Generative AI Slashes Preclinical Animal Use by Up to 50% in Small‑Sample Research
A German‑French team introduced genESOM, a generative AI model that decouples structure learning from data synthesis, restores lost lipid signals in reduced‑sample multiple sclerosis studies, controls false‑positive inflation, and cuts required preclinical animal numbers by 30‑50% while outperforming GMM and CT‑GAN.
