Mike Peresie
Hospitals, health systems and imaging centers are facing a growing radiology capacity challenge.
Imaging volumes continue to rise, radiologist availability remains constrained, and the pressure to deliver timely, high-quality results is increasing across the care continuum. MosaicOS is built to meet that challenge.
By embedding AI into the radiologist’s workflow, Mosaic helps expand radiologists’ capacity and improve turnaround times. Even more importantly, it is designed to strengthen clinical quality, with more consistent reports today and AI capabilities now in clinical study to help radiologists detect clinically significant findings.
In this conversation, Mike Peresie, president of Mosaic Clinical Technologies™, explains how radiologists using Mosaic can deliver faster, higher-quality care, and why hospitals should make Mosaic part of their radiology strategy.
Why was Mosaic created?
Mosaic was created to elevate healthcare for every patient through clinical intelligence. That means bringing together clinical expertise, AI and advanced technology in a way that helps radiologists deliver better care. Radiology sits at the center of diagnosis and care planning, but the specialty is under significant pressure. Mosaic is designed to help address that pressure by giving radiologists better tools, improved workflows and more intelligent support in their care delivery.
Our focus is not technology for technology’s sake. It is about using technology to make healthcare better for patients, radiologists and hospitals and health systems that depend on timely, accurate imaging.
What challenge is Mosaic helping hospitals and health systems solve?
The challenge is capacity. Radiology is already experiencing a meaningful capacity shortage, and imaging demand continues to grow much faster than new radiologists are starting their careers. That creates real operational and clinical challenges for hospitals and health systems: longer turnaround times, growing backlogs, higher costs and delays in care. Those delays matter. When a patient is waiting for imaging results, especially in an urgent or emergent setting, they are waiting for answers that can guide the next step in their care.
Capacity pressure can also affect quality. When radiologists are working against growing backlogs, fatigue and time pressure make it harder to catch every finding. Mosaic is focused on helping reduce that friction by expanding radiologist capacity and improving the flow of high-quality imaging services across the care team.
How does Mosaic approach AI in radiology?
Our approach is simple: Embed AI solutions into the radiologist’s workflow to deliver better outcomes and better radiologist experience. We are not building technology to replace radiologists. We are building technology to empower radiologists. Radiologists bring clinical judgment, accountability and patient context. AI can help by reducing repetitive work, enhancing detection and image interpretation, drafting reports and supporting more consistent workflows.
When radiologists are empowered with AI solutions that actually create value, care quality and patient outcomes can improve, and health systems can achieve better operational performance and a more sustainable approach to imaging growth.
What makes Mosaic different from traditional radiology technology?
Mosaic is not just another solution. It is an operating system for radiology. Traditional radiology technology often requires radiologists to move between disconnected tools, systems and workflows. Mosaic is designed to serve as a single orchestration and intelligence layer centered around the patient. It brings together reporting, workflow, AI and clinical context in one integrated platform. That matters because radiology does not need more fragmentation. Health systems need technology that reduces complexity, supports radiologists in the flow of work and helps deliver more timely, consistent patient care.
It is also built by people who understand radiology from the inside. Mosaic is developed with and for Radiology Partners, the largest radiology practice in the world, and shaped by direct feedback from practicing radiologists. That means it is designed around real clinical workflows, not assumptions about them.
How does Mosaic Reporting™ support radiologists and health systems?
Mosaic Reporting is a next-generation reporting capability that lets radiologists speak naturally rather than tabbing through fields or relying on rigid workflows. Using natural language and large language models, it helps populate report templates, improve consistency and reduce unnecessary dictation burden. It also creates opportunities for automation, such as inserting evidence-based clinical guideline language when appropriate. The result is less time spent on manual tasks and more time focused on clinical decision-making.
Mosaic is built not only for today, but for the future of radiology. We created Mosaic Reporting to work optimally with draft reports because we expect AI drafting to become widely available within the next one to two years. It is a powerful tool on its own and even more valuable as an editor that helps radiologists review, refine and manage those drafts.
For health systems, that can mean more efficient workflows, more standardized reports and a stronger foundation for scaling imaging services. For radiology practices, it can help improve service levels and turnaround times, reduce physician burnout and support the delivery of high-quality patient care.
What are investigational vision-language models (VLMs)*,and why are they important?
The VLMS are developed by Cognita Imaging, Inc. (Cognita), Mosaic’s advanced AI lab. They function a lot like an AI resident. It reviews imaging studies, including current and prior images, along with clinical history, and the VLM surfaces positive findings. Those findings can then be used to generate a draft report for the radiologist to review, edit and sign.
This is different from many traditional AI tools that are designed to detect one finding at a time. The VLMs are intended to review the exam more broadly and support the radiologist across the reporting workflow. It can identify findings, describe where they are located, compare them with priors, provide measurements and help generate report language. The radiologist remains in control. The value is that AI helps flag potential findings for radiologist review, reduce burden, improve consistency and create additional clinical capacity while preserving physician oversight.
