bank account of hospitals, proportional to the total number of bed-days and the number of performed manipulations. Because of this, a vicious system was formed in which it was advantageous to conduct «extra» diagnostics, prescribe unnecessary counselling and delay treatment.

If you pay attention to today's health care reform in the US, we see opportunities for more fine-tuning of Big Data. Its authors adhere to the concept of «accountable assistance», within which the effectiveness of the treatment will be assessed.

Thus, thanks to processing technologies and analysis of large data, they will pay not so much for the treatment process as for the ability to quickly heal and maintain the patients' health.

Another important direction was the forecasting of expenditures. It is based on a multifactorial analysis of statistics such as the number of repeated visits, the percentage of complaints against specific doctors and units, the prevalence of various pathologies, the number of patients with chronic diseases, and epidemiological indicators.

Employees of the National Health Service of the United Kingdom (NHS) use the Big Data analysis of the frequency of repeated hospital admissions and missed appointments, the total time of the patient's stay in the operating room, the availability of medical supplies and materials during the operation and their availability. And also by the total amount of time a surgeon spends on performing operations in comparison with administrative or preparatory measures. The definition of inconsistencies in these indicators allowed the NHS to increase the operational load by 2%, which provided savings of 20 thousand pounds a week. As a result, patients not only began to perform operations for vital signs more quickly, but also discharge them more quickly from the hospital, making room for other people in need.

Thus, using the analysis of «large data» allows foreign providers of medical services to shorten the length of stay of patients in the hospital, to keep costs down and to reduce the number of repeated hospitalizations.

Medical analytical problems that can be solved using the analysis of «large data» can be of various types depending on the level of maturity (in ascending order):

1.Descriptive analyst (answering the question «What happened?»);

2.Diagnostic analytics («Why did this happen?»);

3.Predictive analytics («What will happen in the future?»);

4.Prescribing analyst («What needs to be done to prevent this from happening?»).

With the increasing complexity of tasks, the complexity of the analytical system and algorithms increases as well as the number of necessary data sources, from simple information from medical records and biometric monitoring data to genomic and family data, and even to information from social networks.

MAIN BODY

In health care, modern technologies come in that support all the standard methods of working with data. This design and filling of multidimensional OLAP cubes, the ability to synchronize OLTP and OLAP storages in real time, the rapid development of analytical panels using a library of visual components, the ability to analyse unstructured text data and conduct predictive analytics.

An analysis made in a McKinsey report shows how «large data» can not only create an additional source of cost recovery, but also improve the quality of medical care. The basis of Big Data can be combined information stored in the four main sources of data that are not interconnected today. It:

• Data obtained during research and testing;
• Data from clinics on the history of the disease and diagnosis;
• Data on the behaviour of patients, their purchases, reviews, data from home medical devices and even from clothes and shoes, such as sneakers with sensors;
• Data from medical institutions on the provision of services, pharmacies on the release of drugs, information on prices in the health market.

Based on the analysis of all these data, it is proposed to develop the following areas of use of Big Data:

1. Operations of medical institutions. There is an opportunity to study the effectiveness of treatment by processing all available information about the practice of treatment. Based on an analysis of all known medical histories and diagnostics, the practice of physicians will make extensive use of decision support systems that allow the clinician to have previously unheard of access to the experience of thousands of colleagues throughout the country. Methods of personal and preventive medicine, based on remote monitoring of patients will lead to a significant reduction in costs and improve the quality of life. The proliferation of various sensors of human body activities connected to wearable gadgets makes it possible to reduce the need for conducting laboratory tests, prevent unexpected complications, and an automatic reminder of the need for independent therapeutic and prophylactic manipulations will improve the quality of the prescribed treatment;

2. The system of pricing and payment. Analysis of invoices and receipts using automated procedures based on machine learning and neural networks will reduce the number of errors and thefts in payment. The formation of price plans that take into account the real possibilities of the population and the need for services also increases the total income from patients. Only systems that work with «large data» allow you to switch to a payment based on the quality of care provided and to jointly regulate the costs of medicines and medical personnel;

3. Research and development. The greatest effect here should be expected from new possibilities of predictive modelling in the development of medicines. Equally important, statistical algorithms and large data tools are designed to plan clinical trials and involve patients in such trials. The processing of the results of such tests is another important application of «large data». A special place in research and development in health care is now occupied by innovations in personalized medicine. Based on the processing of huge volumes of genetic information that are becoming more accessible to humans, doctors will be able to prescribe absolutely unique medicines and methods of treatment. Finally, the development of disease patterns will allow us to obtain good prognostic assessments of the development of various types of diseases, identify risk profiles and not only to carry out preventive measures, but also to forecast the need for the development of treatment methods effective for future types of diseases;

4. New business models. Based on digital data in healthcare, these models can complement existing ones or even compete with some of them. These are data aggregators that supply analysed and assembled blocks of data that satisfy specified conditions to third parties. For example, all medical records of patients using a pharmacological drug are important for pharmaceutical companies and they are ready to buy such data. Another potential for new business models are on-line platforms for patients and doctors, medical researchers and pharmacologists;

5. Mass screening and prevention and detection of epidemics. This area is based on Big Data, the development of technologies allows building both geographical and social models of public health and predictive models for the development of epidemic outbreaks.

CONCLUTION

Despite the fact that the degree of penetration of Big Data in Uzbek health care is lower than in the US and Europe, the problems of attitude to these technologies are similar. Although medicine is one of the industries in which the “big data” management technologies have the most dramatic effect, many still feel sceptical about them, perhaps because of the not always clear business benefits and lack of specialists.

References:

1. Hemingway H, et al (2018). Big data from electronic health records for early and late translational cardiovascular research: challenges and potential.
2. Archenaa, J., Anita, E. M (2015). A survey of big data analytics in healthcare and government. Procedia Computer Science.
3. Dash, S., et al (2019). Big data in healthcare: management, analysis and future prospects. J Big Data.
4. He K.Y, Ge D, He M.M (2017). Big data analytics for genomic medicine. Int J Mol Sci.
5. Hrovat, G., et al (2014), M.: Contrasting temporal trend discovery for large healthcare databases. Computer Methods and Programs in Biomedicine.