Peking University, in collaboration with the Suzhou Institute of Biomedical Engineering and Technology of the Chinese Academy of Sciences, has successfully developed the first AI-assisted pathological diagnostic microscope. Equipped with independently developed deep learning algorithms, the device can identify cancerous cells in cervical smears in real time during microscopic observation, achieving an accuracy rate of 99.3%. It has now passed the National Medical Products Administration's special approval process for innovative medical devices.

Utilizing a convolutional neural network architecture, it can process 200 microscopic images per second, reducing diagnosis time for individual cells to 0.2 seconds.

Integrates phase contrast, fluorescence, and brightfield imaging modes to support observation needs for different sample types.

Built-in 5G communication module enables remote expert consultations and real-time annotation guidance.

In multicenter trials involving 36,000 cervical cancer screening cases:

: Unit price controlled under ¥200,000, significantly enhancing tumor screening capabilities in grassroots hospitals.

: Built-in teaching system provides real-time annotation of typical pathological features.

: Already deployed in HPV rapid screening programs.

WHO cancer expert André Ilbawi commented: "This technology will reshape the landscape of cancer screening in low- and middle-income countries."

The device is expected to obtain medical device registration certification in the first quarter of 2025, with an annual production capacity of 5,000 units.