In recent years, digital health management has been accelerating. With the rapid development of the market, it is also faced with problems such as single product functions and service content. Industry standards need to be improved, and data cannot be interconnected between institutions.
“The foundation of China’s health management industry is relatively weak, and there are still many problems. For example, relevant laws and regulations are not perfect, and the development of professional education is uneven. Moreover, physical examination cannot be connected to other health management links effectively. This leads to the problem of single product and service in the health management industry.” Xiaodong Ni, chairman of Hangzhou Seehelath said. The vigorous development of the health management industry has become a general trend, and all parties in the industry need to find a development path that suits them and adapts to the times.
In 2008, Hangzhou Seehealth Information Technology Co., Ltd. was established, which is the predecessor of Seehealth Group, the parent company of Seehealth. Seehealth Group mainly uses medical knowledge base and artificial intelligence as its core technologies, and provides smart health management platforms, regional health management platforms, digital whole-course specialized disease management platforms, smart hospital follow-up platforms, digital commercial insurance platform and other product systems for large and medium-sized hospitals, Health Commissions, insurance companies and other customers. The company has served more than 1,000 institutions, among which more than 800 are hospitals, covering 31 provinces across the country.
Seehealth is a medical and health technology company incubated and independently operated by Seehealth Group, focusing on disease health management and digital therapeutics. At present, the company is building a platform that combines an online disease health management platform and an in-hospital disease management platform. The platform integrates online and offline services for patients, covering the whole course of disease management both in-hospital and out-of-hospital. In addition, Seehealth also seeks to cooperate with hospitals to accumulate patient disease management data actively and explore innovative applications of digital therapeutics that follow evidence-based medicine.
Relying on the industry resources accumulated by the parent company for many years, Seehealth has advantages in building health management networks in-hospital and out-of-hospital, and also lays a foundation for future product market promotion and innovative application of digital therapeutics based on evidence-based medicine. The realization of the functions of the two disease management platforms of Seehealth is inseparable from the support of medical knowledge and artificial intelligence technology.
The “Health Brain” combining medical knowledge with AI
Seehealth has two supports: medical knowledge base and artificial intelligence technology. And these are integrated into the Health Brain of Seehealth and become the technical backbone for it to carry out most of its business.
Specifically, Seehealth’s Health Brain has built-in medical knowledge graph, disease assessment models, disease management solutions and intelligent question answering base. It can realize the establishment of health records, health assessment, monitoring and early warning, intelligent follow-up, specialized disease management, health intervention and other functions.
The construction team of medical knowledge graph of Seehealth ‘s Health Brain consists of more than 10 members with medical background. After years of accumulation, the diseases covered by this knowledge map involve cardiovascular, respiratory, endocrine, oncology, obstetrics and gynecology, pediatrics and other departments. The main sources of these medical knowledge are international and national medical standard terminology, authoritative medical textbooks, the latest domestic and foreign clinical pathway guidelines, clinical practice documents and medical encyclopedia, etc. The sources of some diseases also include clinical data after desensitization in cooperative hospitals.
After acquiring the initial medical knowledge, it will first perform structural processing such as attribute classification. Second, it extracts, normalizes, and medically labels the obtained structured or semi-structured data. Finally, it realizes the construction of medical knowledge graph through data fusion methods such as entity segmentation, tagging, semantic association, and entity association. In addition, Seehealth will also carry out multiple manual verifications of the medical knowledge graph data, and check the labeling conflicts and suspicious results in the labeling results comprehensively. If there are inconsistencies in entity words, incorrect attributes, incorrect semantics and relational sequences, or inconsistencies, it will be fed back to the expert team for further review and corresponding corrections. After laying the foundation of the medical knowledge graph, Seehealth has formed more than 200 evaluation models covering more than 200 common diseases, mainly including cardiovascular diseases, metabolic diseases and tumors.
