The Center for Artificial Intelligence and Robotics (CAIR), a Hong Kong-based research center affiliated with the Chinese Academy of Sciences (CAS), has introduced a novel AI tool designed to aid complex brain surgeries. The tool dubbed CARES Copilot 1.0 AI model aims to enhance clinical diagnosis and medical judgments for neurosurgeons.

In an interview with the South China Morning Post, Liu Hongbin, the executive director of CAIR, stated that the CARES Copilot 1.0 system had undergone internal testing in various hospitals. It has been integrated into doctors' workflows to assist in preparing surgical plans and managing post-surgery procedures efficiently. However, he acknowledged current limitations in the healthcare industry regarding the scarcity of specialized databases for brain surgery procedures, often with inadequate inter-hospital connectivity. To overcome these limitations, Liu highlighted that the research team fed CARES Copilot 1.0 with extensive data from thousands of medical textbooks, academic papers, and international neurosurgical guidelines.

During a live demonstration at the Hong Kong Science Park, Danny Chan Tat-ming, head of the neurosurgery division at the Department of Surgery at the Chinese University of Hong Kong, showcased the capabilities of the CARES Copilot 1.0 model. The AI tool can rapidly generate crucial information from academic papers, including citations, ensuring accurate answers. Chan mentioned that the tool boasts an impressive accuracy rate of up to 95%.

CARES Copilot 1.0, based on Meta Platforms' Llama 2 large language model (LLM), underwent training on diverse multimodality databases, incorporating text, images, MRI, CT scans, and ultrasound imaging tailored for the medical field.

The initiative by CAIR, co-funded by the Hong Kong government's InnoHK research program, reflects the broader efforts in China to advance AI innovation, aligning with the country's goal to catch up with leading entities like OpenAI, the creator of ChatGPT.