At the Microsoft Developer Summit in the beginning of the month, CEO Satya Nadella showed the world the direction of machine learning research, in which a company called Jianan Huaxia became the world's first application in the medical field. one's business Patner.
Founded in 2013, Jianan Huaxia, focusing on personalized and precise diabetes management, is the first company in the world to predict blood glucose trends based on CGM continuous blood glucose technology and to fine-tune and manage daily blood sugar. It is currently in pre-A stage.
When it comes to the indissoluble bond with Microsoft, Yang Feng, the founder and CEO of Jianan Huaxia, worked for Microsoft in the early years, doing data analysis. The first version of Microsoft's first balanced score card came from him. In 2008, Yang Feng left the company to start a business. Due to his relationship with the Bank, his entrepreneurial direction also favored medical IT solutions. In 2015, he created the “Medical Follow-up Glucose Kinetics Testing†service.
After the birth of Microsoft's machine learning, it plans to enter the medical field and find suitable projects to be grafted onto the platform to provide services. After three weeks of screening and evaluation, Jianan Huaxia finally reached a cooperation with Microsoft. Diabetes is the largest in the world, with the largest medical expenditures. The CGM continuous blood glucose technology, which is followed up by medical doctors, combines with Microsoft's machine learning to perform the most effective blood sugar management.
CGM refers to the continuous blood glucose monitoring system. Users need to wear a dynamic blood glucose meter for 7 days. The probe has a minimally invasive device on the probe to measure blood sugar by releasing the enzyme.
To put it simply, the patient's use process is such that after wearing the blood glucose meter, it is necessary to faithfully record the specific content and time of eating, taking medicine, exercise. The blood glucose meter measures blood glucose every 3 minutes, 480 a day, and can get more than 3,000 pieces of data a week, they will form a point-like curve. Medical follow-up by controlling the characteristics of the curve (such as peaks and troughs), control variables (diet, drugs, exercise) to return to normal levels.
Medical follow-up can predict the patient's blood sugar changes through machine learning models and monitored food intake. It can also provide suggestions on the type and magnitude of the diet through the machine's conversion of the nutrient content of the food, greatly improving the judgment. Speed, better for patients. According to the perfect data analysis system, after the patient's blood glucose changes during wearing, he can roughly know his future blood sugar changes.
In this way, each patient can know how to eat and how to exercise in order to maintain normal blood sugar, thus achieving the purpose of precise management.
The management of diabetes is affected by many factors, such as diet, exercise, drugs, illness, insulin secretion, and even the time and location of insulin injection. The biggest problem of management is personalization, because each kind of nutrition will have different effects on the blood sugar of different people. What Jianan Huaxia needs to do is to personally analyze the influence factors of each individual.
"Now more is within the self-influence of diabetics, telling patients about the effects of various factors on his blood sugar and guiding." Yang Feng believes that in this process, a lot of data analysis and tools are needed, if Traditional programming or manual methods can be very costly and not very fast. If you use AI to analyze and sort large chunks, different people give different recommendations, then this thing will be much simpler.
“A lot of people are talking about the data model, so is his model exactly what the patient needs, especially if his data model is to say how many copies of the data will be obtained in the future, and what kind of analysis is there, every company will There are different opinions. And we think that machine learning is good, big data is good, it is connected with the company's own services."
Yang Feng believes that the company must first solve the diabetes management itself, and then use software and machine learning to make it automated and then intelligent. “In other words, you have to do a curative service first, then you can automate it through machine learning, and reduce costs and profit through marginal effects.â€
Greenhouse Plastic Plastic Clamps
Greenhouse Plastic Plastic Clamps,Greenhouse Structure Connecting Clamps,Agricultural Greenhouses Clamps,Plastic Clamps
JIANGSU SKYPLAN GREENHOUSE TECHNOLOGY CO.,LTD , https://www.greenhousehydroponic.com