At HLTH 2025 in Las Vegas, conversations about artificial intelligence in medical imaging often centered on impressive technical capabilities—speed, accuracy, detection rates. But RapidAI CEO Karim Karti offered a different lens: the fundamental distinction between AI that simply identifies problems and AI that actually helps solve them. In a diagnostic imaging market where “AI-powered” has become ubiquitous marketing language, that distinction matters more than ever.
Clinical AI That Matches Patients to Therapies
Karti emphasized that RapidAI‘s differentiator lies in employing artificial intelligence to solve genuine clinical problems rather than merely flagging potential issues. The company’s current focus on ischemic stroke exemplifies this philosophy—confirming whether a patient is an appropriate match for specific therapies, not just detecting the presence of a stroke.
According to Karti, this approach creates three interconnected types of value: clinical, operational, and financial. For stroke patients, where every minute of delayed treatment destroys nearly two million neurons, the ability to rapidly confirm therapy eligibility can mean the difference between full recovery and permanent disability.
RapidAI’s journey began in 2011, initially as RAPID software developed to automate CT perfusion imaging analysis. The medical imaging platform has since evolved into a comprehensive suite of AI-powered diagnostic tools that integrate seamlessly with hospital PACS and EMR systems. The company secured $25 million in Series B funding in 2020 to accelerate its global expansion beyond ischemic stroke into hemorrhagic stroke and aneurysm care. The company now operates in more than 2,200 hospitals across over 100 countries, according to company materials, with peer-reviewed research validating its clinical applications in neurovascular imaging.
Platform Strategy: Beyond Single-Condition Solutions
Rather than building standalone tools for individual conditions, RapidAI has developed a radiology AI platform approach that enables healthcare organizations to leverage multiple imaging analysis applications across different disease states. Beyond the company’s foundational work in stroke imaging, the platform now addresses brain aneurysms, pulmonary embolism, and other time-sensitive neurovascular and vascular conditions.
Karti described how the platform enables quantification and tracking through technologies like 3DV (three-dimensional visualization) and GPI (image processing capabilities), though he noted these technical features serve the broader clinical mission rather than existing as isolated innovations.
This platform strategy reflects a population health perspective—viewing stroke care not just as individual episodes but as a comprehensive challenge within defined geographic coverage areas. The goal is establishing standardized care processes that level the playing field between large academic medical centers and smaller community hospitals.
Augmenting Rather Than Replacing Clinical Expertise
A central theme in Karti’s perspective at HLTH was the role of AI in augmenting radiologists rather than replacing them. “I look forward to bringing my own experience in imaging, artificial intelligence and digital health to help push the boundaries of clinical technology and patient care even further and continue to save lives,” Karti said when he joined RapidAI as CEO in 2022, bringing more than 25 years of medical technology experience from roles including President and CEO of GE Healthcare Imaging.
“I look forward to bringing my own experience in imaging, artificial intelligence and digital health to help push the boundaries of clinical technology and patient care even further and continue to save lives.” — Karim Karti, CEO, RapidAI
RapidAI’s AI augments radiologists, enabling them to become critical members of the patient care team and allowing them to practice at the top of their licensure by automating time-consuming image analysis tasks while preserving their expertise for complex clinical decision-making. For stroke care, where the mantra “time is brain” drives urgency, this division of labor proves particularly valuable.
RapidAI’s worklist prioritization feature exemplifies this augmentation philosophy. Rather than simply processing scans in the order received, the system identifies cases requiring urgent attention and surfaces them immediately to radiologists and stroke teams, integrating directly into existing clinical workflows.
Measuring What Matters: Door-to-Needle Metrics
Healthcare organizations implementing RapidAI track specific care metrics that directly impact patient outcomes. Door-to-decision time and door-to-needle time serve as critical performance indicators for stroke programs, with research published in peer-reviewed journals suggesting the platform may reduce radiology report turnaround times when combined with clinician interpretation.
