Tencent AI Lab Produces Smart Microscope for Medical Diagnosis

Tencent AI Lab unveiled its new intelligent microscope for pathological analysis yesterday at the Tencent Global Partner Conference in Nanjing. The device represents a huge step in the company’s ongoing commitment to ramp up its R&D in AI + medicine.

The Tencent smart microscope exceeds the limit of what regular microscopes can do by actively assisting physicians’ diagnosis through the incorporation of AI techniques such as computer vision and natural language processing. it also employs augmented reality during the slide reading process. Moreover the microscope’s voice interface allows doctors to easily input hands-free commands for detection, identification, quantification, and the generation of reports.

The computer vision technology can assist not only in simple cell metrology, but also with much more difficult tasks such as cancer type identification. Detection results meanwhile can be displayed in real time via the device’s eyepiece, which, along with the speech-recognition-driven voice interface feature further streamlines and speeds up the doctor’s reading process to contribute to greater diagnostic efficiency and accuracy.

The general diagnostic process includes four major steps: medical imaging screening, pathological analysis, treatment planning, and post-operative recovery and rehabilitation.

In the case of colorectal screening for example, polyps are examined to determine whether they are benign or malignant and to determine specific lesion categories (adenomas, adenocarcinomas, etc.). Manual diagnostic screening results can vary between experienced and inexperienced doctors, and the microscope’s AI — trained on a large amount of expert data — can improve consistency in such diagnoses. The doctor need only give the “Polyp Classification” voice command and the AI will automatically read the slide and collect relevant images. The “Generate Report” command will then output a comprehensive report for the doctor to review.

Tencent’s imaging screening products have already been deployed in more than 100 top Chinese hospitals, where they have read more than 100 million images and served more than a million patients. By entering the pathological analysis stage, the company is laying a strong foundation that will support the development of its ultimate goal: an AI-powered solution that can cover the entire medical diagnosis and treatment process.

Author: Jessie Geng | Editor: Michael Sarazen