The problem. Head-to-head imaging benchmarks tell you how the platforms compare to each other, but not whether any of them is right. On tumor tissue especially — heterogeneous, clinically important — you want the imaging assays checked against independent references.

The idea. This comparison profiles matched FFPE tumor samples across the imaging platforms and, unlike a pure three-way bake-off, brings in orthogonal measurements as ground-truth anchors: bulk RNA-seq for expression concordance, GeoMx (region-based) profiling, and multiplex immunofluorescence for protein-level cell identity. That lets it ask not just “which platform detects more” but “which platform agrees with the references, and where they systematically disagree.”

Why it matters. The cancer-core-facility setting is the closest to where the spatial role actually lives, so a tumor-focused, reference-anchored comparison is the most directly relevant of the benchmarks. The methodological move — validate a new assay against established orthogonal ones before trusting it — is exactly the reproducibility discipline I’d bring to a facility, and it’s the same instinct as benchmarking a variant caller against a truth set.

Verdict. Pairs with the other 2025 imaging benchmark as the “how do these assays behave on real tumors, and do they agree with what we already trust” cluster. Same honest limits — sample, panel, and version dependence — so read the two together for the pattern of agreement and disagreement rather than a single verdict. The reference-anchoring is what makes this one worth reading closely.