Lin Qiang underwent a full set of physical examinations, including full-body CT scans and mNGS sequencing. Inspections are expensive and time-consuming. BUT FOR AN EXPERIMENTAL GROUP WITH A 4U CHASSIS FULL OF EIGHT TESLA V100 ACCELERATOR CARDS, THIS PART OF THE COST IS ACTUALLY NOT TOO MUCH TO BEAR.

Strictly speaking, Lin Qiang was the object of the old thing's first practical experiment. Although the researchers in the group have some disagreement on whether it is necessary to apply the old thing to this incurable disease in the first experiment, now that it has been decided, everyone still hopes to go all out and try to collect as much information as possible for the old thing.

The more information there is, the more accurate the diagnosis that the old thing can make will theoretically.

To find out if there is a problem with the design of their AI algorithm, it is certainly not enough to rely on the case reports that have the answers in advance - when manually summarizing and recording cases, people will always unconsciously remove some data that is not "direct" enough to ensure that the report itself is simple and easy to read.

And these processing methods used to ensure that other people read smoothly have actually saved the old thing a lot of trouble - there are too many human symptoms, and whether or not these redundant variables can be eliminated and the cause can be accurately found is the biggest test that the old thing needs to face at present.

What made this group of experimenters feel crazy was that even if they tried their best and tried every means to get all the information for the old thing, they couldn't immediately verify whether the old thing's calculations were correct - no one knew what disease Lin Qiang had, and even the diagnosis conclusions reached by Dean Wu Youqian and the experts of the three military medical universities were quite vague.

AI designers and operations managers are not doctors, they are standard science students. The result of the calculation is either a definite A or B, or there are multiple results within some explicit range. The results are now good, and the calculated answers do not match the standard answers. If it's completely wrong, then at least there is a clear signal that "you are doing something wrong and need to recheck the process". There are clear results, but the reference answers are vague – how can this be referenced?

Wu Youqian looked at the many operators in front of him with anxious faces, and waved his hand indifferently, "Look at you like this, it's too useless." He took a sip of tea with a smile, "When doctors encounter patients with unclear diagnoses in clinical practice, they must also use some means to eliminate them." It's like asking a doctor to hit the bull's-eye on his first thought and find the problem of the disease. Don't talk about him Sun Lien, you can't find any doctor. The old thing has only been corrected for less than a year, isn't the accuracy rate already catching up with him?"

"Dr. Sun's misdiagnosis rate is so low that if every diagnosis hadn't been supported by video, I'd definitely suspect that he cheated. The lead experimenter wiped the sweat from his face, "His average misdiagnosis rate is less than 18%, and the misdiagnosis rate of emergency-related diagnoses is only about 10%. The average misdiagnosis rate of the old stuff is now 25%. If it weren't for Dr. Sun's high rate of misdiagnosis in gynecology and obstetrics, he would really be a monster. ”

"The design direction of the old thing is 'the reference for doctors' clinical diagnosis', and it was not designed to replace doctors in the first place. Wu Youqian looked at the researchers with different expressions in front of him, and explained with a smile, "I can understand your idea of wanting to improve your results to the point of perfection, but scientific research should also respect objective facts - no matter how hard you try, treatment and diagnosis are always the professional work of doctors." It's impossible to replace them, and to be able to make one a useful tool that you can rely on, that's what you're working on at this stage. ”

Although many practitioners are reluctant to mention the topic of misdiagnosis, it must be admitted that misdiagnosis is widespread in medical activities at home and abroad. Leaving aside the deliberate misdiagnosis that needs to be referred to the judiciary, the average misdiagnosis rate among doctors is about 30%. These misdiagnoses may be due to different presenting features of the disease, atypical presentations, or inexperience of doctors, or even the patient's concealment of medical history, resulting in errors in the basis of the diagnosis.

In China, in addition to the above reasons, doctors are also suffering from stress and fatigue. The main reason why Sun was able to achieve an average misdiagnosis rate of 18% was not because of his own outstanding ability, but because of the limited number of cases he handles on a daily basis, and the fact that he is able to devote more time and energy to each case without direct contact with patients. And there is almost no restriction on the ability to perform any test he wants in each case, regardless of whether the patient's family can afford the test or whether they are willing to cooperate.

Sun Lien's situation is a special case, and it is no exaggeration to say that in the "ninth consultation room" of the laboratory building, Sun Lien is definitely the most nourishing doctor in the country and even in the world. The reality is very different from this ideal situation.

