[APOA 수부상지학회 스토리 #8]
“The digital medical record subscription service is made in the form of a capsule that the patient swallows, from a smart toilet that analyzes stool components to a biosensor membrane that measures key vital data such as body temperature, rate pulse and heart rate, and a microcomputer capable of performing blood tests and cell collections It was possible to update subscribers with real-time biometric indicators and status collected by various biosensors, including sensors … After a while , a notification appeared on Smart Stream announcing that Garcia’s digital medical records had arrived … Garcia’s condition was critical. The hospital started an automated AI-powered process to find an antiviral drug that might work for him. ” (AI 2041 / Kaifu Lee, Qiu Fan Chen)
2023 is the first year of the AI war. Chat GTP opened the door, and several global companies started launching new AI services. The recent development of AI can be imagined, by connecting Chat GTP with augmented reality programs provided by various global companies and Apple’s ‘Siri’, the early model of AI assistant Jarvis in the movie ‘Iron Man’ can be realized The speed is very fine quick
The article quoted above (AI 2041) is part of a story that predicts the development of artificial intelligence in the next 20 years in a factual format, taking into account the current technological level, speed of development, and potential practical application. Similar predictions of future situations can be found in many books, articles and movies. Commonly, real-time monitoring and analysis of various variables, including human biomarkers, diagnoses the current condition and predicts the extent of future risk, and uses AI algorithms to manipulate them in a short time to create drugs that n suitable for patients. content that work is possible forms the mainstream.
In addition, given the current technology level, some of them have already reached the commercialization stage, and many of them are very likely to be realized in the future. However, there are relatively few stories about automated surgical treatment using AI other than data analysis, diagnosis and drug development. The reason is that there are many areas that are still difficult to realize for robots with artificial intelligence, such as the unique experience and difference of the surgery itself, coping according to the situation, and surgical technique.
For this reason, it can be said that AI research using surgical medical data, including orthopedic surgery, still lacks much quantitative or qualitative AI research and practical application in internal medicine and radiology. Even that focuses on analyzing pre- and post-operative clinical data, pre-treatment diagnosis and classification, predicting post-operative outcome through pre-operative evaluation, and comparing the accuracy of analysis algorithms. Of course, as many people expect, at the end of AI research in the field of surgery, there will eventually be an AI robot that is more accurate than a human without human intervention. However, it is still difficult to obtain quantitative data on surgical techniques and differences between operators, which makes it difficult to start learning by translating surgery itself into data.
However, it is possible to check the main areas of research using AI only in the field of surgery and orthopedic surgery, or the content in which commercialization is at work. First of all, as a study on post-operative clinical and radiological outcome prediction, AI has compared and learned pre- and post-operative case data for several years and developed an algorithm that increases efficiency and predicts clinical outcomes with high accuracy through It is also commercialized by the company.
Due to the nature of orthopedic surgery, where post-operative functional recovery treatment is particularly important based on these data, several domestic startups have recently suggested methods of diagnosis and treatment using various clinical information of patients, managing the patient’s prognosis and restore function after surgery. more and more efforts to apply AI algorithms to treatment.
There is also a surgical robot that applies AI. Based on computed tomography (CT) images guided by ‘humans’, the accuracy of surgery is increased by navigation during artificial joint surgery, bones are broken during artificial joint surgery using robots, and the Da Vinci surgical robot Surgery minimally invasive performed as the use of minimally invasive surgery is still widely used in various medical institutions.
However, in 2022, a research team at Johns Hopkins University in the United States reported a successful case of laparoscopic surgery for intestinal anastomosis through a ‘smart tissue autonomous robot (STAR)’ without human intervention. In addition, in Korea, a research team at Chonnam National University has been publishing research on a fracture surgery robot where an AI robot directly matches the fracture site for several years, and is in the process of commercialization . There are not many such research movements yet, and there will be many technical, legal and ethical problems to overcome before surgery based on AI judgment without human intervention can be used in practice.
There are still a number of limitations in applying AI algorithms to medical care. It is crucial to ensure excellent data in quantitative and qualitative aspects of data, but it is not easy. In addition, in the case of AI algorithms, there is a “black box”, which makes it difficult to understand the reason even if the prediction rate and accuracy are high. Although various methods for analyzing the “black box” are being developed, if it is difficult to fully confirm the reason, it will inevitably be difficult to apply the results presented by AI to humans.
Before 2023, the first year of the AI war, the industrial application of AI algorithms is no longer an option, but a necessity. The global event ‘CES 2023’ also chose digital healthcare as a keyword and put AI at the center. Therefore, despite the various practical limitations and drawbacks mentioned above, the application of AI algorithms in medical care has become the trend of the age. In the future, when developing AI, it will be necessary to apply AI to different medical fields to develop medical care as well.
Professor Chung Kyu-Hak (APOA Aquatic Physiology Society Academician)
