FIDI deploys Oxipit AI through CARPL for 24/7 chest X-ray decision support in emergency departments

In Brazil's busy emergency rooms, physicians must frequently interpret chest X-rays without on-site radiologist support. Clinicians — often early-career and hired through third-party agencies — faced a three-way dilemma: wait hours for a radiology report, order additional imaging (CT) at added cost and delay, or proceed with diagnostic uncertainty and risk misdiagnosis.

FIDI deployed Oxipit's AI-powered chest X-ray tool through CARPL's vendor-neutral platform, integrated with Philips Vue PACS. Rather than generating standalone reports, the AI overlays visual annotations of potential findings directly onto the X-ray image, providing real-time diagnostic support at the point of care 24/7 — without requiring radiologist presence.

Following a phased rollout from April to July 2024, Oxipit AI became a standard decision support tool across FIDI's partner ER departments lacking on-site radiologists. ER physicians now receive AI-annotated images in real time, reducing reliance on additional imaging and improving diagnostic consistency regardless of which third-party physician group is staffing the department. A related deployment at the HSPE emergency department in São Paulo (cited in an external article) reduced chest X-ray turnaround time from one hour to five minutes.

• Visual annotation directly on the image — rather than a separate report — is better suited to ER clinicians who need instant, interpretable feedback at the point of care.
• Deploying through a validated AI marketplace (CARPL) allowed FIDI to skip lengthy procurement and development cycles; over 4,500 studies were pre-validated before go-live.
• Standardizing AI support across a rotating, agency-staffed ER workforce ensures a consistent baseline of diagnostic quality independent of individual physician experience.