Cohorts and protocols.

A cohort is Radial's working unit: a defined patient population (a treatment site, a technique, a study) that queries, scorecards, and monitoring all operate on. Everything downstream of curation happens to a cohort.

Cases sort themselves, you confirm

Radial groups incoming cases by treatment site and fractionation, then proposes a cohort assignment per group. Every case carries a "why this assignment" view of the evidence, and one decision covers the whole group. Assignments are made as cases arrive and flagged for you to confirm or move. Nothing is hidden, and mis-assigned legacy cases surface for moving instead of hiding in the denominator. When approval should gate what you count, any query can be scoped to curate-approved cases.

Protocols give a cohort its standards

A cohort becomes accountable when it has constraints attached. Radial ships protocol scorecards built from published constraint sets alongside your in-house ones, with clinical goals written in Mayo syntax (the dose-endpoint notation physicists already use, extended for Radial's distance types) and evaluated as cohort pass rates, not one plan at a time.

What a cohort gives you

For a cohort, Radial computes the dose metrics, population DVH, and plan-quality metrics, evaluates its limit sheets, and monitoring watches the pinned metrics as new cases accrue. Define the population once, and every view of it stays consistent. Same cases, same names, same numbers.