Photo by Phil Boorman Photography/Getty Images
As imaging professionals gather in Chicago this week at the annual Radiological Society of North America meeting, there are lots of new announcements focused on AI-enabled products and processes for radiology and other imaging modalities.
But there have been a lot of other AI news across healthcare this past month, including new products focused on zero trust in healthcare cloud and hybrid environments, protecting patients at the point of care and safeguarding communications with medical device companies and their providers.
Evolving radiology intelligence
At RSNA 2025, SimonMed Imaging, a vendor of outpatient imaging, and Lunit, an AI developer, announced that they are jointly deploying one of the industry’s first large-scale custom foundation models to enhance patient care and consistently improve radiological reporting speeds and quality.
The effort is expected to leverage AI to standardize for chest X-ray workflows across SimonMed's 175 U.S. locations while preserving radiologists' expertise and final judgment, they said in a joint announcement on Monday.
It's a "leap forward" in radiology AI deployment – "practical innovation with immediate clinical impact," said Dr. John Simon, SimonMed's founder and CEO.
"By training foundation models on our own patient population and reporting style, we’re able to create AI that is highly intelligent, patient-friendly and uniquely ours," he said. "This approach allows us to share and scale the latest technology with millions of SimonMed patients coast to coast."
Existing generic AI models cannot achieve the adaptability and local specialization essential for radiologist-level accuracy, the companies stated.
SimonMed is using Lunit’s Foundation Model Services to fine-tune its CXR foundation models with its own site-specific imaging data in a HIPAA-compliant environment. Lunit's services include model-performance monitoring and drift-alerting capabilities, the companies noted.
While SimonMed's CXR report-generation model is the first deployed on Lunit's FMS, models for mammography and digital breast tomosynthesis are planned for release next year, according to the announcement.
"With FMS, providers like SimonMed can build high-performing models that reflect their own data and workflows in a matter of weeks," Brandon Suh, Lunit's CEO, said in a statement.
Raidium, a precision radiology company, introduced at RSNA 2025 a new AI-native PACS viewer powered by the Curia radiology foundation model.
A spokesperson for the company told Healthcare IT News by email last week that this is the first AI-native PACS viewer built entirely on a radiology foundation model trained on 1B+ medical images.
"The viewer works like an augmented resident: It interprets complete imaging exams; connects CT, MRI and other modalities; executes complex workflows; and interacts via image, text or voice," they said.
The artificial general intelligence is not a PACS console with bolt-on AI plugins; it is more like a cockpit for radiologists than a single AI copilot, they explained.
Raidium’s viewer is capable of handling full-body exams and agentic workflows because it is a multimodal, multi-organ and multitask model. "It can contextualize findings across modalities, automate lesion detection or follow-up, draft structured reports and run entire imaging workflows from a single prompt."
The model outperforms published baselines across 19 radiology tasks from anatomy, oncology, infectious diseases, degenerative disorders and emergency imaging, according to the company. The viewer also features a dedicated tumor-analysis AI copilot that performs 3D tumor segmentation, automatic longitudinal tracking, total tumor burden analysis and survival prediction based on internal representations.
RapidAI and Amazon Web Services announced on Monday that they will accelerate the global deployment of deep clinical AI by co-developing an FM infrastructure that combines clinical depth with cloud scale.
Access to AI-driven insights, based on disease-specific expertise, available to hospitals worldwide could improve diagnostic precision, streamline clinical workflows and expand access to life-saving technologies, the companies said.
"RapidAI is advancing the frontier of clinical AI by pairing domain-specific intelligence with scalable cloud infrastructure," Dr. Rowland Illing, AWS global chief medical officer and director of healthcare and life sciences, said in a statement.
Also, on Nov. 19, Mosaic Clinical Technologies said it acquired Cognita Imaging to scale radiology intelligence and address the acute global shortage of radiologists.
"Nearly half of the world’s population has limited or no access to basic diagnostics," said Louis Blankemeier, Cognita's CEO, in the announcement.
Cognita’s proprietary vision language models power Mosaic Drafting, a genAI copilot that analyzes X-rays and head CTs and drafts preliminary results for physicians to review and confirm. Testing showed enhanced diagnostic accuracy, reduced variability and expanded clinical capacity, according to the company.
"The next era of radiology will be defined by the convergence of clinical expertise and advanced technology," Rich Whitney, Mosaic's CEO, said in a statement.
Advances in trust
Aviatrix, an enterprise cloud security vendor, launched a zero-trust product for workloads as cross-cloud enforcement for AI and cloud-native environments to shield health systems' vast repositories of protected health information, the company said on Nov. 12.
Nearly a quarter of healthcare organizations – 22% – have experienced a cyberattack targeting their medical devices, according to a report by RunSafe Security, according to Tom Davis, Aviatrix's vice president of industry solutions.
