Since the 1960s, artificial intelligence has experienced three ups and downs. At present, artificial intelligence is experiencing a new round of outbreaks, and various industries are actively exploring and developing artificial intelligence. "Artificial Intelligence + Radiation Therapy" enables high-quality medical resources in Beishang to sink at the grassroots level; "Artificial Intelligence + New Drug Development", micro-problems in drug development are revealed; "Artificial Intelligence + Diagnostic Assistance" saves time At the same time, it can reduce the error rate of diagnosis. Since 2015, the wave of artificial intelligence has come back. In the medical field, "artificial intelligence + image" took the lead. Despite occasional doubts, we can't deny that this wave of artificial intelligence is sweeping the entire industry. After the "artificial intelligence + image", the "artificial intelligence + new drug research and development", "artificial intelligence + diagnostic aid" has come. At Huaxing Capital’s annual Medical and Life Technology Leadership Summit, “AI†has become a hot topic of discussion without any suspense. After Alpha Go successively defeated Li Shishi and Ke Jie, society and industry began to think, in the medical field, will artificial intelligence eventually replace doctors? Industry experts from Insilico Medicine, Honghui Capital, Global Medical, Jingtai Technology, and Senyi Intelligent discussed the application scenarios, application thresholds, and data attribution of artificial intelligence in the medical field, hoping to find out the answer. AI in primary care: the downward force In the view of Kang Shigong, co-founder and vice president of the global medical industry, the two most typical application scenarios of AI in the field of radiation therapy are resource sinking and quality control. The AI ​​technology enables the radiotherapy resources of the three hospitals in Beishang, Guangzhou and Shenzhen to sink to the grassroots level. "Most of the grassroots patients' financial ability is not enough to support him to go to the big cities to find oncologists," he said. "The addition of AI technology is like putting a sharp-edged scalpel at the grassroots level." Of course, quality control of AI is especially important in this process. Through big data, all the conditions that lead to poor quality control are judged and collected one by one, and models and algorithms are used to automatically monitor and guide the operation of the primary radiotherapy. The development of AI+ new drugs began to move toward the cusp in 2017, and the entry of artificial intelligence made microscopic problems in drug development and development. Helping drug development, going forward "It is a tool, like a high-resolution microscope." Talk about the application of AI in the development of new drugs, said Li Lipeng, co-founder of Jingtai Technology and head of the Big Data and Artificial Intelligence R&D Center. In the process of drug action, how to combine small molecules and proteins is difficult to imagine by human resources alone, but AI can learn and discover the laws through a large amount of data. In Lai Lipeng's view, the application of AI in drug development can be described as “continuously comingâ€. The large amount of data accumulated in the past contains failed data, but the so-called failure is actually a clinical failure, and does not mean that the data is worthless. “Continuing to go†means extracting information that has not been noticed in the past from past data through statistical and machine learning methods. In the “open†section, models based on deep learning can help developers explore larger chemical spaces and do more groundbreaking work. For example, an article in the 2017 "Nature" mentioned that the chemical space of the drug can be 10 to the 60th power, but the physical molecular library that can be realized at present is about 10 to the 13th power. “There are still more than 40 orders of magnitude difference here.†Lai Lipeng said, “The drug molecules that are really researched in the laboratory are just the tip of the iceberg in the entire drug space.†The addition of AI technology can be directed in a huge space to find The drug molecule we need. “In addition, I think AI can reshape the workflow of new drug development,†he added. “Many existing methods are now impossible to cover the complex system of living things.†For example, in the first-stage clinical toxicity of drugs, the same drug may have different effects in humans and animals. AI emphasizes end-to-end prediction, and it is very promising to directly predict a series of toxic and side effects of drug candidates on the human body based on molecular structure and other conditions, greatly reducing the possibility of drug clinical trial failure. “The more important point is in the development of methods for drug discovery and crystal prediction,†he continued. At Jingtai Technology, they continuously improve the computational efficiency through the combination of physical models and AI models. AI and diagnostic aids: instead of manual repetitive work The place where Senyi intelligently chose to fall is the landing of AI in the auxiliary medical treatment. Venous thromboembolism (VTE) is common in patients who have been hospitalized for a long time, have been hospitalized for a long time after surgery, or have been bedridden for a long time after childbirth. This disease has a certain probability of developing pulmonary embolism, and the mortality rate of pulmonary embolism is very high. In order to prevent the patient's risk, the hospital will send nurses daily for manual monitoring, and to track and feedback the patients. However, relying on manual methods consumes a lot of time and