Perspectives

Artificial intelligence is becoming commonplace. In the medical sector, doctors increasingly appreciate the value that AI can bring to help augment their diagnosis and treatment of diseases.

But there remain challenges. It is difficult transforming unstructured data from real scenarios in to knowledge a computer can work with.

For AI to “think” it needs knowledge graphs; a way of organising information enabling computers to form “knowledge” about a subject.

Behind the scenes computer scientists and engineers are expanding the boundaries of knowledge graphs to help AI make better informed decisions. Ping An Smart City recently unveiled a Chinese medical knowledge graph, which was developed by Ping An Smart Healthcare together with the Institute of Medical Information/Medical Library of the Chinese Academy of Medical Sciences and Peking Union Medical College.

This work involved researchers deploying natural language processing technology, which allows computers to understand and process human language. This technology was used to identify the concepts and relationships, covering diseases, drugs, symptoms and examinations, from medical text books and literature in Chinese language. This was a wealth of knowledge previously inaccessible to machines. They also used the knowledge graph integration technology to merge this medical knowledge from different sources. This resulted in the knowledge graph transforming unstructured medical knowledge into a format which can be comprehended by machines.

Already, this knowledge graph has been used to develop AI-assisted diagnostic and treatment models for several thousand disease types. Based on these models, the AI assistance makes recommendations to human doctors covering diagnosis, further medical examination to be undertaken and potential treatments. All this goes toward improving overall treatment efficacy or effectiveness.

Our work to date has achieved accuracy levels of 95 per cent for AI–assisted diagnostic models for common diseases in General Practice Medicine. Xie Guotong, Chief Medical Scientist of Ping An Group, said this knowledge graph has been gradually applied to build solutions for the entire process of pre- and post-diagnosis.

The Chinese medical knowledge graph is also built with deep learning abilities, allowing it to accumulate experience by continuously analysing effective treatment data and cases. This is a growth model similar to how doctors learn from experience in real life.

Ping An Smart Healthcare’s AI-assisted diagnosis and treatment system has been deployed in nearly 1,000 medical institutions in China, covering more than 60 cities in 17 provinces including Heilongjiang, Gansu, Hunan and Hebei, with more to come in the future.

This website uses cookies to help us provide you with better experience and allow us to improve our service. By continuing to browse the site, you understand and agree to our Privacy Policy and Terms of Use .

This website is not supported by IE. Get the latest version of Firefox or Chrome for better browsing experience.

Noted