AI model developed by alibaba's damo academy enhances liver cancer detection in clinical trial

A liver cancer diagnostic AI model, DAMO LiON, developed by Alibaba's DAMO Academy in collaboration with Shengjing Hospital and China Medical University, has been published in the prestigious medical journal Nature Medicine. The model is designed to identify small liver lesions, particularly those that are often missed, such as liver metastases, through enhanced CT images. In a two-month real-world clinical trial, DAMO LiON assisted in analyzing images for over 10,000 patients, identifying 15 previously undetected malignant tumors. Most of these lesions were approximately 1 cm in diameter, allowing patients to receive timely treatment through surgery or medication. Research findings indicate that the use of AI-assisted image analysis reduced doctors' reading time by 27% and increased their sensitivity to malignant tumors by 11.5%. Junior doctors were also able to achieve the diagnostic accuracy of senior doctors with the aid of the AI model. DAMO LiON employs an improved network architecture that focuses on the relationship between lesions and the entire liver, while preserving local texture and boundary details. The model can integrate different phases of enhanced CT images, capturing changes in small lesions across various scanning stages. This capability enhances the detection of complex cases, including fatty liver, liver cirrhosis, and postoperative livers.