A variety of workloads and legacy cyber postures exacerbate operational complexities in health IT. That underscores "how quickly these incidents escalate into direct patient safety risks," he said by email last week.
"The troubling reality is that attackers rarely target isolated assets; they exploit the hybrid cloud seams and trusted pathways created by necessary business decisions."
The company's Cloud Native Security Fabric platform is a unified enforcement layer built directly into the network fabric, capable of securing data flows regardless of workload type, the company said. It leverages a distributed cloud firewall for policy and high-performance encryption to embed zero-trust security across the full hybrid network path to radically minimize lateral movement, contain threat propagation across multi-cloud and on-premises boundaries, and safeguard regulated cloud-to-EHR data flows, the company said.
"The challenge is rooted in technical gaps and fragmented governance," Davis said. "Attackers exploit blind spots created by inconsistent security policy across these multi-cloud and different workload types, and hybrid boundaries, quietly establishing breach chains that compromise sensitive PHI and disrupt vital clinical systems."
DigiCert, an intelligence company, and Citrix NetScaler, part of Cloud Software Group, are now automating certificate management across hybrid environments to deliver centralized visibility, automation and crypto-agility for enterprise application delivery.
The new integration automates the entire lifecycle of SSL/TLS certificates, from issuance to renewal, to ensure uptime, compliance and crypto-agility across hybrid and multi-cloud environments, the companies said in their Nov. 13 announcement.
"The cryptographic landscape is undergoing rapid transformation, and shorter certificate lifetimes are just one part of that change," said Anthony Ricci, DigiCert's AVP of solutions engineering.
Also, Medcrypt, a medical device and vulnerability management cybersecurity platform vendor, and Thirdwayv, which offers software security for connected medical devices, signed an agreement to advance cybersecurity and interoperability standards across the connected medical device ecosystem.
Their goal is to help medical device manufacturers embed security earlier in development, maintain compliance throughout the device lifecycle, appropriately document cybersecurity controls and traceability in regulatory submissions, and deploy secure over-the-air software updates for connected devices.
"With the increasing need to connect mission-critical medical devices to the internet and directly to each other, having access to an existing, previously certified secure connectivity solution is a game changer," Mike Kijewski, Medcrypt's CEO, stated on Oct. 28. "Thirdwayv's decade-long track record of providing comprehensive software and intellectual property solutions for hundreds of millions of smartphone-connected insulin pumps is truly impressive."
Protecting patients
Conduent, a global tech services company, said it is now leveraging generative AI for reportable event detection to enhance the speed and accuracy needed for U.S. Food & Drug Administration compliance requirements.
Medical devices, life science companies and others interact with patients by email, chat, voice call, text, fax, and social and digital platforms. The genAI platform, built on Microsoft Azure OpenAI, reviews interactions like adverse reactions, product complaints and usability issues across all these channels, using genAI to analyze language, context and compliance rules.
"Patient assistance programs and medical information engagements are complex and data-rich, which is perfect for genAI," Kimberly Marshall, Conduent's head of commercial solutions and account management, said in the Nov. 5 announcement. "By combining AI with deep compliance expertise, we’re helping clients report faster, more accurately and more consistently while saving time and money."
Ambience Healthcare, an AI platform for documentation, coding and clinical workflows, launched a new clinical agent tool to support inpatient documentation.
Conditions Advisor is a new capability that helps clinicians identify and document patient complexities at the point of care and bring relevant information to their attention during patient encounters, the company said. It analyzes complete patient data, including records available from outside organizations, to surface evidence that potentially supports missed diagnoses, the company said.
"Clinicians are managing complex conditions, but don't always have time to review extensive chart data," Dr. Will Morris, Ambience's chief medical officer, said in the Nov. 3 announcement.
Also, Levitate, a software-as-a-service (SaaS) platform, announced that it met HIPAA Security Rule requirements on the Vanta compliance platform and is expanding into personalized patient care on Nov. 3.
Healthcare organizations that depend on trust and personal connection require tailored communications that help reduce no-shows, increase repeat visits, communicate important reminders and allow staff to allocate more time to patient care through minimizing repetitive outreach responsibilities, according to Jesse Lipson, Levitate's founder and CEO.
The SaaS product offers email-to-text outreach, custom content and automated reminders, retention campaigns, a social media content library with SEO-optimized websites and AI-powered suggested responses.
"Stepping into the expansive healthcare space is a natural extension of our mission to help professionals nurture authentic, long-term relationships at scale," Lipson said in a statement.
Andrea Fox is senior editor of Healthcare IT News.
Email: afox@himss.org
Healthcare IT News is a HIMSS Media publication.
