Physicians and data scientists have teamed up to establish Big Data Clinical Utilization Research Society to encourage the right use of medical big data.
Founded in December, the society held an inaugural seminar in January. As the rising number of people with dementia gave more significance to big data-based research, the organization signed a memorandum of understanding with the Korean Dementia Association to collaborate in research in March. The society also works with the Korean Association for Geriatric Psychiatry to study depression among the elderly, which raises the risk of suicide.
Korea Biomedical Review met with Kim Hun-sung, president of the Big Data Clinical Utilization Research Society and a professor at the Endocrinology Department at Seoul St. Mary’s Hospital, at the society’s seminar at Samsung Medical Center on Friday.
Kim emphasized that physicians have a crucial role in turning medical big data into meaningful knowledge.
|Kim Hun-sung, president of the Big Data Clinical Utilization Research Society and a professor at the Endocrinology Department at Seoul St. Mary’s Hospital, poses during the society’s seminar in Seoul on Friday.|
Question: Why is the role of physicians important in big data research?
Answer: Big data has become a hot issue. Many scientists and physicians have earned simple information from big data but failed to go to the next step to make it meaningful knowledge. This is because they did not reflect medical concepts on big data. A strange thing is happening because people extract simple information and coerce medical explanation later. So, wrong information is overflowing.
Physicians have to actively intervene in all stages from gathering raw data to turning them into information, and to knowledge. There is no one else but doctors who know medical data the best. This is why we founded the Big Data Research Society.
Q: How have you come to establish the Big Data Research Society?
A: I saw many physicians extracting information from abundant data. As there is no medical concept reflected on big data, the results and interpretations were widely different depending on who uses what kind of statistics to come up with the inference.
I once asked a data analysis firm to analyze my data. I saw the results that a patient who came on Monday morning had good numbers in blood glucose, and on Friday afternoon, bad numbers. The problem was, the company advertised to patients, “If you want to receive good medical consultation, visit the hospital on Monday morning.” This was not the fault of the data analysis company. From a medical point of view, patients who see a doctor on Monday morning tend to be more relaxed and take care of themselves relatively well. On Friday afternoon, however, patients are busy and eager to get the prescription and leave the hospital as soon as possible. In other words, there are two different groups of patients. The data analysis company earned information from data but completely failed to obtain knowledge.
Physicians should pay much attention to how they should interpret medical data.
Then, I found myself committed to clinically utilizing knowledge, obtained through big data.
I gathered physicians who know how to handle data and industry officials to establish the society for research.
Q: Who are the members of the Big Data Research Society?
A: Our operation team consists of 32 people who are industry officials, doctors, and scholars. We have 180 registered members, and the number of people who subscribed to our website online has reached nearly 1,300. Non-physicians who are interested in medical big data account for about 60 percent of all members.
Q: What does the Big Data Research Society plan to do?
A: To utilize medical big data for clinical use, physicians and data scientists should work together. Doctors should not hope that scientists can turn any data into information or knowledge.
Doctors are very interested in big data but also afraid that artificial intelligence (AI) can replace them. However, the physicians at the society are not. We think AI can help us treat patients. It is impossible for AI to replace doctors.
When I was a trainee doctor, we used paper charts. If I had a CT scan on a patient with appendicitis, my senior professor scolded me for “depending on a machine” instead of checking the patient with my hand. At that time, people used to say CT machines will replace doctors who treated appendicitis. However, now we have a diagnostic radiology department. It is the same with AI. Physicians should be faithful to their work, not overwhelmed by a flood of information. I hope conservative doctors could actively support the use of big data.