A.I. and the Future of Medical Science
A.I. and the Future of Medical Science
  • Ahn Sung-min Professor
  • 승인 2017.03.15 11:29
  • 댓글 0
이 기사를 공유합니다

Last year, Gachon University Gil Hospital adopted Watson for Oncology to diagnose a patient with colorectal cancer with accuracy similar to that made by doctors. The patient had third stage colorectal cancer and had laparoscopic surgery. With the data all input, Watson for Oncology chose FOLFOX and CapeOX for the main treatments, which are general anticancer drugs.
The recent introduction of Watson for Oncology, a computer-based medical consultation aiding system for cancer, into the domestic market has initiated an enthusiastic discussion about A.I. (Artificial Intelligence) and the future of medical services, which had been wholly based on the decisions made by humans. To understand and acknowledge the part that A.I. takes in the realm of medical sciences, there has to be some perception that medical science is being developed along with data science as the accumulation of medical data is one of the very fundamentals of medical treatments.
The amount of data or information that doctors need to know in order to understand and thus perfectly treat one patient is exponentially increasing. Not just various image materials and diagnosis analysis but also gene data and patient’s lifestyle have come under the realm of medical science, acknowledged as additionally required information for doctors to understand their patients. According to IBM (International Business Machines)’s analysis, the data based on gene and lifestyle for a patient in his whole life respectively accounts for about 6 and 1100 terabyte, respectively. Considering that the amount of data for modern clinical treatments of a patient is about 0.4 terabyte, the additional amount of information that doctors should be aware of to approach perfect medical treatment of a patient is extraordinary.
The National Institute of Health has proclaimed that for Precision Medicine, which corresponds to specialized treatments for individual patients, must be based on several things. The institute claims the following three will establish a foundation of validity for Precision Medicine: development in the analysis techniques of genome, health screening technologies, and the common usages of Big Data. The advancements in data production/restoration and the usages of Big Data will make Precision Medicine possible. Therefore, Precision Medicine is indeed a data-dependent medical science. Then, in this technological context, what role does A.I. play?
Basically, A.I. is an instrument that enables us to efficiently deal with loads of data. Now, it is practically impossible for a single doctor to successfully analyze all information about a patient, understand the cutting edge guidelines and documents about the diseases of the patient, and eventually deduce a perfect treatment. Taking multidisciplinary approach for cancer patients as an example, for deciding treatment of one cancer patient, doctors of various majors such as surgery, internal medicine, radiation oncology, pathology, diagnostic radiology, etc gather. Like this, A.I. can increase the efficiency of medical science in the realm where a collection of various professional data is required. Of course, as many people say, A.I. cannot completely substitute doctors. However, it must be acknowledged that things can now be divided into what humans have to do, what humans are good at, and what A.I. must take part in. Vinod Khosla, the founder of an authoritative IT group of Silicon Valley (Sunmicrosystems), predicted in 2013 that “Ten years from now, the contribution made by data science will overrun that of the entire field of biology”.
Along with the expansion of medical services converged with Big Data and A.I., intelligent medical services based on personal gene information, lifestyles, etc. has come to a practical stage. A.I. such as IBM’s Watson and Google’s Verily utilized for diagnosis interpretation onsite is increasing daily. IBM, for example, is planning to establish an open type cloud of medical information so that doctors, researchers, medical companies, and insurance companies can freely make use of the Big Data and keep on innovating personal medical care. Medical science is seeing is a prelude to a huge innovation of data science led by A.I.