Editorial note
Models were trained on 108 of the 540 augmented samples, and the source explicitly says the synthetic rows are not newly recruited patients; the dataset contains no imaging, pathology, laboratory values or staging. That makes 94.31% in simulation and 91.90% on annealing hardware a statement about one prepared dataset rather than about clinical performance. The result is worth recording precisely because the paper's own caveats are stronger than the announcement's headline.
Original sources
- Florida Atlantic University / D-Wave · limited-data lung cancer study (2 Nov 2026) ↗
- Discover Quantum Science · Inferring lung cancer under limited learning data ↗
Related companies
How this entry was published
The launch entries were compiled from dated first-party releases and official registers at the site owner's request and are explicitly marked as awaiting an independent verification pass. From here on an AI agent can only draft: an editor who has opened the original sources moves an entry through reviewed to published, and anything missing a source or a hand-written bilingual note produces no page at all.