What results has Mosaic seen so far?
We are already seeing meaningful impact. Mosaic Reporting has been deployed to more than 2,000 radiologists across 65 Radiology Partners-affiliated practices. More than 1,500 radiologists are using the VLMs as part of our clinical study, and we are continuing to expand usage and improve the technology.
In our ongoing clinical study of our VLMs, we observed early signals of improved detection of clinically significant findings, such as pancreatic lesions, stenosis, suspicious bone lesions and pneumothorax. On average, we have seen more than 30% efficiency improvement for drafted exams and approximately 20% efficiency improvement in ultrasound capture. We have seen more than 200,000 additional findings surfaced through the VLMs. We are seeing improvements in turnaround time and reduced backlogs, and we expect those gains to grow as more exam types become eligible and more capabilities are added to the platform.
Why should hospitals and health systems pay attention now?
Because the capacity challenge is not going away and the enhanced detection rates from vision language models demonstrate the potential to save lives. Health systems are being asked to do more with less. Patients and referring physicians expect faster answers, and traditional staffing models alone cannot close the gap. When radiologists are under constant pressure to keep up with growing backlogs, burnout rises and quality can suffer.
Mosaic gives hospitals and health systems a way to think differently about radiology capacity, with one operating system designed to scale and strengthen clinical quality.
This is an opportunity to modernize imaging operations in a way that supports radiologists, improves service levels and helps patients receive care faster.
How does Mosaic improve patient care?
Patient care is the reason Mosaic exists. When radiology workflows are delayed, patient care is delayed. Faster, higher quality imaging results can help physicians make decisions sooner, move patients through the care journey more efficiently and reduce unnecessary waiting.
AI, specifically the advanced vision language models we have developed, have demonstrated remarkable potential to improve detection of clinically significant findings. AI is able to ingest all the pixels across all the slices of an image and analyze them all in an instant. As a result, advanced AI models can detect findings that might be hard to see by the human eye on a first review. With AI models scanning images for missing findings, radiologists are no longer reviewing an image alone. Together, detection rates and accuracy are improved.
Mosaic’s AI has demonstrated the ability to support enhanced detection, better reporting consistency and faster turnaround times. Over time, that can help hospitals improve care coordination, support timely downstream care and strengthen the role of radiology as a strategic part of the care team.
How does Mosaic continue to improve over time?
Radiology Partners’ scale gives us a large, real-world environment where radiologists are using the technology, giving feedback and helping shape the roadmap. That feedback helps us improve the platform, refine AI models, enhance the user experience and add new capabilities.
As the platform improves, it can help deliver better outcomes for clients, patients and radiologists. That creates more trust, more adoption and more data to continue improving the system.
We are focused on getting better every day. Even small improvements compound over time, and that is how we believe Mosaic will continue to evolve quickly and meaningfully.
What is coming next for Mosaic?
We are moving quickly. We are continuing to enhance reporting features and workflows, expand our clinical study to deploy new VLMs across additional exam types, and build out broader MosaicOS capabilities, including worklist, information hub and viewer functionality. We are pursuing regulatory clearance for our vision language models to enable external customers to gain access to the VLMs and obtain the benefits from the breakthrough innovation we are building. Our goal is to make Mosaic more comprehensive, intuitive and valuable for radiologists and the organizations they serve. We are still early, but the pace of progress is significant.
What should radiology practice, hospital and health system leaders do if they are interested in Mosaic?
Reach out and start a conversation. If your organization is facing radiology capacity constraints, turnaround time pressure, workflow fragmentation or rising demand for imaging services, Mosaic was built to help address those challenges.
We welcome the opportunity to show you what Mosaic looks like in practice, discuss your organization’s radiology strategy and explore how Mosaic could support your sites.
Radiology is changing quickly. Health systems that act now have an opportunity to build a more scalable, intelligent and future-ready imaging model, one that helps radiologists deliver high-quality care to every patient.
*CAUTION: These vision-language models are an investigational device not currently approved for sale in the United States.
Mosaic Clinical Technologies is the technology services division of Radiology Partners (RP), powering the future of radiology through MosaicOS—a proprietary imaging intelligence platform designed to meet the specialty’s most pressing challenges. A fully cloud-native and AI-native operating system, MosaicOS is where innovation meets impact, seamlessly integrating diagnostic technologies, AI-powered tools and smart workflows into a single scalable solution. Mosaic Clinical Technologies™ supports RP’s national network of affiliated practices and commercial partners across the imaging landscape, redefining what is possible in enterprise imaging. Connect with us on LinkedIn and YouTube. Contact us at info@MosaicClinical.ai.
