Skip to content

Instantly share code, notes, and snippets.

@jejuro
Last active February 27, 2018 08:30
Show Gist options
  • Save jejuro/4689f5f4ca73d5c081ad6148d795cbf9 to your computer and use it in GitHub Desktop.
Save jejuro/4689f5f4ca73d5c081ad6148d795cbf9 to your computer and use it in GitHub Desktop.
openhash-health-eng.md

The text presupposes the following knowledge.

  • [Open Hash Platform White Paper] ()

  • [Open Hash Platform Lecture] ()

The characteristics of the block chain technology, which can not be falsified, can also be used appropriately for the purpose of constructing the national medical system ** fairly and efficiently. The facts will become clear through the text.

This project is one of the hands-on exercises that can be done with the learning of the ** standard curriculum ** and ** the autonomous course **. ** The Health Insurance Corporation ** is reproduced by the Openhash application, and the text presents an overview of the coding work. Specifically, we present a concrete plan and a system implemented to incorporate some of the work of the NHIC into the OpenHash Platform-based ** National Healthcare System **. This is accomplishing two goals that can not be achieved with conventional technology, and thus are contradictory to each other.

It will reduce the burden on the nation's NHS.

The health insurance coverage rate is 90%.

A block-based medical information system has many advantages.

  • The calculation of health insurance benefits is transparent.

  • Responsibility is clear when a medical accident occurs.

  • Compliance with the use of medicines, equipment or consumer goods (such as disposable syringes) is clear.

  • Significantly reduce the manpower to monitor the fairness of the medical / pharmaceutical industry.

  • Various AI modules can be easily applied throughout the medical procedure. Otherwise, it can be improved in keeping with the development of science and technology. This is described in detail in Section 5.

  • Without increasing the number of doctors, the per capita consumption of health care services can exceed the OECD average.

The proposed ** national medical system ** does not assume a standard currency. However, using standard currency can be more effective. In addition, the standard currency can be more easily spread by government policies such as the benefits of use, such as the reduction of the value added tax.

  • When trading in standard currency, 9.5% VAT applies (current 10%)

  • 5% reduction of income tax when reporting tax return in standard currency (by current income standard)

1. Overview

The National Health Insurance Corporation administers the following matters (National Health Insurance Act Article 14 (1)).

  • Qualification management of the National Health Insurance subscribers and their dependents

  • Collection and collection of national health insurance premiums and other levies

  • Management of National Health Insurance Benefits

  • Preventive business needed to maintain and promote the health of members and dependents of National Health Insurance

  • Payment of national health insurance benefit expense

  • Management, operation and proliferation of assets

  • Operation of Medical Facilities

  • Training and promotion of health insurance

  • Research and international cooperation on national health insurance

  • What the National Health Insurance Act stipulates in the work of the Corporation

  • Tasks entrusted pursuant to the Collection of Consignment Act (「National Pension Act」, 「Act on Collection of Insurance Premiums for Employment Insurance and Industrial Accident Compensation Insurance」, 「Wage Claims Guarantee Act」, and 「Asbestos Damage Relief Act」)

Other services entrusted under the National Health Insurance Act or other laws and regulations

  • Other duties deemed necessary by the Minister of Health and Welfare in relation to health insurance

1.1 Qualification Management of National Health Insurance Members and Dependents

After determining who will bear the cost of medical care for the country, he or she will be administered through the qualifications management of the subscriber and dependents.

The cost burden of the individual is proportional to his / her property and income. However, the weights of the property items and income items are different.

1.1.1 Property

There are a lot of discussions about taxation criteria and methods, even though they have a lot of land and real estate, but there is no cash income. For example, a land worth 100 million won and a car worth 1 million won may be an asset of the same value in terms of the health insurance. Since such judgment is based on political ideology, this project is out of the question.

In addition, it is the domain of the 'national sub-system' that records the individual's property and periodically evaluates it.

1.1.2 Income

As of 2018, the income of Korean workers is transparent, so it is relatively transparent to pay for the income. On the other hand, failure to pay income tax due to self-employed businesses such as income tax breaks is an issue related to the national tax collection mechanism.

