Frequently Asked Questions

Rad AI Impressions is an AI‑enabled software solution that automatically generates a draft impression of a radiology report based on the radiologist's dictated findings, helping save time, improve accuracy, and reduce burnout. It runs alongside your existing reporting software and inserts a draft impression with consensus-based guideline recommendation directly into the report for the radiologist to review and edit.

No. Rad AI Impressions is a physician‑augmentation tool, not a replacement for medical judgment. A licensed radiologist must always review the AI‑generated impression, edit it as needed and sign the final report.

Rad AI Impressions integrates with leading radiology reporting systems (such as PowerScribe and Fluency) via a lightweight desktop client. As you dictate findings, Rad AI Impressions securely captures the report text, generates a draft impression in seconds and inserts it directly into the Impression section of your existing report editor — no change to your core reporting platform or voice recognition vendor.

Impressions supports common radiology modalities, including:

• CR X‑ray (computed radiography)
• CT (computed tomography)
• PET/CT (positron emission tomography/computed tomography)
• MR (magnetic resonance)
• US (ultrasound)
• MG (mammography)

The AI model automatically adapts the impression based on the detected modality.

Rad AI Impressions is trained on a large, diverse corpus of de‑identified historical radiology reports (no images) to build a generalized model for drafting impressions. For each site, de‑identified historical reports (typically 3–5 years) are used to customize the model to individual radiologists so that output more closely matches their language, style, and reporting patterns.

Rad AI Continuity is an AI-powered follow-up management platform that automatically detects imaging follow-up recommendations in radiology reports, helps navigators coordinate communication with providers and patients, and tracks every recommendation through to resolution so patients receive timely, appropriate follow-up care. It gives health systems and radiology groups a unified, closed-loop workflow for imaging follow-up.

No. Rad AI Continuity doesn’t replace clinical judgment or navigator roles. Radiologists still determine whether follow-up is needed and document recommendations in the report, and navigators and providers still decide how best to act on those recommendations. Rad AI Continuity’s role is to surface the right patients, automate routine steps, and keep status up to date so teams can focus on higher-value clinical work rather than manual tracking.

Rad AI Continuity integrates with your existing EHR and radiology information system using standard interfaces (such as HL7 results, orders, and ADT messages), so there’s no need to change your core systems. Finalized radiology reports are sent securely to Rad AI Continuity for AI analysis; follow-up entries then flow back into worklists, provider and patient communications, and status updates that stay in sync with your EHR scheduling and results workflows.

Rad AI Continuity can track follow-up imaging recommendations across a wide range of categories and programs, such as pulmonary nodules (including Fleischner-based follow-up) and incidental thyroid nodules (TI-RADS), among others. Findings are organized into standardized categories and subcategories, so your team can manage diverse follow-up workflows from a single, prioritized worklist.

Rad AI Continuity uses AI and natural language processing to identify follow-up recommendations and map them into clinically meaningful categories and timeframes. During implementation, Rad AI configures Continuity to your environment — tailoring categories, subcategories, due-date logic and exam-code mappings — so that entries align with your existing reporting patterns, consensus guidelines and routing rules. Over time, configuration and rules can be refined to reflect your evolving clinical policies and program goals.

Rad AI uses open standards, including DICOM, HL7, FHIR and OIDM, and integrates seamlessly into existing PACS/RIS/EHR environments. Most groups go live quickly using their current workflows, templates and microphones.

Yes. Rad AI's cloud-native architecture is built for scale, with centralized governance, standardized templates, enterprise integrations and support for multi-site workflows.

RSNA Ventures has partnered with Rad AI to bring RSNA's trusted, peer-reviewed radiology knowledge into the Rad AI workflow. This integration will enhance Rad AI Reporting by automatically surfacing case-based insights at the moment of interpretation, helping radiologists create more consistent, evidence-based reports without additional steps.

Rad AI is HIPAA compliant with enterprise-grade security controls, encryption practices and compliance protocols to protect patient data across all environments.

No. Rad AI augments radiologists — it doesn’t replace them. By reducing repetitive dictation and redundant steps, and providing intelligent assistance throughout the reporting workflow, it helps radiologists work more efficiently and reduce errors.

Clinical judgment and final sign-off always remain with the radiologist. Rad AI is built to support better, faster care — not to make decisions independently.

Radiology reporting software helps radiologists create and manage medical imaging reports. Rad AI Reporting enhances this process with a personalized experience that harnesses the power of generative AI (GenAI), structured templates, advanced speech intelligence and workflow automation to improve accuracy, efficiency and consistency.

Yes. Rad AI Reporting supports both structured templates and free‑form dictation. We can migrate your existing templates, picklists and macros or help you create new ones — and you can still dictate freely or use legacy workflows. Radiologists can work the way they prefer, with no forced change to their style.

Implementation is lightweight and fast. Because Rad AI supports existing templates, workflows and hardware, most radiologists are productive on day one with minimal disruption.

Rad AI's speech intelligence was built on the latest healthcare speech models, then deeply tuned for radiology with thousands of hours of radiologist dictation and specialty-specific terminology. The result is significantly lower error rates, a faster, real-time dictation flow and fewer interruptions than legacy solutions, allowing users to remain focused on their interpretation.

Rad AI learns each radiologist's language, style and reporting patterns from historical reports and live dictation to generate text that sounds like them. This includes impressions and unchanged follow-up exams where only true changes need to be dictated. It also adapts to how radiologists work — preserving their preferred templates, macros, guidelines, layouts and dictation habits — so the entire reporting experience feels personal while maintaining structure and consistency.

Rad AI Reporting performs real-time quality checks such as sex and age, clinical guideline recommendation insertion and structured organization to reduce clinically significant errors and increase report consistency. It understands clinical context, not just transcription.

Yes. Rad AI maintains and updates guideline logic within the system and allows practices to customize guideline language to their preferences. Guideline recommendations are inserted automatically where appropriate.