Why a platform, not scripts.

Physicists have been scripting data out of the TPS and OIS for years. We've written our share, and good scripts get real answers. The hard part starts after the script runs. Radial doesn't replace your science; it replaces the wiring.

Can you trust the dataset?

Cord vs SpinalCord vs SC_PRV03: before any analysis, someone has to decide what every structure actually is, and which of six plans was the one treated. A script absorbs those decisions silently, inside its author's head. Radial standardizes deterministically, with review queues: errors are surfaced, not absorbed, and every name and assignment carries provenance.

The second question shouldn't cost what the first did

Every new question sends a script back through the whole archive, file by file. Radial ingests once, into a database built for cohort-wide queries. After that, answers take seconds, and follow-up questions are free to ask.

Vendor coupling

In-house scripts usually target one vendor's export or API, and need rework when it changes. Radial is DICOM-native: it reads what your systems already produce, which means collaborators, legacy archives, and partner sites are the same import path as yesterday's cases.

Reproducibility

Scripts tend to leave with their author. Radial is the same validated engine for everyone (benchmarked against dicompyler-core, the public gold standard), so two people asking the same question get the same number, and the methods section becomes one sentence.

Continuity

A script produces one CSV, once. Radial keeps working after the answer: new cases are mapped, assigned, scored, and watched as they accrue, so the dataset you defended last quarter is still current this quarter.