The New England Journal of Medicine has retracted a clinical image after discovering that artificial intelligence was used to alter it. The image, initially published on April 18, accompanied a report on an 87-year-old man’s lung injuries following forest fire exposure.
The image featured striking black structures retrieved from the patient’s airways, with a measuring tape at the top for scale. However, authors Yuling Wang and Xiangdong Mu, affiliated with Daxing Teaching Hospital and Beijing Tsinghua Changgung Hospital, confessed to using AI tools to reposition the tape.
In the retraction notice dated April 29, the authors stated, “We were unaware of the Journal’s policies on image manipulation and used AI to adjust the ruler’s position. We, therefore, withdraw our image and case report.”
A PubPeer commenter initially raised concerns about the image’s scale, noting irregular numbering on the tape. Mu responded, clarifying that the patient was critically ill at the time of imaging, and adjustments were made for better clarity. Mu emphasized that these changes did not impact the clinical data or conclusions.
The journal has not yet commented further, and attempts to reach Harleen Marwah, the editor for the image, have been unsuccessful.
This incident marks the first retraction for the journal since 2020, following a similar case involving data from the private company Surgisphere.
