Beijing,Feb.22(SANA) A Chinese research team has developed an advanced system for diagnosing rare diseases using artificial intelligence (AI) technologies, named “DeepRare,” setting a new global record for diagnostic accuracy.
Xinhua News Agency reported that the new system was developed by a joint team from Shenhua Hospital, affiliated with Shanghai Jiao Tong University School of Medicine, and the university’s School of Artificial Intelligence.
Since its online platform was launched last July, the system has attracted more than 1,000 specialized users from over 600 medical and research institutions worldwide.
According to the research team, tests showed that the system achieved 57.18 percent accuracy in diagnosis based solely on clinical information, without relying on genetic data.
This represents an increase of nearly 24 percentage points compared to the best previous global models. When genetic data was integrated, the diagnostic accuracy exceeded 70 percent.
DeepRayer relies on immediate access to a vast knowledge base encompassing medical literature and real-world clinical case data. It also employs an iterative analytical cycle involving hypothesis formulation, validation, and self-review to enhance the reliability of diagnostic evidence and address any potential logical gaps.
Researchers point out that diagnosing rare diseases remains a significant challenge, particularly in areas lacking genetic testing services. Traditional AI systems suffer from low reliability due to the difficulty in tracing their reasoning mechanisms.
The new system offers a practical solution to these challenges by enhancing transparency and accuracy in the diagnostic process.