effort, and errors may occur. "AI technology can improve this situation." Zhang Shaodian, founder and CEO of Senyi Intelligence, told the Arterial Network. The medical AI products developed by Senyi Intelligent have two functions: First, the assessment will automatically score the patients according to the patient's condition, medical history, including hospitalization, surgery, and inspection and examination. It is an early warning, based on the patient's data to predict whether it is possible to be a high-risk patient. If it is a high-risk patient, an alert message will be sent in advance and pushed to the doctor's workstation. What is the role of such a product? Zhang Shaodian revealed that in the two months of cooperation with the top three hospitals, these products can help to ignore the time of 95% of the manual evaluation of patients, and can also increase the recognition rate of high-risk patients by 70%. “This is one of the examples we have tried in the field of assisted treatment,†he added. How does the data threshold cross? But whether it is drug development or diagnostic aids, or radiation therapy, AI technology is always inseparable from big data. In a messy, varied data, how do you get better data and produce better results? Perhaps you have to set the threshold before data collection. “Labs are the most basic scenario for data generation, but the cost of getting data here is high,†said Artur Kadurin, chief AI officer at Insilico Medicine. Through early efforts, Insilico Medicine found a way to obtain basic experimental data, and further data can be obtained. Kadurin believes that a large amount of data can be obtained in China, which cannot be done overseas. But Insilico Medicine's purpose in coming to China is not just data, they hope to have a longer development in China, and even in Asia as a whole. “China is a very huge market and we want to be able to work with local partners,†he said. Lai Lipeng agrees with Kadurin's point of view that the cost of obtaining highly accurate data is not low. Jingtai Technology's early development of drug development tools mainly comes from two aspects: one is public domain data; the other is that it comes from its own high-precision data. The amount of public data is relatively large, but the cleaning process is very laborious due to the uneven quality and format of the data. The accuracy of the internal high actuarial calculation is very high, but because of its scale, it can reach hundreds of millions or billions, and the cost of acquisition is not low. With the deep cooperation with customers, Jingtai Technology also obtained some data from partners. These data are the closest to first-line R&D and specific issues. However, since some of the data is not collected for AI modeling, it is possible that the key information is not fully recorded. Data is the foundation, but ownership is not in the enterprise "There is no doubt that the data is the most basic thing in terms of AI technology," added Hong Hui Capital Partner. In addition to being processed and institutionalized, these databases are also capable of field extraction and understanding, more intelligent puzzles, and recognition. This is hard to find in China at this stage. "If you don't accumulate these structured data, it's hard to get a strong diagnosis without basic things," he continued. He revealed to the arterial network that when looking at the enterprise, the capital pays great attention to the source of the enterprise data, whether it is legal, whether it has been adequately desensitized during use, whether the hospital’s rights and patient privacy are well protected. . "We believe that technology companies should not think about owning data ownership." Zhang Shaodian agreed. He believes that technology companies should base themselves on their core technologies and capabilities, fear technology, provide better solutions for hospitals, and then use part of the data in the hospital through the product. But the most important challenge facing AI companies today is the issue of data governance. Why is the first image landing artificial intelligence? Because this type of data is relatively standard. The data of medical records and medical products requires a lot of data management. For example, Senyi Intelligent's VIE early warning monitoring system for the top three hospitals. With this system alone, Senyi Intelligent has docked more than 20 systems in this hospital. "The data structure, standardization and data governance behind this is a huge project." Zhang Shaodian revealed to the arterial network. As for whether AI will eventually replace humans, replacing doctors or nurses in the medical system can actually be answered from the above. What AI is doing now is to help hospitals save unnecessary resources and experiences, such as replacing redundant labor, shortening time, increasing output, and reducing misdiagnosis. At the drug development end, the role of AI is to improve accuracy and reduce material loss. Therefore, no matter from what angle, AI will be the method and tool for reducing costs, improving efficiency and precision of the entire medical system. "In a short or even long period of time, AI is unlikely to really replace humans. But it can be a very good helper." He said.
Salted Shredded Kelp
The so-called original ecological kelp means that we pack the young kelp with saturated salt water directly after washing without any preservatives.Original ecological products are the favourite of housewives.
How to eat:
The difference between deep sea kelp and traditional ordinary kelp
Nutrational Effects
Salted Shredded Kelp,Shredded Seaweed,Wakame Kelp,Organic Sea Kelp Shandong Haizhibao Ocean Science and Technology Co.,Ltd. , https://www.haizhibaoseafood.com
