Siriraj’s AI reads chest X-rays with 95% accuracy, used in 500,000 cases

SATURDAY, FEBRUARY 21, 2026

Siriraj Hospital develops AI for chest X-ray reading with 95%+ accuracy, used in 500,000+ cases and rolled out to 145+ public hospitals

  • The Faculty of Medicine Siriraj Hospital has developed AI to help interpret lung/chest X-rays, with accuracy above 95%, comparable to expert radiologists.
  • The technology has been used in real practice in more than 500,000 patient cases and can report results in under 10 seconds.
  • Siriraj has transferred the AI technology to more than 145 hospitals under the Ministry of Public Health nationwide, and plans to develop AI for brain CT scans and breast cancer screening (digital mammography).

The Faculty of Medicine Siriraj Hospital, Mahidol University, through its Department of Radiology, says it has achieved major success in developing an AI programme to increase the efficiency of interpreting chest X-ray images. Development began during the Covid-19 pandemic to cope with large patient volumes and has continued to the present.

The innovation has been certified by the Royal College of Radiologists of Thailand as having accuracy above 95%, comparable to interpretation by specialist radiologists.

A key advantage is speed, with results produced in under 10 seconds per case. The technology has now been used in real-world practice in more than 500,000 patient cases.

For nationwide expansion, Siriraj has transferred the AI technology to more than 145 hospitals under the Ministry of Public Health and aims to expand its use to cover 450 hospitals in 2026, to reduce inequality in access to quality healthcare.

Siriraj is also extending the work by developing AI to interpret brain CT scans and to support breast cancer screening using digital mammography. It is collaborating with specialised private-sector partners to build AI with a better understanding of breast anatomy specifically among Asian women, to strengthen screening for major diseases and improve health outcomes for Thai people on a wider scale.