Shulou(Shulou.com)06/02 Report--

Shangyun, the basis of the Application of Medical big data

Shantou University Medical College is located in Shantou City, Shantou is one of the birthplaces of Chaoshan culture, but also a gathering place of Chaoshan people. Chaoshan population originated from Henan and Shanxi in the Central Plains and was formed by migration thousands of years ago. It has the characteristics of relatively uniform genetic background, stable residents, many children, and a large population at home (25 million) and overseas (35 million). It is an ideal place to study the health (genetic and environmental factors) of Han people.

Earlier, the Medical College of Shantou University joined five affiliated hospitals: the first affiliated Hospital (including trusteeship), the second affiliated Hospital, the affiliated Cancer Hospital, the Mental Health Center and the International Ophthalmology Center. The affiliated hospital has nearly 5000-6000 beds, providing 50-70% of medical services for local people.

In recent years, relying on the construction of the information platform, the Medical College of Shantou University has collected a large amount of clinical diagnosis and treatment data, which provides a basis for hospitals and doctors to carry out big data scientific research on patients, and the further mining of these data. it is helpful for hospitals to build a scientific research and analysis platform from form making to statistical analysis reports, and to promote the development of transformational medicine. However, limited by hardware and software and other conditions, the development of hospital scientific research platform is uneven, and scientific research resources can not be shared with affiliated hospitals.

In order to establish an application service platform based on the regional medical data center to provide services for scientific research, medical treatment, management, teaching and public services, the Medical College of Shantou University has built a cloud platform for the integration of "Internet + Medical Teaching and Research". The platform carries four application scenarios of high-performance computing, big data, clinical medicine and science and education, and requires strong computing, storage, agile response and other capabilities.

The Medical College of Shantou University has accumulated a large number of genome data and image data in the process of scientific research, and the affiliated hospitals have also accumulated a variety of medical image data such as CT, nuclear magnetic resonance, X-ray, ultrasound and so on, which need to be preserved for a long time, which will test the storage capacity of the cloud platform.

The fusion and transformation cloud platform carries the scientific research and teaching scene. Taking the dental clinic as an example, the doctor scans the patient's teeth with an instrument, uploads the collected image data to the cloud platform, and carries out 3D modeling through the computing engine of the cloud platform. The parameters after modeling are transferred to the back-end virtual simulation teaching class, allowing students to experience the whole process of diagnosis and treatment. This requires very high timeliness of information transmission, calculation speed and accuracy of modeling.

The integration and transformation of the cloud platform needs to carry the clinical medical business, and the information system of the oral clinic is deployed on the cloud platform. The high availability and timeliness of clinical medical scene services put forward higher requirements for the comprehensive basic capabilities of the cloud platform.

How to design a landing "medical cloud"

Shantou University School of Medicine and Kyushu Cloud have reached a cooperation to build a cloud platform for the integration of "Internet + Medical Teaching and Research". Kyushu Cloud OpenStack open source solution is used to realize the hybrid architecture of virtualization and physical environment, and the virtualization technology, distributed storage and convergence architecture are used to achieve the characteristics of unity, sharing, high utilization and unified management, and can host all kinds of medical business applications.

The implementation of enterprise cloud computing is a complex and systematic project, and the planning, design and implementation process is much more complex than traditional infrastructure projects. After years of cloud data center practice, Kyushu Cloud team has summarized and refined a set of effective planning, design and implementation methods, as shown in the following figure:

According to the above approach, in the top-level design of the big data platform, OpenStack's open source solution will be used for overall design and planning. Make use of the open source feature of OpenStack to realize the vendor-free binding, architecture standardization, strong scalability and customization of the cloud platform.

Cloud infrastructure design will be implemented in detail according to the "overall design solution", mainly considering the specific implementation of the following business requirements:

▌ Computing Resource Pool Design →

Considering the complexity of medical business system, computing resource pool needs to be compatible with the cloud implementation of traditional business, so at the computing level, it can provide virtual machines, physical machines and containers for future Internet scenarios.

For virtualization scenarios, OpenStack is compatible with mainstream virtualization technologies. This project uses KVM as the virtualization implementation technology, and can manage technologies such as VMware and Hyper-V, while providing bare metal management capabilities. In order to achieve the core business of high concurrency, intensive computing requirements. At the same time, it can also accommodate the cloud demand of the old system.

At the same time, it supports the sensitive business in the Internet scenario, and realizes the architecture support of IT two-state. From the perspective of infrastructure management, standardized hardware such as X86 servers and converged architecture servers will be used to eliminate heterogeneity at the hardware level. Simplify operation and maintenance management and hardware selection. It is easy to expand and maintain the platform in the future.

