In the AI+ medical field, the most common form of landing is medical imaging, which uses AI to assist doctors in reading image information. However, in the field of text processing, there are still fewer incoming people.
Ronton Medical cuts into the field of assistive medicine with text analysis, which can help general practitioners quickly complete outpatient medical records and achieve clinical diagnosis. Specifically, when the doctor fills in the “medical complaintâ€, “medical historyâ€, “checking body†and other information in the electronic medical record system, the system will automatically recommend the items to be tested (such as blood routines), the name of the suspected disease, and the treatment plan. Wait. After completing the laboratory test, the doctor can also complete the test list interpretation through AI to provide a basis for the diagnosis.
AI-assisted diagnosis and treatment: laying foundation for graded diagnosis and treatment system
The background of the birth of this intelligent auxiliary diagnosis system is that the country is vigorously promoting the grading diagnosis and treatment system.
The imbalance in the distribution of medical resources in China is a well-known fact. Large hospitals are expensive to see and see a doctor. First-tier cities have even spawned registered yellow cattle, and the patient experience is poor. For the hospital, it is also a big problem for medical staff to work over load, leading to misdiagnosis and missed diagnosis, and reducing the average level of diagnosis and treatment. The problems in primary hospitals are equally obvious, the doctors are not qualified enough, and the level of diagnosis and treatment is low, so it is difficult to harvest the trust of patients.
The solution proposed by the government is a grading diagnosis and treatment system, “the first diagnosis at the grassroots level, two-way referral, rapid division and treatment, and linkage between the upper and lower sidesâ€. After the initial diagnosis by the primary hospital, the serious cases will be transferred to the large hospitals; at the same time, the patients who have recovered in the top three hospitals will be transferred back to the primary hospitals to release medical resources. In August this year, the State Council issued the "Guiding Opinions on Promoting the Construction of Graded Diagnosis and Treatment System" to promote the development of the triage system from the policy level.
But the so-called grading diagnosis and treatment, there is still a dead knot, that is, the basic doctors have insufficient diagnostic ability. Faced with a large number of illnesses, many of the conditions are similar in external performance, and the grassroots doctors sometimes cannot judge the condition smoothly.
The entrepreneurial concept of Langtong Medical is to improve the ability of grassroots doctors through AI-assisted treatment, which currently covers 2,500 hospitals.
Take the data of Shao Yifu Hospital and launch the deep learning artificial intelligence
Before exploring the field of artificial intelligence, Langtong Medical's business was to build a new rural cooperative business system and data center platform for government (the government-organized rural medical social security). From 2004 to 2014, Langtong Medical Services has served 80% of the Zhejiang New Rural Cooperative Market.
CEO Xu Zhe decided to transform in 2015, he is ready to combine electronic health records with auxiliary medical treatment, the middle of the adhesive is artificial intelligence.
At that time, there was no corresponding product in China, and Langtong Medical cooperated with the British Jiukang Medical and completed the overall framework. Xu Zhe recalled, "At that time, we invited British technical experts to guide the development of the framework. They originally hoped to have a joint venture, but considering that we would have access to massive clinical data later, we decided not to introduce foreign capital."
“Our core is the medical knowledge base and artificial intelligence inference algorithm. An efficient auxiliary diagnosis and treatment system needs to integrate the prior knowledge of the medical field and the actual case data. Our knowledge base is based on this idea. The AI ​​reasoning algorithm Based on the knowledge base, it realizes the functions of natural language processing, intelligent reasoning diagnosis, providing inspection programs, and pushing treatment programs." Xu Zhe told Hunting.com.
Once the framework is built, the next step is to collect the data and guide the artificial intelligence to complete the iteration through deep learning. Xu Zhe found Shao Yifu Hospital. The other party opened the desensitization data for 10 years to Langtong Medical. After eliminating the invalid data, a total of more than 1.8 million medical records were collected. Up to now, nearly 26w cases have been imported into the system. Covers 739 common diseases.
In addition to covering professional terminology, the medical knowledge base also summarizes the natural language expressions that doctors may use, and synthesizes semantic expressions such as "abdominal pain", "abdominal pain", "stomach pain" and so on. According to reports, the update of the knowledge base will continue.
At the end of 2015, Langtong Medical launched the Diagbot, a clinical assistant decision-making system based on deep learning. It began to be used in 2016. After small-scale trials and product optimization, the products were basically mature at the beginning of this year.
The system also implements the interpretation of text test sheets. According to Xu Zhe, the test results on the past test sheets were actually judged and entered by humans. Some indicators that were obtained from different tests by different patients exceeded the standard, which may be caused by different causes. The related reasoning of Langtong Medical can be used to interpret the test list for patients. In addition, some illnesses that are not well understood by the grassroots doctors can be viewed by clicking on the entry on the right side of the interface.
Ronton Medical is not prepared to do its own application. It prefers to be the underlying service, implant other medical applications or directly access the hospital's medical system.
At present, there are more than 70 people in the team, and more than 30 of them are technicians in the field of artificial intelligence. In 2016, the company received tens of thousands of Pre-A rounds of financing from Cybertron.
Full Body Massage,Body Massage Cost,Perfect Body Massage,Neck And Shoulder Massager
Shenzhen Jie Zhong Lian Investment Co., Ltd. , https://www.meizons.com