Radium
by Radium
· Based in United States →
— The fastest inference engine for building production-ready AI systems. Deploy, manage, and scale your AI models with enterprise-grade infrastructure.
Per-token basis, varies by model tier (hal, clarke, tycho) and by input and output tokens.
Regulatory Status Disclosed
Overview
Radium by Radium is an AI-powered platform designed to assist healthcare providers with administrative and clinical documentation tasks. It is primarily intended for use by physicians and other clinicians across various specialties and care settings, including hospitals, clinics, and private practices.
- What it does: Radium automates the generation of clinical notes, summaries, and other documentation by leveraging AI to process patient information. It aims to reduce the time spent on administrative burdens, allowing clinicians to focus more on patient care.
- Who it is for: The tool is applicable to a broad range of medical specialties that require extensive documentation, such as primary care, internal medicine, and specialists who manage complex patient cases.
- How it fits a clinical or practice workflow: Radium integrates into existing clinical workflows by capturing patient encounters (e.g., through ambient listening or transcription of dictated notes) and then drafting relevant documentation. Clinicians can then review, edit, and finalize these AI-generated drafts, ensuring accuracy and compliance.
- Notable capabilities: Key features include the ability to generate various types of clinical notes (e.g., progress notes, discharge summaries), extract key information from patient records, and potentially integrate with electronic health record (EHR) systems to streamline data entry.
Reviewed by Pouyan Golshani, MD — Interventional Radiologist
Key Features
- AI-powered second read for chest X-rays
- Abnormality detection
- Workflow integration
- Efficiency enhancement
- Augments human expertise
Use Cases
- Assisting radiologists with chest X-ray interpretation
- Improving diagnostic accuracy in radiology
- Streamlining radiology reporting workflows
- Reducing diagnostic errors in medical imaging
- Prioritizing critical cases in high-volume settings
What Physicians Need to Know
Evidence Base
Radium has developed the world's first radiological AI-native foundation model, referred to as the 'GPT of radiology.' This model is designed to create an imaging biomarker factory for both clinical practice and research. The AI assistant analyzes patient data, referencing the latest medical guidelines and recommendations to identify inconsistencies or potential risks.
Clinical Validation Studies
Radium's R. Read solution is currently available as a standalone pipeline for clinical trial use, with use in clinical practice pending regulatory clearances. It has been deployed at Moffitt Cancer Center for clinical research settings, replacing a legacy radiomics tool. The platform supports complex oncology imaging workflows, including whole-body lesion detection, AI segmentation models, and longitudinal transfer of lesions across studies. Raidium aims to decrease inter-reader variability by three times through organ-agnostic automated RECIST measurements. The company is pursuing 510(k) clearance for a subset of features and expects to announce clearance before year-end 2026.
Alert Fatigue Management
While not explicitly detailed for Radium, AI in general helps combat alert fatigue by filtering irrelevant or low-threat alerts, allowing clinicians to focus on critical events. AI can also prioritize alerts by assigning severity levels based on threat intelligence and potential impact. The Radium AI platform for RPA bot management includes real-time event alerts and monitoring, with a focus on reducing manual effort and faster resolution through auto-healing insights.
Drug Interaction Checking
While Radium's primary focus appears to be radiology, AI in healthcare is being developed to identify harmful drug interactions by analyzing large datasets quickly. AI models can process vast amounts of data in real-time with high accuracy, improving predictions and lowering adverse events. Some AI models have achieved high accuracy (e.g., 99.9%) in predicting harmful drug interactions.
Differential Diagnosis Support
Radium's Medical AI Assistant module is designed to support clinical decision-making by conducting instant analysis of patient data, aiding physicians in diagnostics and treatment planning. It references the latest medical guidelines and recommendations. In a broader context, AI diagnosis tools can help expand, update, and organize differential diagnoses within the clinical workflow.
Clinical Workflow Integration
Radium was built with an 'AI-native viewer from the ground up,' designed to fit real-world radiologist workflows, aiming to reduce friction and improve usability. The platform consolidates critical functionalities into a single system to ensure seamless data flow and an intuitive working environment. Radium's core philosophy emphasizes 'Infrastructure First, AI Second,' with its PACS infrastructure providing deep integration into clinical workflows. The R. Read solution aims to streamline oncology imaging review, reducing manual effort across search, review, and follow-up, and facilitating longitudinal tracking. It also aims to streamline radiologist-clinician communication for better medical decision-making.
