CARES Copilot

CAIR
Overview

CARES Copilot is designed to meet the intelligent needs of doctors and build AI intelligent assistants based on a large language model. It provides support in a variety of medical scenarios, including assisting in diagnosis, automated reporting, data collation and collection, and helping physicians understand the surgical process to improve productivity and empower medical practice.

  • CARES Copilot 0
  • CARES Copilot 1
  • CARES Copilot 2
Research completion
released Ver2.0, opened trial accounts for 100+ surgeons
Commercialisation opportunities
1 SaaS (Software-as-a-Service) subscription model2 AI Medical Device Integration (Embedded AI Solutions)3 Medical AI Model-as-a-Service (MaaS)
Problem addressed

At present, there is no large-scale model system designed for doctors, who are faced with heavy tasks of diagnosis and treatment, report writing and data sorting in their daily work, and the existing tools are difficult to meet the professional needs of the medical industry

Innovation
  • The advanced continuous learning method is adopted to alleviate the "catastrophic forgetting" problem that may be caused by fine tuning of the base large model, ensure that the new medical knowledge will not affect the original ability, and achieve stable and efficient knowledge update.
  • The model is made to reason according to the logic of evidence-based medicine, actively retrieve high-quality medical evidence, and support diagnosis and treatment recommendations based on authoritative data to ensure that doctors get intelligent assistance based on the best evidence in the clinical decision-making process.
  • Combined with the knowledge base of medical experts and real-time search engine, the model can not only ensure authoritative and reliable medical knowledge, but also continue to supplement the latest research results, and improve the medical reasoning ability and information accuracy.
Key impact
  • With its powerful natural language processing and medical knowledge integration capabilities, CARES Copilot provides efficient support for assisted diagnosis, medical record report writing, case analysis, and more, reducing doctors' repetitive work and allowing them to focus on higher-value medical decisions.
  • Combined with a knowledge base of experts and a search engine, CARES Copilot dynamically updates medical information and maintains efficient access to the latest research, clinical guidelines and medical evidence, ensuring physicians receive the latest, most reliable medical support.
  • Through evidence-based thinking, CARES Copilot is able to reason based on clinical trial data, authoritative guidelines and medical papers to help doctors make the best decisions in line with medical evidence and improve the science and safety of diagnosis and treatment.
  • CARES Copilot can be customized and optimized according to the needs of different departments and specialists, combined with connection learning technology, to ensure that the model continues to adapt to new knowledge while maintaining existing capabilities, providing doctors with more accurate, efficient and personalized intelligent support.
Application
  • Combining medical records, image data and the latest medical research, it provides accurate diagnosis suggestions to help doctors improve diagnosis efficiency and accuracy.
  • Quickly generate medical records, examination reports, surgical records and other documents, reduce the writing burden of doctors, and improve the quality of medical documents.
  • Integrate expert knowledge base and search engine to provide real-time updated evidence-based medical information to support clinical decision making.
  • Combined with VR/AR technology, it helps doctors understand complex surgical procedures, improve preoperative planning and intraoperative decision-making ability. Combined with robot technology, as the brain to realize the application of embodied intelligence

Under the Memorandum of Understanding signed between HKSAR Government and Chinese Academy of Sciences (CAS), CAS established the Hong Kong Institute of Science & Innovation of Chinese Academy of Sciences (HKISI), its affiliated institution in Hong Kong, with a view to facilitating its presence in the research clusters on healthcare technologies and on artificial intelligence and robotics technologies. The institution aims to promote transfer and translation of scientific research results, coordinate collaboration between the CAS and local universities, and undertake technological education and promotion activities.

The Centre for Artificial Intelligence and Robotics (CAIR) of HKISI, was established in 2019 and is located at Science Park, Hong Kong. Its development is led by the Institute of Automation, Chinese Academy of Sciences (CASIA), also supported by the InnoHK programme from Innovation and Technology Commission. It is positioned to become a world-class research institute for AI and robotics, emphasizing applications in the range of healthcare, fintech and arts. CAIR aims to become an international hub for scientific research and technological innovation centre with global influence in the Guangdong-Hong Kong-Macao Greater Bay Area.

Enquiry