China has published new ethical guidelines for research and the clinical use of artificial intelligence in medical imaging. The rules cover the development of algorithms, their testing and validation, and their implementation in medical practice.
Medical imaging uses artificial intelligence to analyze scans from computed tomography, magnetic resonance imaging and ultrasound examinations. The technology can assist doctors in detecting pathological changes and speed up the diagnostic process.
However, wider use of these systems also carries risks. They include leaks of sensitive health data, algorithmic discrimination and uncertainty over who is responsible when an artificial-intelligence system makes a mistake.
The new guidelines are based on six principles: protecting and improving human well-being, fairness, respect for human autonomy, protection of personal data and information security, safety and controllability, and transparency.
Patients’ interests and health must come first. Researchers are required to obtain informed consent from participants, who may withdraw that consent at any time.
The rules specifically state that artificial intelligence cannot make a final diagnosis. Responsibility remains with people, and the doctor retains the final say when making medical decisions.
The guidelines also prohibit exaggerating the capabilities and results of artificial-intelligence models. They stress the need for algorithmic fairness so that systems do not produce diagnostic bias because of insufficiently representative or unbalanced training datasets.
Through the new requirements, China aims to combine the development of artificial intelligence in medicine with stronger patient protection, more reliable technologies and human control over medical decisions.
|
9490 | 30 Aug. 2026 | 17:00
Mobile verison
RSS
