AI cancer detection boosts healthcare

Driven by a vision to overcome geographic barriers, Liu Wen, a professor at the Xinjiang Institute of Engineering, has pioneered an AI-powered diagnostic system aimed at improving early cancer detection and expanding access to medical services for those living in remote communities.

Liu's journey with artificial intelligence in the medical field began in a university lab, where he became interested in the potential of intelligent systems to address real-world challenges. He later focused on a persistent issue in grassroots healthcare — early cancer detection.

"Early discovery and treatment can increase the survival rate of cancer patients dramatically, with cure rates for early-stage cancers as high as 90 percent," Liu said.

In the Xinjiang Uygur autonomous region, where communities are often separated by vast distances, access to specialized diagnostic expertise has been limited. Many patients have had to travel long distances to major cities for authoritative checkups, sometimes missing the critical early window for intervention.

In 2019, Liu assembled a cross-disciplinary team, partnering with leading regional hospitals and bringing together engineers and medical professionals. Their goal was to develop an AI-assisted diagnostic system capable of providing expert-level screening at the township level.

In 2020, the team developed an AI-assisted diagnostic system for breast and thyroid cancers. According to Liu, the system can identify suspected cancerous cells with millimeter-level precision in 12 areas, including the breast, thyroid and abdomen, achieving an accuracy rate of more than 96 percent in distinguishing benign from malignant tumors.

Since May 2023, Liu has promoted the system through malignant tumor screening initiatives and academic and technology exchange activities at more than 100 primary medical institutions across Xinjiang. The system has assisted primary healthcare doctors in completing more than 1.3 million screenings.

In the past five years, Liu's team has traveled extensively across the region to collect medical imaging data to train their algorithms and to conduct cancer screenings and technology exchange activities. To date, they have held 27 training sessions and trained more than 9,000 medical personnel.

As a member of the Xinjiang regional committee of the Chinese People's Political Consultative Conference, Liu said he sees his role as helping translate technological innovation into social benefit.

"The AI-assisted diagnosis technology has brought transformative change to grassroots healthcare in Xinjiang. It makes quality medical services accessible to people in grassroots communities," he said. "Residents in remote areas now get expert-level medical advice at their doorstep without long travel, realizing the vision of AI technology benefiting thousands of households in Xinjiang."

Liu said he hopes to see a day when grassroots doctors feel more confident handling complex tumor diagnoses and herders on the grasslands no longer have to travel long distances for a single checkup.

"When timely and accurate diagnoses become a solid foundation for health, more families can move beyond the shadow of disease and hold onto reunion and hope. I believe this is the best answer to how AI can serve people's livelihoods," he said.