This is also the purpose of AI diagnosis proposed by the project team, if a set of cloud-based AI algorithms can be used to provide doctors with a "second opinion" that will not have any complaints, and will not make mistakes due to fatigue and work pressure, then it will be able to significantly reduce the misdiagnosis rate.

At the moment, however, it seems that the old stuff still has a long way to go.

·

·

·

"Is it right to sign these two documents?" The next morning, Sun Lien received a call from Mrs. Lin in the hospital. There is no psychiatric department in the Fourth Hospital, so after meeting with Sun Lien and signing the corresponding permission and notice of the experimental team, Wu Youqian took people to directly take Sun Lien, Mrs. Lin, and Lin Qiang to the Affiliated Hospital of Ningyuan Medical College. There is a psychiatric department, which can receive Lin Qiang, who was preliminarily diagnosed with "chronic fatigue syndrome".

Mrs. Lin and Sun Lien went to the toll window to hang up the number of the psychiatric department, and avoided Lin Qiang from beginning to end from registration to payment. In order to prevent some mistakes in the middle, even the registered items are written as "special needs outpatient clinic".

"You have been feeling very unwell lately, and you have muscle ......pain," said the doctor in charge of the consultation, the deputy director of the psychiatric department, whom Wu Youqian had told hundreds of times the night before. As the deputy director of a large tertiary hospital directly affiliated to a medical school, Director Mei should have been very pompous. It's a pity that Wu Youqian was already the dean when Director Mei went to school, and the teacher had a life, and the students really didn't dare to blow up the thorns. Director Mei could only honestly conceal his true identity and start the diagnosis directly. After re-questioning, he excused himself to "need a final consultation" and asked Lin Qiang to have breakfast first, while he went to the room where Wu Youqian and Sun Lien were waiting.

Director Mei took off his mask and said to the two, "There is no problem in diagnosing chronic fatigue syndrome, if there is no absolute CD4 cell count drop, in fact, I am more inclined to mass hysteria." ”

"That means it can be treated, right?" Wu Youqian nodded, he glanced at his watch, "Then let's strike while the iron is hot and quickly deal with this problem." ”

"The idea of using indigo rouge injection as a placebo is really creative. Director Mei was very interested in this treatment plan, and put forward his own views on the details of the treatment, "Ordinary syringes are transparent, and the color of indigo rouge injection is the same as stained urine. I suggested going to the dental department to borrow a metal syringe and stick the observation window with black tape, so that the gods couldn't see what it was. ”

"You can also ask the nurse who injected to wear a lead suit in the CT room. The brainstorming sharing session was good for this, and Sun Lien had an idea and put forward his own suggestion, "Anyway, the more serious it looks, the better the treatment effect of placebo should be." ”

The placebo effect is generally directly related to two conditions, one is "expected" and the other is "conditional". The two can regulate and interact with each other. Sun Lien has already planned for Lin Qiang in advance, "One injection will get rid of the disease." At the same time, this expectation is reinforced by means of conditionalization. Next, it's time to see the real chapter.

Lin Qiang sat in the consultation room, a little cramped and uneasy. His wife was sitting not far from him, but she seemed to be wearing a warning of "don't talk to me". He tried to speak several times, but finally gave up on the idea and stared anxiously at the door of the examination room.

"Lin Qiang, right?" The heavily armed nurse lady walked in with a plastic shock absorbing box, and when she saw Mrs. Lin who was also in the room, she was a little uncertain, "You are a family member?"

"Yes. Mrs. Lin nodded, stood up and said to the nurse, "I'll wait outside." With that, she went out the door on her own. I didn't wait for the nurse sister to reply at all.

Lin Qiang glanced at his wife's back with some anxiety, but his attention was quickly attracted by the box in the nurse's hand.

"This is a special medicine for you. The nurse sister adjusted the box in her hand without a trace, and put it on the table to open it - a silver-white syringe with a metallic luster appeared in front of Lin Qiang, and introduced, "Because this medicine is radioactive and reacts to light, it must be injected with a special metal syringe." While disinfecting Lin Qiang's arm, she asked, "Did the doctor tell you about the effective process of this medicine?"

"Nope. Lin Qiang shook his head a little nervously, "They only said that this medicine takes effect very quickly." ”

"That's right. However, this medicine does not work for some people. The nurse continued with the scheduled line, "When the injection is over, you can go to the bathroom to check your urine in about fifteen to twenty minutes." If it takes effect, the urine will turn blue. If there is no discoloration, it is not effective, so you have to tell the doctor quickly, you know?"