China creates AI model for esophageal cancer screening

Chinese scientists have developed an artificial intelligence model called Damo Eagle for esophageal cancer screening.

This was reported by the technology corporation Alibaba.

The model was jointly developed by Alibaba's research division — the DAMO Academy, the Sichuan Provincial Cancer Hospital, and the Sun Yat-sen University Cancer Center.

The model does not require intubation or the administration of a contrast agent. Damo Eagle only needs to analyze a standard chest computed tomography scan to identify the risk of esophageal cancer, including early stages and precancerous lesions.

The model's sensitivity in cases of esophageal cancer (the probability that a sick person will be classified precisely as sick) reaches 90%, while its sensitivity for precancerous lesions with very weak signals is 52.5%. The model's specificity (the probability that subjects who are not sick will be identified precisely as not sick) reaches 99.2%.

The model was tested on more than 80,000 cases in three countries. A scientific paper devoted to this model was published in the international medical journal Nature Medicine. According to statistics from China's National Cancer Center, about 220,000 new cases of esophageal cancer are diagnosed annually in the PRC, of which 190,000 end in death.