Decision Audit Trail
An AI audit trail is a complete, tamper-evident record of what an AI system did and why, including the data used, the decision or output produced, and the actions taken. For AI-assisted systems, audit trails need to capture the specific model version, a reference to the input data, and the identity of any human reviewer. This is crucial for accountability, risk management, and compliance, allowing for the reconstruction of any decision.
Physician Tip
Leverage Radium's AI-native viewer for oncology imaging to streamline complex reviews and reduce manual tasks, particularly for longitudinal tracking and RECIST measurements. The integrated AI assistant can provide instant analysis of patient data, aiding in diagnostics and treatment planning by referencing current guidelines. While the R. Read solution is currently for clinical trial use, its design for real-world workflows suggests future benefits in clinical practice for enhancing efficiency and consistency in radiology. Pay attention to the forthcoming regulatory clearances for broader clinical application.
Radium is built on a PACS infrastructure, allowing for deep integration into clinical workflows. The company's approach is modular, designed to integrate seamlessly into existing environments without extensive prerequisites. For its RPA bot management offerings, Radium AI integrates with leading RPA platforms like UiPath, Automation Anywhere, and Blue Prism, and can integrate with ITSM systems like ServiceNow for automated ticket generation. Radium also implements an OpenAI-compatible API contract, facilitating integration with existing AI stacks through an endpoint swap.
Details
| Category | Clinical Decision Support & Reference, Radiology & Imaging AI |
| Pricing |
Per-token basis, varies by model tier (hal, clarke, tycho) and by input and output tokens.
|
| Free Trial | No |
| Deployment | Cloud, on-prem (for RPA bot management platform). |
| Mobile App | No (Note: There are unrelated apps named 'Radium' for webradio and local shopping). |
| API Available | Unknown |
| Languages | English (for the API services). |
| Target Size | Companies running AI through API endpoints; companies building or deploying private models; companies new to AI. |
| Compliance | |
| BAA Available | Unknown AI-estimated |
| HIPAA Compliant | Unknown AI-estimated |
| FDA Status |
Not applicable AI-estimated Radium (radium.cloud) does not have FDA clearance. However, a French medtech startup named Raidium (raidu.ai), founded in 2022, is pursuing regulatory approvals for FDA and CE certifications for its radiological AI-native foundation model. Another company, RADIN Health, which integrates with AZmed's FDA-cleared AI solution for radiology, is a separate entity. |
| Integrations | |
| EHR | Not specified |
| Specialties | Radiology |
Social Proof
| Customers | unknown |
| Notable | Pre-Radium era team worked on Mint.comHarvard Business Publishing. Radium AI (RPA bot management) integrates with UiPathAutomation AnywhereBlue Prismand Microsoft Power Platform. |
Ratings & Reviews
No reviews yet. Be the first to review this tool!
Rate Radium
Press & Coverage
Radium AI
Radium Introduces AI-Powered Oncology Imaging Workflow to U.S. Cancer Centers
Radium has launched Raidium Read (R.Read) in the U.S., an AI-native imaging solution for oncology research centers, already deployed at Moffitt Cancer Center. The platform offers whole-body lesion detection, AI segmentation models, and longitudinal transfer of lesions across studies to support complex oncology imaging workflows.
2026-07
Tech.eu
ChatGPT of radiology: Raidium secures $13M for AI-powered precision diagnostics
French medtech startup Raidium raised $13 million in seed funding to accelerate the global rollout of its AI-powered precision radiology technology. The company plans to expand to the U.S., strengthen European operations, pursue regulatory approvals (FDA and CE), and enhance its R&D efforts.
2024-12
Imaging Technology News
Radium Debuts New AI-Native PACS Viewer | Imaging Technology News
At RSNA 2025, Raidium will introduce its new AI-native PACS Viewer, powered by Curia, the first Foundation Model dedicated to radiology. This multimodal model, trained on over a billion images, can interpret entire imaging exams and execute complete workflows.
2025-12
VentureBeat
Radium looks to speed up AI and ML jobs in cloud datacenters | VentureBeat
Radium, a startup using AI and machine learning to enhance cloud hardware computing power, emerged from stealth mode in 2021, deploying solutions to Cyxtera's cloud datacenters. Their main product, Launchpad, optimizes performance for AI algorithms by allowing projects to run on bare metal machines.
2021-12
Ontario International
Radium | Ontario at World AI Cannes Festival 2026
Radium will be participating in the World AI Cannes Festival 2026, showcasing its vertically integrated infrastructure stack for building, deploying, and running private AI models at scale. The company aims to provide faster performance and lower costs by eliminating traditional cloud inefficiencies.
