Definition: A term used to describe the increasingly common practice of combining advanced imaging (radiomics) and genomic data to produce impressive-looking studies that promise clinical transformation — but deliver little beyond glossy figures, moderate AUCs, and speculative conclusions.

Characteristics:

• Overuse of deep learning models without external validation
• Ambiguous clinical endpoints masked by heatmaps and pipelines
• Conclusions inflated to justify grants or headlines
• Real-world applicability: close to zero
• Frequently published in journals chasing translational buzzwords

Translation: > *Science that performs better on posters than in practice. It looks like progress, sounds like innovation — but doesn’t change how we treat patients.*