Compared with autonomous driving technology, artificial intelligence is much less controllable in the medical field. The organs and functions of the human body are relatively fixed and will not be as complicated as the road conditions of some roads]

[Industry statistics show that the global precision medical market will reach $173 billion by 2024. ]

The application of artificial intelligence in medical treatment has been controversial, but with the deeper development of machine learning technology, it will become the trend of the future for doctors to diagnose and manage patients with artificial intelligence.

Pilot artificial intelligence

At the annual conference hosted by Shanghai University of Science and Technology, many top international scholars including Turing Award winner Ivan Edward Sutherland shared the latest research achievements and core technologies in artificial intelligence and vision.

One of the highlights of this year is that machine vision will have a broader application prospect in medical care.

Professor Alan Yuille, an expert in artificial intelligence at Johns Hopkins University, told First Financial News that his team is working with Johns Hopkins Hospital to use artificial intelligence and deep learning to assist patient diagnosis and management. Yule is a student of the famous physicist Hawking. After completing his Ph.D. in the 1980s, he turned from physics to artificial intelligence.

Professor Yule believes that artificial intelligence will be the first to achieve large-scale application in medical care compared to unmanned driving. “We analyzed the medical images of tens of thousands of case databases and used them to provide training in radiology and pathology to help doctors to target more accurately and efficiently.” Professor Yule said that the project is still in progress. At the research stage, it is still subject to a series of policies such as the US FDA to enter the clinic. However, he believes that artificial intelligence has a promising application in medical imaging.

Although the application of artificial intelligence in the medical field is still in its early stages, the top hospitals in the United States have already begun to act. As early as two years ago, GPU giant Nvidia announced a partnership with the Massachusetts General Hospital Medical Data Science Center. The center hopes to become the center of artificial intelligence medical applications, detecting, diagnosing, treating and managing diseases. NVIDIA's DGX-1 Deep Learning Supercomputer has also been installed at Massachusetts General Hospital.

It is understood that the database of Massachusetts General Hospital contains 10 billion medical images. If the NVIDIA DGX-1 can achieve the expected results, the future medical data science center will be able to expand into the fields of electronic medical records and genetics.

The most popular artificial intelligence projects in the top five hospitals in the United States include predictive analytics, which monitor patient status through data analysis and key metrics, and individualized treatment options for patients through machine learning.

At the beginning of last year, the Mayo Clinic Personalized Medical Center teamed up with the startup Tempus to develop a personalized cancer treatment program using a machine learning platform. Tempus conducted molecular sequencing and analysis of 1,000 Mayo Clinic patients involved in immunotherapy-related research, targeting cancer, melanoma, bladder, breast and lymphoma. Industry statistics show that the global precision medical market will reach $173 billion by 2024.

Getting involved in medical care

Coincidentally, Yule is the mentor of Zhu Xi, the founder of the Shanghai Artificial Intelligence Unicorn Company.

Yule gave high praise to Chinese startups like Etu: "With the support of leading natural language processing (NLP) technology, computer vision technology, and speech recognition technology, Yitu is actively deploying medical fields. This is a Good direction."

Just last month, Yitu announced a strategic level cooperation with Huaxi Hospital to accelerate AI medical care for the people's livelihood. The two sides said that they will use AI to manage multimodal medical big data and build a full-scale clinical research platform to promote hospital intelligence. Yitu and Huaxi Hospital also released two joint research and development results of lung cancer artificial intelligence: the world's first lung cancer clinical research intelligent disease database, and the world's first multidisciplinary intelligent diagnosis system for lung cancer.

According to Wang Wenzhou, director of marketing products of Yitu Medical Products, “In the first national lung cancer clinical research intelligent disease database released by Yitu Medical, Yitu Medical and Huaxi Hospital completed the cross-system of 28,000 cases of lung cancer patients with full cycle data. Integration, including more than 1 million clinical documents and reports, more than 10 million original medical images, laying a solid foundation for research in the field of lung cancer, and based on this, developed the world's first multidisciplinary intelligent diagnosis system for lung cancer, greatly improved The level of lung cancer diagnosis and treatment in primary health care institutions."

Dr. Yang Jun, an analyst with Yitu, also said: "Deep learning and artificial intelligence technology use powerful computing power to build complex mathematical networks that can't be completed by common statistics and mathematical models of machine learning prediction and risk grading."

For example, he said that in the first time after the onset of stroke patients, the CT and MR imaging data of the input computer were highly intelligently and quantitatively evaluated, and the gaps in image decision-making output on the multidisciplinary connection were filled to achieve true clinical service. Auxiliary clinicians can carry out "multidisciplinary consultation" of informationization, intelligence and digitization to improve the overall diagnosis and treatment efficiency of stroke. "The faster and more accurate the image interpretation, the more accurate the diagnosis and treatment recommendations, the less the brain function of the patient is damaged, and this piece is precisely the strength of artificial intelligence." Yang Jun said.

Recently, Intel and the analyst firm ConvergysAnalytics conducted a survey involving 200 US medical decision makers. The survey focused on attitudes toward artificial intelligence and perceived barriers to its adoption. Research shows that while most professionals believe that artificial intelligence will be beneficial, one in three said they would not believe in artificial intelligence, mainly because of the fatal mistakes it caused.

In this regard, Professor Yule said: "Compared with autonomous driving technology, artificial intelligence is much less controllable in the medical field. The reason is that the organs and functions of the human body are relatively fixed and will not be like the road conditions of some roads. So complicated."