2026-02
Forbes Georgia
Georgian Startup Radium Pioneers Healthcare Transformation with AI-Powered Platform Medspace.me - Forbes.ge
Georgian tech startup Radium is transforming the nation's healthcare sector with its AI-powered platform, medspace.me, designed to enhance efficiency, accuracy, and accessibility of clinical operations. The platform offers an integrated solution, including a Cloud PACS for radiological image archiving and management.
2025-05
Investors Hangout
Radium's AI Solution Targets Oncology Imaging Overhaul - Investors Hangout
Radium's AI-native viewer is designed to streamline oncology imaging workflows, reducing manual effort, inter-reader variability, and aiding longitudinal tracking and communication. The system is currently available for clinical trials and is pursuing FDA clearance for broader clinical practice.
2026-07
AuntMinnie.com
Radium raises nearly $17M in funding round - AuntMinnie
Precision radiology startup Raidium secured 16 million euros (approximately $16.9 million) in a seed funding round to advance its AI development and global expansion. The company's radiological AI-native foundation model provides 3D segmentation, advanced measurements, and automated analysis of complex biomarkers.
2024-11
Videos
Product demos, reviews, and walkthroughs for Radium.
The True Story of the Radium Girls
Dr Ben Miles
The Radium Worked Until His Jaw Fell Off
Kyle Hill
The Story of the Radium Girls | A Short Documentary | Fascinating Horror
Fascinating Horror
This is more radioactive than uranium
NileRed
The Horrifying Story of The Radium Girls - Science's Dark Secret
Python Programmer
Who Were The Radium Girls?
BuzzFeed Unsolved Network
Frequently Asked Questions
Radium-223 dichloride (Xofigo) is indicated for the treatment of patients with castration-resistant prostate cancer (CRPC), symptomatic bone metastases, and no known visceral metastatic disease. It is an alpha-emitting radiopharmaceutical that selectively targets bone metastases.
Radium-223 dichloride is contraindicated in women who are pregnant or may become pregnant. Significant warnings include myelosuppression (thrombocytopenia, neutropenia, anemia), which requires regular monitoring of blood counts, and an increased risk of fractures and myelodysplastic syndrome/acute myeloid leukemia with concomitant use of abiraterone and prednisone/prednisolone.
The recommended dose of radium-223 dichloride is 55 kBq (1.49 microcuries) per kg body weight, administered as an intravenous injection every 4 weeks for 6 doses. Patient compliance involves adhering to the scheduled injections and monitoring appointments, as well as managing potential side effects.
Alternatives include other systemic therapies such as enzalutamide, abiraterone, docetaxel, and cabazitaxel, as well as bone-targeted agents like denosumab and zoledronic acid. These alternatives differ in their mechanisms of action, efficacy in prolonging overall survival or delaying skeletal-related events, and their distinct side effect profiles, including hormonal effects, chemotherapy-related toxicities, and osteonecrosis of the jaw.
The cost of radium-223 dichloride therapy can be substantial, and it's important to consider insurance coverage and patient financial assistance programs. Comparative costs with alternatives like oral androgen receptor inhibitors or chemotherapy regimens can vary significantly based on drug pricing, duration of treatment, and supportive care needs.
Long-term limitations and late toxicities can include persistent myelosuppression, an increased risk of secondary malignancies such as myelodysplastic syndrome and acute myeloid leukemia, and potential for increased bone fractures, especially with prolonged use or in combination with other bone-targeting agents.
Monitoring for efficacy involves assessing pain control, functional status, and prostate-specific antigen (PSA) levels, although PSA response is not a reliable indicator of radium-223 efficacy. Adverse events require regular complete blood count monitoring (before each dose and as clinically indicated), and vigilant assessment for new or worsening bone pain, fractures, and signs of myelosuppression.
Compare Radium
VS
DeepXray vs Radium
DeepXray focuses on AI-powered analysis of X-rays specifically for osteoarthritis and osteoporosis, offering automated measurements and disease assessment for orthopedic and rheumatology specialties. Radium Read, on the other…
VS
BriefCase vs Radium
Aidoc's BriefCase is an FDA-cleared AI solution focused on communication and collaboration for critical medical imaging findings across various CT scans, aiming to streamline workflow and prioritize urgent…
Suggest an Edit →
| Last Verified: 2026-07-20 | First Added: 2026-07-19
Investors who backed Radium
Funded through the company that built this tool.