2.2 Collection and collection of national health insurance premiums and other levies

The collection is also overly efficient, so it is out of the question.

1.3 Management of National Health Insurance Benefits

Insurance benefits refers to the services provided by the National Health Insurance Corporation in kind or in cash, as prescribed by the Act on the prevention, diagnosis, treatment, rehabilitation, birth, death and health promotion of sicknesses and injuries to subscribers and dependents.

It is one of the main coding targets for this project, which is to decide what kind of prescription to the patient and how much of the cost to bear. It is described in detail in Section 2.

1.4 Preventive Business Needed to Maintain and Promote Health of Subscribers and Dependents of National Health Insurance

It is one of the core targets of this project. Section 3 describes in detail.

2. Prerequisites

Summarize the current situation and preconditions in relation to health insurance and government medical care policies.

  • The Korean Standard Disease / Sign Classification (KCD7) classifies disease and death data according to similarity, such as medical records and statistics on the causes of death. The species is about 13,000 species.

  • The average OECD coverage rate is 80%, and the Korean government tries to reduce the medical burden of patients by raising the health insurance coverage rate.

With respect to the Openhash Platform,

  • It is assumed that a space web server cluster is established.

  • The stub server records the treatment / prescription / treatment events when the healthcare service consumer and the supplier's address server are the same.

  • Hierarchical forest server records other events.

  • Every citizen has a space web ID. Preferably, it also has a standard currency deposit account.

  • Medical expenses are paid in national currency. Preferably, they are paid in standard currency.

  • The pharmacy is a machine that excludes pharmacists.

For example, if Goyangbu residing in Hankyung-myeon, Jeju city is treated / prescribed / treated at Hankyung clinic located in Hankyung-myeon, the event will be recorded on a stub server that controls Hankyung-myeon.

The project will be described in detail below.

# 3 Healthcare Tx with NHS

The doctor's diagnosis / diagnosis / prescription is recorded as ** Medical Tx ** and confirmed by both the doctor and the patient. The confirmed record is Hashing, so it can not be modified later.

Below is a summary of the process of creating and implementing ** Medical Tx **.

Various forms, such as medical application form, diagnostic DB entry, medical certificate, and payment, are taken from ** Medical Document Form DB ** and attached to ** Medical Tx **.

  • ** Patient interface of medical Tx ** ** Medical institution ERP system **.

  • ** Medical Tx ** creation begins by recording the patient's space web ID in the document including the public key of the medical institution.

  • The provider enters diagnostic and prescription data into the ** Medical Tx **. This data is of Smart Contract type. If certain conditions are met, the costs are charged. For example, payment of medicines, administration of injections, operations and procedures may be the respective conditions.

  • Billing and settlement are by background processes. There is no need for the provider and the patient to be aware of the individual billing / payment process.

  • The monitoring of the billing / payment process is the job of ** NHBO **.

  • One medical Tx ** is terminated by discharge of the patient and includes two Hashes (the final Tx of the physician and the final Tx of the patient). The newly created medical Tx is inserted into Openhash after the new Hash is attached.

  • An unspecified number of medical Txs interact (Weaving), as one provider creates a variety of patients and medical Tx, and one patient creates a variety of providers and medical Tx. Thus, the recorded medical Tx can never be tampered with.

3.1 Medical Tx

The following figure shows the transaction flow of one or two block chains. The transactions between A and B, B and C, C and D are cascaded.

In the case of Openhash-based medical Tx, it is a transaction between a certain number of health care providers and an unspecified number of medical consumers. Therefore, the transaction between A and B, A and C, A and D proceeds.

Note that, like other Openhash Platform application systems, the two-layer hashing algorithm is Hashing Tx Data.

  • A mechanism for interworking a set of data generated within an organization with hashing. For example, one inpatient may develop multiple Tx during hospital stay.

  • A mechanism whereby all the Tx Data that constitute the national healthcare system are interlocked by hashing the data to be published outside the organization. For example, one Tx can be generated at the time of admission and discharge of a particular patient.

