Professor Seo Byeong-gwan from Kyung Hee University

Introduction

A project for establishing Korean Medicine clinical big data is actively carried out according to the government’s “4^th Comprehensive Plan to Develop Korean Medicine (2021–2025).” The project is planned to complete the development of Korean Medicine AI algorithm beyond phased-in standardization and quantification of Korean Medicine clinical data.

The Western Medicine has already been developing various services using big data such as an AI doctor. This makes us wonder just how unique the future AI algorithm based on Korean Medicine big data will be.

1. Why do you think AI trend is rising in the medical industry?

As you said, AI is trending in the medical sector. AI technologies are pervading our everyday life. You can see this in detail in films about the future. For example, a child who has been ill for a long time enters a so-called diagnosis box and the name of a disease and treatment method are instantly revealed.

It is a scene where a “person” is excluded or minimized as a “medical provider” in the series of processes for accessing the latest knowledge system to identify the patient’s problem accurately and finding the optimized answer to solve the problem. In short, the future will be heading towards a way of ruling out “human errors.”

To think about a reason, it takes a lot of time and costs to produce medical professionals who can provide quality medical service. In addition, it requires long hours of training for medical professionals to organize a massive amount of information generated in the treatment process, such as test results of a patient, reaction to certain drugs, the nature of the related technology, and side effects, and classify them according to priority.

However, AI can accomplish these processes in only a short period of time. AI technologies can imitate human intelligence to quickly judge which tasks to perform, collect corresponding data, and evaluate the value of data. That’s why there are more and more cases of AI being used in different areas of medical sites like new drug development and surgical robots.

2. We heard that you are studying AI based on Korean Medicine big data to be used for clinical trials. Where do you put your focus on during the research?

In the process of treating a patient, a medical professional collects comprehensive personal health records (PHRs) through counseling, history taking, physical examination, and more. Korean Medicine is a consecutive process of diagnosing a patient's condition based on personal information, recommending the appropriate treatment method to the patient, and providing improved medical treatment by evaluating the result. This process is based on the understanding of all information related to a patient, including medical treatment, general health information, health insurance, genetic information, and public information, in other words, big data.

Korean Medicine is a personalized medicine in nature, and its goal is to present the most suitable answers to patients based on comprehensive information. And that is what AI research based on Korean Medicine big data aims for. We are trying to help patients and medical professionals find the best alternative for the patient in a more convenient way.

3. What are the difficulties in researching AI development based on Korean Medicine big data?

To collect a vast amount of information in the health and medical field, it is essential to use the standard EMR. As the EMR is based on clinical site data, it can be used by various organizations, thereby enabling evaluation and comparative study of the safety, effectiveness, and economic feasibility of Korean Medicine treatment. It can also prepare effective measures for improvement based on the data generated at sites of Korean Medicine.

Keeping up with this trend, the modeling of various data is designed based on the disease-specific CPG, and an “input system prototype”¹ is being developed based on the standard EMR DB structure found through the modeling.

There is a prerequisite to complete the AI development research based on Korean Medicine big data, which is preparing a standard framework for converting the terms, information, and format/content of medical charts used in medical treatment into significant data applicable to actual medical sites. In other words, we must prepare a standardized information system that can be used in any medical site uniformly and prevent data redundancy or inconsistency. To this end, we need a multilateral approach and support, such as a classification system including the principle of name, definition, form, and rules about the data information elements.

4. How do you expect such research will have an impact on Korea and the world?

Today, artificial intelligence is being developed as a tool for attending to the needs of medical sites along with the advancement of deep learning technology. Accumulating extensive data on patients is necessary to achieve goals, from a simple task of evaluating the test result values and determining the problem to risk evaluation through information analysis, suggestion of diagnosis and treatment methods, and prognosis evaluation.

The era for establishing, disclosing, and using medical data is coming. There will be a sea of information that integrates numerous independent health and medical big data, including data generated from medical institutions and local communities, data generated from various research and clinical trials, and data produced from literature materials like journals.

This vast array of information will become the grounds for the development of AI technology based on Korean Medicine and greatly contribute to raising the medical quality of the world by making technical development of Korean Medicine treatment from multilateral aspects, such as diagnostic techniques of Korean Medicine, introduction of treatment technologies, and prognosis evaluation methods.