Atalaya - Radiology Information System from Order to Signed Report

Dashboard, protocolling queue, reading worklist, reporting workspace and critical findings. Every one of the 22 demo studies opens its own reporting workspace.

Laterality in the impression matches the exam, every AI finding has a response, any critical finding is logged and the template's required fields are complete. Sign and Sign and next stay disabled until all four pass.

Orange findings get 12 hours and yellow 7 days, counted down from signature. A phone call only counts as acknowledged when read-back is confirmed.

Across 9 centres and 15 exam types, with 12 orders to protocol, 11 critical findings, 7 payers including 3 regional conciertos, and flags from 5 labelled AI models.

Nine Centres, Nine Systems

Each acquired centre kept its own RIS, so reading for another site meant a second login, and orders came in by fax, messaging app and partner hospital feeds. There was no single list of what was waiting, where, or for whom.

Protocols Chosen by Whoever Was Free

Real Decreto 601/2019 requires every medical exposure to be justified by a practitioner and optimised against diagnostic reference levels. Protocols picked at the scanner missed contrast allergies, low eGFR before gadolinium, possible pregnancy and repeat CTs of the same region.

Urgency Lost in the Queue

STAT studies waited behind routine ones, and conciertos with regional health services set contractual report turnarounds nobody tracked. AI triage tools could help, but they are medical devices under the EU MDR and likely high-risk under the AI Act, so their output must stay decision support.

Critical Findings Without Proof

Urgent results were phoned through with no record of who was reached or whether they repeated it back. Health data is a special category under the GDPR and the LOPDGDD, so every signature, call and view of a report also has to be traceable.