The company reports that its Rapid CTP technology has been utilized in 75% of Comprehensive Stroke Centers in the United States and served as core technology for pivotal stroke trials that expanded treatment windows from six to 24 hours. These landmark clinical trials, published in the New England Journal of Medicine, helped reshape American Heart Association guidelines for acute stroke treatment. RapidAI’s Rapid LVO was also among the first software products to qualify for Medicare’s New Technology Add-on Payment in 2020, signaling payer recognition of its clinical value.
The clinical validation approach distinguishes RapidAI in a market where many AI tools receive regulatory clearance based on technical performance rather than demonstrated impact on patient outcomes or care processes.
Competitive Landscape: A Crowded Medical Imaging AI Market
RapidAI operates in an increasingly competitive clinical AI market for stroke detection and triage. Primary competitors include Viz.ai, which has gained significant market traction with its LVO detection and care coordination platform, Brainomix, a University of Oxford spinout that recently secured FDA clearance and is expanding in the U.S. market, and Aidoc, which offers AI-powered radiology triage across multiple clinical conditions including stroke.
Recent comparative research has highlighted performance differences among these platforms. A February 2025 study presented at the International Stroke Conference found that RapidAI’s Rapid LVO software detected 33% more large vessel occlusion cases than Viz.ai’s competing product in a 1,525-patient comparison. The study author noted the importance of trusting AI accuracy for time-sensitive conditions like stroke. Separately, research on Viz.ai’s platform has demonstrated significant improvements in treatment times and hospital financial performance, underscoring that multiple AI approaches can deliver clinical value.
However, systematic reviews published in medical journals suggest that multiple AI platforms demonstrate high diagnostic accuracy for stroke detection, with implementation differences and workflow integration often proving as important as pure detection metrics. A Canadian health technology assessment of RapidAI noted that while the platform shows potential to improve acute stroke care efficiency, evidence gaps remain regarding long-term patient outcomes and cost-effectiveness—limitations affecting the entire category of clinical AI tools, not just RapidAI.
Implementation Challenges and Equity Considerations
Despite promising clinical results, the field of AI-powered diagnostic imaging for stroke faces significant implementation challenges. Research examining radiology AI adoption has raised questions about the depth of peer-reviewed evidence supporting some commercial products, though RapidAI stands among the most extensively studied medical imaging platforms with multiple published validation studies.
Equity of access represents another critical consideration. While large academic medical centers and comprehensive stroke centers have relatively straightforward paths to implementing advanced AI tools, smaller community hospitals and rural facilities often face technical, financial, and expertise barriers. This disparity is particularly concerning given that stroke mortality rates are often highest in rural counties with limited access to specialized care.
RapidAI’s emphasis on standardization and its integration approach—designed to work with existing hospital infrastructure rather than requiring specialized equipment—attempt to address these equity concerns. The company’s work with smaller health systems suggests a recognition that clinical AI’s value should extend beyond elite medical centers.
The Road Ahead: Expansion Beyond Stroke
Looking forward, Karti envisions RapidAI expanding well beyond its stroke care foundation to address numerous other clinical conditions where time-sensitive imaging analysis and care coordination prove critical. The company’s aneurysm detection capabilities and work in pulmonary embolism represent early steps in this direction.
This expansion reflects a broader market evolution—from point solutions addressing single clinical scenarios to comprehensive platforms that healthcare organizations can leverage across multiple service lines. The successful vendors will likely be those that can demonstrate not just technical sophistication but genuine clinical utility: reduced time to treatment, improved patient outcomes, better care team efficiency, and measurable financial impact.
Karti’s perspective from HLTH 2025 underscores a maturation in how healthcare leaders evaluate medical imaging AI technologies. The conversation has shifted from “can AI detect this condition?” to “does AI help us deliver better care?” That shift from detection to problem-solving represents the next chapter in diagnostic imaging AI’s evolution—one where the technology serves as a genuine tool for radiologists and clinicians rather than a replacement for human judgment. For stroke patients and their families, that distinction could make all the difference.
This original article was created with AI support.
