Nine years ago one of the world’s leading artificial intelligence scientists chose an endangered working type.
“People should.” Stop training radiologists now,“Geoffrey Hinton said, adding that it was” absolutely obvious “that in five years AI overcome people in this area.
Today, radiologists – doctors in medical imaging who look inside the body to diagnose and treat disease – still high demand. Recent study From the American College of Radiology, she assumed an ever -increasing workforce by 2055.
Dr. The Hinton, which was awarded the Nobel Prize for Physics last year for pioneering research in AI, was generally correct that this technology would have a significant impact – simply not as a murderer.
This applies to the Mayo Clinic radiologists, one of the leading medical systems of the nation whose main campus is in Rochester in Minn. In recent years, AI has begun to focus on images, automate routine tasks, identify medical abnormalities and predict disease. AI can also serve as a “second set of eyes”.
“But would it replace radiologists? We didn’t think so,” Dr. Matthew Callom, Radiological Chairman of Mayo Clinic, remembered the prediction of 2016. “We knew how difficult it was and everything that was concerned.”
Computer scientists, work experts and politicians have long been discussing how AI will eventually play in the workforce. Will it be a smart helper, strengthening human performance or robotic spare and sending millions of workers?
The debate has intensified because the top technology for chatbots seems to improve faster than expected. Openai, anthropic and other Silicon Valley societies now predict that AI will be in most cognitive tasks within a few years. However, many scientists predict a more gradual transformation in accordance with seismic inventions of the past, such as electricity or the Internet.
The estimated extinction of radiologists provides narration a case study. So far, AI will prove to be a powerful medical instrument to increase the efficacy and enlargement of human abilities, rather than someone’s work.
As for the development and deployment of AI in medicine, radiology was the main goal. Of the more than 1,000 AI applications approved by Food and Drug Administration for use in medicine, about three quarters are in radiology. AI usually excels in identifying and measurement of specific abnormalities such as lung lesion or breast lump.
“There has been amazing progress, but these AI tools are largely looking for one thing,” Dr. Charles E. Kahn Jr. Radiology: artificial intelligence.
Radiologists do much more than study pictures. They advise other doctors and surgeons, talk to patients, write news and analyze medical records. After identifying the suspect cluster of tissue in the organ interpret what this could mean for an individual patient with a specific medical history and the use of years of experience.
The predictions that AI will often steal jobs “underestimate the complexity of the work that people really do – just like radiologists do much more than reading scanning,” said David author, economist at the Massachusetts Institute of Technology.
At the Mayo clinic, AI tools were examined, developed and adapted to match the work routines of busy doctors. Since the forecast of Doom Dr. Hinton to more than 400 radiologists increased the staff by 55 percent.
In 2016, income by warning and progress in recognizing the image of the driven AI, the leaders of the radiology department compiled the group to assess the potential impact of technology.
“We thought the first thing we should do is use this technology to improve us,” Dr. Callom. “That was our first goal.”
They decided to invest. Today, the Radiology Department has an AI team with 40 people, including AI scientists, radiological researchers, data analysts and software engineers. They developed a number of AI instruments, from tissue analyzers to disease predictors.
This team cooperates with experts such as Dr. Theodora Potetzke, which focuses on the kidneys, bladder and reproductive organs. It describes the role of a radiologist as a “doctor for other doctors”, which clearly communicates the results of the display and helps and advises.
Dr. Potetzke cooperated on AI, which measures the volume of the kidneys. The kidney growth, when combined with cysts, can predict a decrease in kidney function before it occurs in blood tests. In the past, the kidney volume was mostly manually, with the equivalent of the ruler on the screen and guessing. The results varied and the chore was time consuming.
Dr. Potetzke worked as a consultant, end user and tester when working with the AI department. It helped to design a software program that has color coding for different tissues and checked the measurement.
Today, it triggers a picture on the computer screen and clicks on the icon and the kidney volume measurement will immediately appear. She saves her 15 to 30 minutes every time she examines the kidney image and is permanently accurate.
“It is a good example of something I have been very comfortable with AI for efficiency and accuracy,” Dr. Potetzke. “It can expand, help and quantify, but I’m not at the place where I give up the interpretation conclusions of technology.”
Francis Baffour, a radiologist of the staff, explained a diverse way that AI was applied to the field, often in the background. MRI and CT scanners use AI algorithms to speed up and clean them, he said.
AI can also automatically identify images showing the highest probability of abnormal growth, basically tell the radiologist: “Look here first”. Another program scans images for blood clots in heart or lungs, although medical focus may be elsewhere.
“AI is now everywhere in our workflow,” Dr. Baffour.
Overall, the Mayo Clinic uses more than 250 AI models, developed internally and licensed from suppliers. The department of radiology and cardiology are the largest consumers.
In some cases, the new technology opens the door to the knowledge above human abilities. One model AI analyzes data from electrocardiograms to predict patients more likely to develop atrial fibrillation, the abnormality of heart rhytm.
The research project in radiology uses the AI algorithm to recognize fine changes in the shape and texture of the pancreatic to detect cancer up to two years before conventional diagnoses. The Mayo Clinic Team team cooperates with other medical institutions for further testing of more data algorithm.
“Mathematics can see what the human eye can’t,” Dr. John Halamka, President of the Mayo Clinic Platform, which oversees the digital initiative of the medical system.
Dr. Halamka, Optimist AI, believes that technology transforms medicine.
“In five years, it will not be negligible to use AI,” he said. “But it will work and AI cooperate.”
Dr. Hinton agrees. In retrospect, he believes that he spoke too wide in 2016, he said in -mail. He did not find out that he was talking purely about the image analysis and was wrong about the timing, but not the direction, he added.
In a few years, most of the interpretation of medical images will be performed “by a combination of AI and radiologist, and it will be much more efficient in radiologists in addition to improving accuracy,” Hinton.
