- Published on
Alibaba Damo Academy Launches Liver Cancer AI Model Capable of Detecting Tumors as Small as 1 cm
- Authors
- Name
- DeepHorus
Alibaba DAMO Academy recently announced that its self-developed liver cancer AI model, 'DAMO LiON,' has made a breakthrough in real-world clinical trials. In a two-month prospective clinical trial, the model successfully identified 15 malignant tumors that were originally missed by radiologists, most of which were tiny lesions approximately 1 centimeter in diameter. This achievement has won valuable early treatment time for patients, and the related research paper has been successfully published in the top international medical journal 'Nature Medicine.' DAMO Academy stated that the launch of this model marks a new milestone in the field of medical AI.
It is understood that since its establishment in 2017, DAMO Academy has prioritized medical AI as one of its key strategic directions and was one of the first institutions in the industry to propose using artificial intelligence technology to identify subtle lesions in medical images that are difficult for the human eye to detect. Over the past few years, DAMO Academy has successively launched several cancer screening AI models, including 'DAMO PANDA' for pancreatic cancer screening, 'DAMO GRAPE' for gastric cancer screening, and 'DAMO COCA' for colorectal cancer screening. These models have all been published three times in 'Nature Medicine' and have received multiple authoritative recognitions, including the innovative channel certification from China's National Medical Products Administration and the 'Breakthrough Device' designation from the U.S. FDA.
The latest 'DAMO LiON' model serves as an 'AI safety officer' to assist doctors in reading CT images. Its core advantage lies not only in accurately identifying primary liver cancer but also in discovering liver metastases that are easily overlooked. Clinical trial data shows that the AI model's accuracy in identifying malignant tumors is higher than that of radiologists. When doctors read images with AI assistance, the average working time is reduced by 27%, while the sensitivity to malignant tumors is increased by 11.5%, effectively reducing the risk of missed diagnoses.
This technological advancement has had a significant impact on medical practice. The study points out that with the assistance of the AI system, junior radiologists can achieve diagnostic levels comparable to those of senior doctors, which helps alleviate the uneven distribution of medical resources and improve the diagnostic and treatment capabilities of grassroots medical institutions. At the same time, the increased efficiency in reading images means that doctors can dedicate more time to communicating with patients and analyzing complex cases, optimizing the overall medical service process.
Looking ahead, DAMO Academy plans to further integrate the 'DAMO LiON' model into clinical workflows and explore its applications in more aspects such as liver cancer screening, diagnosis, and treatment decision support. The team is also committed to promoting similar AI technologies to a broader range of diseases and medical scenarios, using technological means to continuously advance the development of precision medicine. Industry analysts believe that the continuous emergence and validation of such medical AI models will profoundly change future disease diagnosis and treatment models, benefiting more patients through early detection and precise intervention.
