Students develop AI system for minimally invasive breast tumors surgery

Members of a student team from Nankai University, Tsinghua University and the Beijing Institute of Technology pose for a photo during a student innovation showcase at Zhongguancun Academy in Beijing, June 12, 2026. The team is developing an AI-powered minimally invasive breast surgery system aimed at improving early diagnosis and treatment of breast tumors. [Photo by Liu Caiyi/China.org.cn]

A student team at Zhongguancun Academy is developing an AI-powered surgical system to improve early diagnosis and treatment of breast tumors.

The project, which combines advanced imaging, surgical robotics and embodied AI technologies, was presented Friday at a student-led innovation showcase jointly organized by Zhongguancun Academy and the Zhongguancun Institute of Artificial Intelligence in Beijing.

The six-member team, drawn from Nankai University, Tsinghua University and the Beijing Institute of Technology, launched the project to address key challenges in breast tumors diagnosis and treatment.

Zhang Boya, a student at Zhongguancun Academy and one of the project's initiators, said the idea grew out of her clinical training, where she recognized the growing importance of minimally invasive surgery in reducing treatment risks and improving patient outcomes.

The team identified two major challenges facing minimally invasive breast surgery. Current imaging technologies often struggle to provide clear, real-time visualization of lesions and surrounding tissues, while existing surgical tools can make precise operations difficult.

To address these issues, the team is developing what members describe as a combination of a "smart eye" and a "smart scalpel."

The imaging component focuses on real-time 3D ultrasound reconstruction. It uses original AI algorithms to sharpen resolution and contrast, even when imaging data is incomplete.

The team hopes the technology will eventually complete 3D reconstruction within seconds while achieving segmentation accuracy of up to 95%, giving surgeons more detailed and timely information during procedures.

On the surgical side, the team is developing an intelligent robotic system capable of adapting to tissue characteristics during operations. The system incorporates embodied AI technologies, including physical property perception, world models and experience-based decision-making.

The robotic platform is designed to generate adaptive surgical strategies and make real-time adjustments during procedures, with the goal of improving precision and safety in minimally invasive interventions.

The project also includes a virtual simulation platform that visualizes the surgical process and helps verify puncture path planning and human-machine interaction before clinical use.

The team chose breast tumors as its first application because of its high incidence rate, substantial treatment costs and strong demand for minimally invasive procedures. Members said the underlying technologies could eventually be extended to surgeries involving other organs.

Looking ahead, the team hopes the system will help bring high-quality medical resources to more patients, particularly in areas with limited access to specialist care, while contributing to the development of next-generation domestic medical equipment.