▌ Storage Resource Pool Design →

For medical business systems and data, taking into account the high concurrency and data integration of structured data (EMR, HIS, LIS, RIS, PACS, etc.), as well as the unified long-term preservation of non-institutionalized data (mainly PACS and related image data), different types of storage resource pools are needed.

Storage resource pool uses the current mainstream storage technologies (SDS, AFA, HCI) and commercial storage (SAN) to achieve all kinds of data storage requirements. Through SAN or high-performance AFA to achieve the core high-concurrency business requirements. Using ceph and object storage to realize the storage of hot and cold data of all kinds of business in different pools.

▌ network resource design

Making full use of the current network conditions, the platform can achieve tenant-based network division requirements and security requirements, and can achieve the needs of large-scale user scenarios such as VLAN and VxLAN. Provide soft and hard solutions such as load balancing and firewalls, while supporting the expansion of the network to SDN and the support of traditional networks.

Based on the above design ideas, according to the needs of the business into the pool, the overall capacity planning and hardware procurement.

Cloud platform promotes the transformation and upgrading of medical teaching and research

Based on the cloud management platform and service of Kyushu Cloud Animbus, the Medical College of Shantou University abandoned the traditional "chimney" system construction method, and realized the goals of IT resource pooling and resource sharing, intelligent management and easy-to-operate background.

IT resource pooling and resource sharing: colleges and universities can meet the needs of massive applications without too many physical servers. Converged cloud platform will ensure the application expansion needs in the next 5-10 years, and realize the development of green, environmental protection and energy saving.

Intelligent management: through the visual graphical interface operation and customized management system of Animbus cloud management platform, the Medical College of Shantou University has realized the infrastructure transformation of IT system information, networking, digitalization and intelligence.

Simple and easy to operate background: the "Internet + Medical Education and Research" platform based on Animbus cloud management platform has the characteristics of simple and easy operation, which makes the management flexible and convenient, improves work efficiency and saves a lot of infrastructure funds.

Kyushu Cloud supports the Medical College of Shantou University and its affiliated hospitals to establish an application service platform based on the regional medical data center, and completes the basic construction of the medical big data platform. It also provides computing, storage, network, disaster preparedness and other infrastructure cloud platform. The following medical business objectives have been achieved:

1) based on the unified standard of cloud architecture, it can provide a unified cloud deployment platform for the output construction standards and specifications of hospitals in the Medical Alliance, and reduce the repeated investment and construction of business application systems.

2) support resource coordination and data collection at the medical data center and hospital level, and at the same time serve as a business expansion and disaster recovery supplement for medical entities.

3) support the hosting model: meet the needs of some medical institutions for hosting in regional medical data centers, and some medical institutions that lack the capacity to build can deploy business hosting on the medical data center platform.

Big data has changed the way the industry makes decisions, cloud computing is reshaping the IT architecture, mobile and social technologies have given new meaning to interaction, and the new technology new business type's new model has promoted the transformation and upgrading of traditional industries. The construction of the integration and transformation cloud platform breaks down the industry barriers between Shantou University Medical College and medical institutions, and enables it to achieve better resource sharing, co-construction and common knowledge with local medical information institutions.

Wu Yaqiong, project manager of the bdp project of Shantou University Medical College, said: "next, we hope to dock with the platform of the Health and Family Planning Commission of Shantou City to get through the information systems, electronic medical record databases, and health file databases of grass-roots medical and health institutions, so as to form a closed-loop model of collaborative innovation among government, industry, learning, research, and application, which can provide data support for regional policy formulation in the future."

In the future, cloud computing will help more medical institutions to improve the efficiency of operation and management, medical quality and service ability, and finally realize the construction of "patient-centered" medical information system. As a provider and service provider of cloud solutions in the industry, Kyushu Cloud will help more and more medical institutions to steadily embark on the road of cloud.

About Kyushu Cloud 99Cloud:

Founded in 2012, Kyushu Cloud is an early professional company engaged in OpenStack and related open source services in China. Since its establishment six years ago, adhering to the concept of "open source enabling change", the company has constantly consolidated its own strength, and has realized the expansion and accumulation of business from core cloud to marginal cloud and then to vertical industry applications. it has successively provided high-quality cloud computing services to enterprise customers in government, finance, operators, energy, manufacturing, commerce, transportation, logistics, education, health care and other major industries. At present, it has many heavyweight customers, such as State Grid, × ×, China UnionPay, China Mobile, China Telecom, China Unicom, China Resources Satellite, eBay, International dry Port Group, China Life Insurance, Wanda Information, Dongfeng Motor, Nokia and so on.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.