English Chinese
Hospital-Enterprise Collaborative Innovation State Key R&D Program DHC launched CDSS with top hospitals in China
2022-08-05 Share:

As an important achievement of the National Key R&D Program during the 13th Five-Year Plan, DHC and Peking Union Medical College Hospital jointly developed the Critical Care Clinical Decision Support System CDSS. It is worth noting that this research is an innovative model of hospital-enterprise collaboration.

Forging a new AI clinical assistance model in critical care first

The Department of Critical Care Medicine at Peking Union Medical College Hospital has been known as the birthplace of critical care medicine in China in this field. The establishment of this hospital has played an indispensable role in the earliest dissemination of critical care medicine and the standardization and rapid development of the discipline since then.

For the national ranking of critical care specialties, Union Hospital still tops the list until today. When the top specialties of the top hospitals are combined with the solid accumulation of domestic medical big data companies, it must bring great changes to the industry.

In August 2021, the Concordia Critical Care Big Data and Artificial Intelligence R&D and Clinical Decision Support Seminar and Training Conference was held. The Department of Critical Care Medicine and Information Center of Concordia Hospital, jointly with DHC, announced the Critical Care Clinical Decision Support System.

The system follows the standards proposed by the China Critical Care Primary Care Diagnosis and Treatment Process, and also provides real-time multidimensional and intelligent clinical auxiliary decision support based on the knowledge base of critical care medicine with Chinese characteristics. It satisfies the demand for critical care diagnosis and treatment in primary care institutions, and addresses the problems of standardized diagnosis and treatment and quality control management of critical care in a targeted manner.

Also, the system is an important achievement of the "13th Five-Year Plan" National Key Research and Development Program "New Service Model Solution of Clinical Assisted Decision Support Based on Artificial Intelligence" project.

The project is undertaken by DHC, with Shi Wenzhao, Chairman and CEO of the company, as project leader, and Peking Union Medical College Hospital, Chinese Academy of Medical Sciences, Cancer Hospital of Chinese Academy of Medical Sciences, Fudan University, Sun Yat-sen University and Zhejiang University as project participants.

In July 2022, the project successfully passed the acceptance of the Ministry of Science and Technology of the People's Republic of China, forming a knowledge map based on clinical guidelines and medical literature and oriented to the functions of clinical diagnosis and treatment assisted decision-making and clinical event inference and prediction. It is a knowledge database for diagnosis, treatment and rehabilitation, as well as a cloud platform for clinical decision support covering the whole process of consultation, initial diagnosis, diagnosis and treatment. It creates a new clinical assistance model for the closed-loop process of medical treatment.

The project focuses on providing decision support recommendations to physicians based on the principles and body of evidence of evidence-based medicine. Its compliance with current clinical acceptability contributes to the efficiency of physicians. Through machine deep learning and big data mining, it automatically maps SNOMED-CT, LONIC, and ICD-11 to transform unstructured and semi-structured medical record data into more application-worthy clinical decision and research information. The highly structured and standardized nature of this part of data lays the foundation for the accuracy of the system.

As an important result of the project, CDSS implements the model into a standardized WEB application in a configurable way in terms of technical functionality.

The standardized WEB application implements machine learning models or other complex decision models by collecting model input information, executing the core logic of the predictive model, and visualizing the model output.

Moreover, this application supports standardized diagnosis and treatment opinions for severe infections and shock and offers a source basis. Its standardized terminology system and common data storage model can flexibly carry out the development of a variety of machine learning and big data analysis models based on clinical data.

Clinical diagnosis and treatment is mainly to provide a variety of rule engines and AI algorithms, which can realize the auxiliary diagnosis and treatment of complex business processes such as the calculation of common ICU scores, so as to process and respond to patient data in real time.


Clinical experience for many years to summarize highly suitable for the demand of primary critical care treatment 

The role of critical care medicine in the treatment of critical illnesses at the beginning of the new pneumonia outbreak is obvious to the public. However, the fact cannot be ignored that the strength of critical care medicine in primary care institutions is still weak, and there is still much room for improvement in the standardization and homogenization of critical care treatment.

The intensive care unit brings together a variety of highly sophisticated equipment for life support. Each patient is a collection of massive amounts of information. Therefore, how to handle the data and tap the value of the data becomes a challenge that must be overcome. DHC has undertaken several national big data platform experiences, which provides a great boost to the project advancement.

Based on previous research experience, DHC followed three principles when developing the critical care CDSS in collaboration with Peking Union Medical College Hospital. First, the principles and evidence system of evidence-based medicine provide physicians with recommendations for decision support, ensure clinical acceptability, reduce and eliminate false triggers and missed triggers, and enhance physician productivity. Second, the high degree of structured and standardized data can provide a solid foundation for system accuracy. Third, a data-driven clinical decision support system based on artificial intelligence technology will be further explored in the future, with the exploration of knowledge and evidence mining as the main direction.

Industry-academia-research-medicine collaborative innovation and hospital-enterprise cooperation towards the future

The cooperation between DHC and the clinical expert team and information team of Union Medical College Hospital focuses on the critical care field to launch specialized products. This is not only a focused demonstration of the advantages of clinical decision support system (CDSS), but also a model of collaborative innovation between industry, academia, research and medical institutions, and joint scientific research, technical research and clinical application between medical technology enterprises and medical institutions.

According to evidence-based medical evidence and complete data analysis, CDSS provides real-time decision support for healthcare professionals in clinical applications, assisting in optimizing treatment plans, automatically reviewing the rationality of disposal orders, etc., as well as providing personalized treatment recommendations for patients' conditions.

The CDSS system has a built-in machine learning system and knowledge base system The customer can conduct machine learning on the data through the machine learning algorithm library and train the model. The proven models are then applied to the clinic to assist doctors in making diagnoses.

In parallel, customers can assist physicians in making diagnoses from a knowledge base summarized from known guidelines.

In the Concordia Critical Care Program, CDSS supports applications such as recommendation of similar patient records and intelligent guidance support, and has gained hospital recognition.

The degree of standard of the knowledge base directly affects whether the CDSS diagnosis and recommendations are accurate, derived from guidelines and serving the clinic. Its CDSS, which combines Shenzhou's medical AI and medical big data technology, has the ability to self-learn, accelerate knowledge updates and iterations of computing models, and continuously improve the degree of standardization of the knowledge base.

As an innovative enterprise of medical big data and artificial intelligence, DHC believes that the tight combination of critical care medicine with artificial intelligence and big data technology will collide with even more brilliant sparks in the future. The new direction of this combination will promote the breakthrough innovation of modern critical care medicine, transform the power of science and technology into social temperature, and contribute stronger power to the confrontation between human beings and diseases.