According to Xiaodong Ni, many public disease assessment models currently in the market mostly use simple single-factor weighted methods or multi-factor analysis methods. Seehealth’s disease assessment model adopts the comprehensive analysis method of all factors. The evaluation parameters include disease-related risk factors, and predict the probability of a user’s multiple chronic diseases in the next 5-10 years comprehensively. In addition, regional differences, ethnic differences, dietary and cultural differences were taken into account in the disease assessment model, and regional modification parameters were added. Xiaodong Ni said: “This is our unique set of risk assessment model.”
Based on the medical knowledge grraph and disease assessment models, Seehealth adheres to the concept of formal medical care and forms a personalized and customized disease management plan. The personalization not only refers to different departments, but also refers to different courses of the same disease.
Up to now, the departments covered by Seehealth’s disease management program include cardiology, respiratory, digestive, endocrine, oncology, obstetrics and gynecology, and pediatrics. In addition to the disease health management of patients with the same disease, Seehealth will also form a personalized plan based on different treatment measures, different populations, different causes of disease and other differentiated characteristics.
In addition, all of Seehealth’s disease management programs not only include health education, diet guidance, follow-up guidance and tracking, etc., but also include risk assessment of diseases and complications, medication management, psychological intervention, TCM physical conditioning, rehabilitation guidance and other management projects with strong medical attributes.
The medical intelligent question answering base is one of the application forms of medical knowledge graph. The system will first identify the patient’s questions in natural language to convert them into structured sentences, and then give relevant answers to the questions from the knowledge base according to the algorithm.
In addition to disease knowledge, Seehealth also added the application of psychology, nutrition, exercise and other knowledge sectors in the medical intelligent question answering base. It can be applied to multiple scenarios such as medical knowledge popularization, intelligent guidance, health risk warning, and health care.
Seehealth ‘s Health Brain, which integrates medical knowledge graph, disease assessment models, disease management solutions and intelligent question answering base, can replace at least 60% of items that require manual operations, such as automated data collection, intelligent disease assessment, intelligent solution matching, and personalized health education. This can effectively improve the efficiency and quality of health management.
“More than 200 disease risk assessment models, more than 500 disease health management solutions, medical knowledge graph and intelligent question answering base are the core competitiveness of Seehealth after 14 years of development.” Xiaodong Ni said,” we are one of the few companies in the industry that has both clinical knowledge and informatization, and this is also our core competitiveness in the market in the future.”
Innovative application of digital therapeutics relying on evidence-based medicine
When talking about the development plan of Seehealth, Xiaodong Ni said that the short-term goal of Seehealth is to seek cooperation with hospitals on the health and disease management of discharged patients. The long-term strategy is to cooperate with hospitals to accumulate patient disease management data in order to explore innovative applications of evidence-based digital therapeutics in the future.
The essence of digital therapeutics is the digitization of services. It is to integrate the doctor’s experience into the software, and carry out data accumulation and continuous optimization. Ultimately, all service content that can be digitized will become digital therapies.
Although digital therapeutics cannot completely replace human services, such as surgery. However, the base of the domestic medical and health market is huge. And with the accumulation of data, the optimization of technology, and the improvement of market acceptance, its development potential will be even greater. Therefore, a large number of companies have poured into this track.
Seehealth will carry out researches on digital therapeutics for child growth and development, child precocious puberty, child language development, gestational diabetes, apnea syndrome, and menopausal syndrome. Among them, the cooperative institutions of digital therapy programs for child growth and development, child precocious puberty, child language development, and gestational diabetes are Zhejiang Chinese Medical University and Jiaxing maternal and child health care hospital. The partner institution of the digital therapy program for apnea syndrome is the Fourth Affiliated Hospital of Zhejiang University School of Medicine. The planned partner institution for digital therapy for menopause syndrome is the Obstetrics and Gynecology Hospital affiliated to Zhejiang University School of Medicine.
On the track of digital therapeutics, Seehealth’s performance still needs time to be verified.