Clinical evidence, integration guides, and regulatory updates for radiology AI in community hospitals.
High-sensitivity models generate alerts. Too many alerts train radiologists to ignore them. The right metric for radiology AI deployment isn't sensitivity alone — it's alert precision at the workflow level.
Most radiology AI vendors sell you a detector for one thing — LVO, ICH, PE. Pacslens takes a different bet. Here's why worklist-level triage is a fundamentally different product strategy, and why it matters for community hospitals.
Published LVO detection studies report sensitivity ranging from 87% to 97% depending on cohort and threshold. Here's how to read those numbers without being misled by cherry-picked benchmarks.
Multiple published studies show AI-assisted ICH flagging reduces time to first radiologist read. We examine the evidence base, the methodological limitations, and what 'time-to-read' actually means in a community hospital.
CT pulmonary angiogram AI detection achieves AUC 0.90–0.94 for central PE in retrospective series. But incidental PE, subsegmental disease, and artifact-heavy scans remain challenging. Here's an honest assessment.
There are four distinct DICOM integration patterns for radiology AI deployment. Each has different IT complexity, latency implications, and PACS compatibility requirements.
The FHIR R4 ImagingStudy resource is increasingly how radiology AI results reach EHRs and RIS systems. We walk through the resource structure and what Epic Radiant's FHIR implementation exposes for AI vendors.
The American College of Radiology's AI-LAB framework provides evaluation criteria for AI tools in radiology. Most community hospitals have no dedicated AI governance committee to apply it.
The FDA has cleared over 700 AI/ML-based medical devices as of 2024, with radiology representing the largest category. But what '510(k) cleared' actually means for procurement is more nuanced than most vendors explain.
Academic medical centers have AI governance committees, dedicated IT engineers, and subspecialty radiologists. Community hospitals have one radiologist on call overnight and an IT team that also covers the EMR, billing, and the parking gate.