3.2 Tx Data

Medical services are also traded and recorded with the same mechanisms as other goods. The seller provides the goods and the buyer provides the money. Transactions are recorded in the ** Standard Ledger ** of the Openhash Platform. ** Standard Ledger ** Records are based on the six principle.

3.3 Medical Document Form DB

A medical document can be assigned to a 16-bit number in the IPv6 Subnet ID field, and up to 65,636 types can be specified in the first field (16-bits) of the interface area.

Traditional sense automation => Automation

** Medical Tx ** recording system based on block chaining can easily accomplish intentions or goals that are difficult to implement with conventional technology. Below ** comparison ** can be performed by the system, not by the person.

  • By comparing each product purchased and consumed for a certain period of time, problems such as the use of past medicines or the reuse of disposables are easily identified.

By comparing the types of medicines and services used in each disease, transient / underdetermined care / prescriptions can be easily identified.

  • By measuring the amount of work done by each healthcare worker, it is easy to identify working conditions that violate the Labor Standards Act.

  • In case of a medical accident, the responsibility is clear.

** The greatest advantage of medical Tx ** is that it can be improved in keeping with the development of technology.

5. Automation by Machine Running => Autonomy

Machine learning technology provides a solution for tasks that are difficult to implement with previous science and technology.

5.1 Diagnosis and Prescription

The doctor's treatment and prescription is based on data, and the data is based on interviews, stethoscopes, blood analysis and culture, and photographs. Medical Tx systems can easily incorporate machine learning techniques into such diagnostics.

MRI, CT, blood culture analyzers and other medical equipment used by medical institutions ** medical Tx ** can be fused. This not only assures the objectivity of the diagnosis, but also makes it possible to utilize the AI ​​function widely in the diagnosis process.

From Tx Data, the machine can learn two things.

  • Correlation between the input data of the device and the diagnosis of the physician

  • Interrelationship between doctor's diagnosis and prescription

For this purpose, it is necessary that the ** national medical system ** has control over the equipment / devices of the domestic medical institution.

By comparing the two items below, you can control each of the two systems. From people to AI, AI is learning to learn from people.

  • Diagnosis / prescription by doctor and its prognosis

  • Diagnosis / prescription by AI and its prognosis

Once AI's diagnostic capability reaches a certain level, ** physicians ** will be able to use the AI ​​module as an Intelligent Agency to assist in their diagnosis and prescribing process. Therefore, the total amount of medical services can be dramatically increased without increasing the number of physicians.

This data can be used to calculate the amount of the following NHI salary.

5.2 Calculate the amount of health insurance benefit

By applying the AI ​​module which learned the existing method, it is possible to automate some of the calculation of the benefit in the future. For example, what is the appropriate cost of treatment when you have a cold? Due to many variables such as patient age, health status, the current number of medical services can not be answered. However, with a simple machine learning algorithm, a standard prescription / treatment process and cost for each of a variety of patient groups can be determined.

Machine learning proceeds in two main directions.

5.1 Inductive Method: Alphago Lee

The following contents are subject to completion of ** Level 8 Artificial Intelligence Elementary Course ** provided by Coded Korea.

  • Observe the results after different prescriptions by gender, age, and health status.

  • After presenting the prescription that corrected the previous results, observe the results.

  • Repeat the above procedure until it reaches the equilibrium point.

  • By doing so, establish a ** standard treatment ** according to the patient's condition.

This is a way of learning through experience.

  • Organize the ingredients of each medicine sold on the market.

After comparing the physician's prescription with the patient's progress, select the ingredients with optimal results by age, gender, and health status, and when and how to administer them.

It is a relatively simple algorithm (Supervised or Unsupervised Learning) and requires accumulated data.

Distinctive Method: Alphago Zero

Math and logical reasoning. In particular, it combines physical and chemical laws.

It constitutes a logical causal relationship between the drug substance and physical symptoms.

Logically derive the components of a drug to be used as a treatment for a specific disease.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment