Embracing large data and speeding up medical information construction

At present, we are in a large data era of data explosion, the wide range of information systems in the medical and health institutions and the gradual digitization of medical equipment and instruments to make hospitals accumulate more data resources, these data resources are very valuable medical and health information, for disease diagnosis, treatment, The control of medical expenses is very valuable. How to do well in the large data age of medical and health information construction, it is worth our thinking.

New requirements for medical information construction in the age of large data

It is estimated that the amount of medical data accumulated in a medium-sized Chinese city (1 million population) in 50 will reach the 10PB level. With the rapid development of medical informationization in China and the explosive growth of medical and health data, the medical and health field has ushered in its own "large data Age", and the large data-related characteristics have put forward new requirements for medical information construction.

(i) The explosive growth of data volume requires that the medical information architecture be expanded

Medical and health data mainly comes from three aspects:

1. Life Science field and medicine Research and development field

In the field of life sciences, with the development and gradual application of high-throughput sequencing technology, the data volumes in the Life sciences field are growing at a high speed, with each high flux sequencing device producing about 100GB of data per day, and only one Chinese gene company, China's largest gene, has more than 100 such sequencing machines running at full capacity every day, produces 10TB of data. In the field of pharmaceutical research and development, drug research and development is a very intensive process, a small and medium-sized pharmaceutical Enterprises in the annual cumulative amount of data is more than TB.

2. medical field

With respect to medical organization data, diagnosis and treatment data structure complex contains a large number of semi-structured or unstructured data, a single semi-structured data (such as ECG, B-Ultrasound, CR, CT, etc.) the amount of data is much larger than a single structured data (such as XML documents), such as a general CT image about 150MB, A standard pathological figure near 5GB, and an XML document size of about dozens of K, with the medical institutions of information construction to the clinical information system, more and more hospitals will focus on the construction of PACs, LIS system, resulting in a large number of unstructured diagnosis and treatment data.

Medical field Large data in addition to medical institutions, patients in the process of medical institutions will also produce registration, payment, the use of new agricultural and health funds, medical insurance funds use, diagnosis conclusions, treatment process and other data. According to the "Health information Management Method (Trial)" promulgated by the Wei-planning authority in 2014, the population health information such as electronic health files, electronic medical records, full population information needs to be preserved for a long time, and the amount of data in medical field will accumulate every year.

3. Mobile medical Field

With the rapid development of mobile Internet and IoT technology, wearable medical equipment is gradually moving from concept to reality. Wearable equipment is the most important application in the field of health management, from Google Glasses, Apple Watches to Nike wrist strap, etc. have a certain health management functions, to provide users with physical fitness indicators monitoring, disease data tracking and other services. With the rapid development of wearable equipment, a large number of medical and health data will be produced in the future.

In the face of rich and growing sources of medical and health data, at present, the storage architecture of medical informatization can not meet the needs of large data application, when processing and querying large data sets, it is necessary to design new data-centric computing model and system architecture, and to separate, disperse and Different brands or different levels of storage products unified into one or several large storage pools, the formation of a logically unified whole, and in accordance with data integration or application integration needs to migrate data to the corresponding storage space, so as to achieve the medical information storage architecture in the unified planning and deployment.

(ii) High data value, requiring medical information construction to pay attention to data mining

Medical and health data structure is more complicated, in addition to the ordinary structured data, most of the semi-structured or unstructured data, such as ECG, B-Ultrasound, CT, Mr, CR, and other clinical image files are unstructured data. In the era of "big Data", the traditional database analysis system is facing a historic change.

There are already many institutions abroad that have begun to delve into the use of medical data, and have found potential health-related benefits from large data, such as: a few weeks before the outbreak of swine flu in 2009, Google developed a "Google Flu trend" extension service, By analyzing a large number of users ' search records for flu-related entries to identify the flu outbreak, Google could predict the outbreak 1-2 weeks ahead of the official agency, with a 97% per cent correlation between the results and official data. Apple's former president, Steve Jobs, has also applied large data technologies in the fight against pancreatic cancer, he was the first person in the world to sort all his DNA and tumor DNA, and by analyzing large data files on the entire genetic data, jobs ' doctors were able to use drugs on demand based on his specific genetic makeup. Data show that patients with pancreatic cancer have a very high mortality rate, the average life expectancy is only 9 months, 5 year survival rate of less than 2%, but jobs through large data technology developed by personalized drugs to prolong life for several years.

Compared with the development of data mining of medical health in foreign countries, China's medical and health data mining is only just beginning, most of the medical and health institutions remain in the data accuracy level, rather than from the data correlation analysis of data mining value, need to increase the input of large data analysis of health care, through the effective storage of medical and health data, processing, Enquiries and analysis, assist doctors to make more scientific and accurate diagnosis and medication decision-making, to help hospitals in accordance with the potential needs of patients to develop new personalized services and automatic services, to help relevant research institutions to break through medical methods and drug innovation, support the region and even the national medical industry authorities to optimize medical resources and service allocation.

Suggestions on the construction of medical informatization in China in the era of large data

Medical information construction in large data age from the industrial point of view, the future will create a patient-centric, rapid data flow, accurate analysis of the value chain. In this chain, the government, medical and health institutions, medical information, software and hardware providers should clearly identify themselves and cooperate in order to do a good job in the large data era of medical information construction.

(a) Government should play a role in policy formulation and industry development

The government mainly plays the role of policy making and industry development in the medical information construction. At present, the construction of medical informatization standards in China is not complete, the basic electronic medical record and the standard of electronic health file construction have been promulgated, but there is no standard for data collection, transmission, processing and related medical information architecture in the construction of medical informatization in the age of large data.

(ii) Medical and health institutions should look forward to the deployment of medical information construction

Medical and health institutions are the main body of medical information construction, in the process of medical information construction should have a forward-looking awareness, the courage to the next generation of information technology into the process of medical information construction. At present, some of the three hospitals have carried out such as mobile medical, telemedicine and other attempts, but for large data in the medical and health institutions are still more conservative application, still stay in the BI with large data-assisted office, for clinical precision medication less, in the future with large data-related technology gradually mature, The medical and health institutions should gradually expand the large data from Assistant office to improve the quality of clinical medical service.

(iii) ICT manufacturers should increase upstream and downstream cooperation, aiming at large data medical information construction actively develop new products

Software and hardware manufacturers, as providers of medical information technology, should aim at the new opportunity of medical information construction under the large data to develop products, increase the upstream and downstream industry chain cooperation, such as the production of wearable equipment for a large number of health monitoring data, medical information software and hardware manufacturers can work with downstream health institutions, Putting medical services in front of the patient so that patients can discover possible health risks as soon as possible, even at home, through the intelligent terminal can regularly view the health information recorded in wearable equipment, and the medical institutions of the relevant doctors to give medical and health diagnosis information and advice, to achieve "treatment of disease" to the "treatment of disease" transfer; due to the small size of wearable equipment , thin characteristics of battery endurance and CPU, wireless transmission chip power to put forward more stringent requirements, medical information software and hardware manufacturers need hardware and software vendors and upstream processor, memory, power, wireless communications, soft board, sensors, actuators and other major parts and components manufacturers to cooperate, Jointly develop wearable medical products with low power consumption and high function.