Nikon's Small World in Motion competition, now in its 16th year, accepts microscope-based films and digital time-lapse entries and aims to showcase works at the intersection of scientific observation and visual expression.
The initial winner this year was Dr. Ning Xu of Tsinghua University in China, who submitted a short video displaying fine hair-like structures called cilia moving within the airways of a child with primary ciliary dyskinesia (PCD). Cilia help clear mucus and inhaled particles from healthy airways, so observing their behavior has both visual and medical interest.
The entry drew scrutiny after several researchers noted anomalies: cell and cilia structures and sizes appeared and disappeared in ways that observers considered atypical. A faculty member at the University of Florida reported that some structures ‘‘appear and vanish’’ and that cilia ‘‘seem to come out of nowhere,’’ which experts said can be a hallmark of artificial intelligence (AI) generation. Robert Hirst, a specialist in PCD diagnostics at the University of Leicester, also questioned the footage’s authenticity; speaking to the BBC, he said he has diagnosed PCD for twenty years by examining ciliary waveform, length, cell size and shape, and that the cells in the video do not resemble those typically seen in PCD patients.
Ian Donovan, a PhD student at UT Southwestern Medical Center, identified a distinctive watermark in a source image, which he said could indicate AI-generation. Two former judges also raised concerns about the video, according to reporting by TechSpot.
Dr. Ning Xu posted on LinkedIn (the post was later deleted) acknowledging the use of AI in post-processing the video. Xu said the AI was used only to ‘‘distinguish and display features of reconstructed grayscale images from super-resolution optical imaging’’ during post-processing, and asserted that the video itself, including the cilia and their motion, was not generated by AI. Xu maintained that the processing did not violate the competition rules.
That explanation did not resolve doubts. Observed irregularities continued to prompt questions, and experts suggested that showing the raw grayscale footage could have clarified what changes were made during processing.
Nikon’s investigation concluded that Dr. Ning Xu had breached the competition rules, and the prize was subsequently revoked. The award for the best entry was reassigned to Nguyen Nam Nhat of Vietnam, who had submitted a video of a tiny nematode and a unicellular organism.
In a statement, Nikon said it will review rules and judging procedures for future entries, and emphasized that the decision was not intended to call into question the excluded researcher’s professional work or scientific achievements.
The incident highlights wider challenges posed by AI in scientific imaging: while AI can be a powerful tool for recognizing patterns and analyzing large datasets, rigorous boundaries and transparency are particularly important in contests that aim to present microscope-observed reality. Nikon plans to clarify its rules to help prevent similar disputes in future competitions.
