AI offers breakthrough in cancer diagnosis

Huawei Cloud, a subsidiary of the Chinese technology giant Huawei Technologies, and Kingmed Diagnostics, a Guangzhou-based medical diagnostic testing company, have announced a technological breakthrough in the diagnosis of cervical cancer.

By leveraging AI technologies, they can diagnose the disease with a true positive rate of over 99.9 percent, and a true negative rate of more than 99 percent, which is currently the highest level in the world, according to Kingmed Diagnostics.

"It takes only 36 seconds to examine cervical smear cells, which would usually need an average six minutes by cytopathologists," said Luo Pifu, director of the Kingmed Diagnostics pathological department.

Cervical cancer remains one of the biggest threats to women, with about 100,000 new cases being reported each year in China, according to Luo.

"Early examination is key to treating cervical cancer," said Luo.

However, only 70 million women have been examined since China began a nationwide examination program against cervical and breast cancers in 2009, according to Liang Xiaoman, a Guangzhou-based cytologist.

Liang said China lags in the development of pathologists, especially in rural areas.

"The lack of enough pathologists has become a major factor affecting the nationwide examination of cervical cancer," said Liang.

The companies began collaborating in July, conducting research in the use of artificial technology in assisting examinations, and was based on 43.5 million cervical cell samples collected by Kingmed Diagnostics.

The breakthrough was also based on Huawei Cloud's Model-Arts, which provides data processing, intelligent labeling, a readily packaged development environment, automatic model generation and comprehensive AI development.

"The AI technology will help greatly increase efficiency of the diagnosis of cervical cancer, benefiting the country's 350 million women," said Liang.

According to Pan Jie, director of Huawei Cloud Guangdong business section, the AI technology may be expanded into areas of pathology such as breasts, digestion, kidneys and blood.

"We are exploring more uses of the cutting-edge AI technology in medical areas, aiming to provide more assistance to increase medical testing efficiency," said Pan.