SYDNEY, Australia & BOSTON, USA – March 6, 2026
Harrison.ai announced that the U.S. Food and Drug Administration has granted 510(k) clearance for its artificial intelligence system designed to triage acute infarcts on non-contrast CT brain scans, marking a significant advancement in AI-powered stroke detection and radiology workflow automation. The clearance represents the company’s ninth FDA authorization covering 13 radiological indications and its second breakthrough-designated device to achieve marketing authorization. The AI tool is designed to identify Acute Ischemic Stroke earlier in the diagnostic process, enabling clinicians to prioritize suspected stroke patients faster and accelerate life-saving treatment decisions.
AI Technology Enables Early Stroke Detection on First CT Scan
Harrison.ai’s newly cleared system introduces a novel approach to stroke triage by analyzing non-contrast CT brain scans, which are typically the first imaging tests performed when a stroke is suspected. Most existing stroke detection AI tools focus on CT angiography scans that require contrast injection and are often performed later in the clinical workflow. In contrast, Harrison.ai’s technology analyzes the initial CT scan performed in emergency departments, enabling earlier identification of ischemic brain injury before advanced imaging is conducted. The AI system can detect infarcts across six vascular territories of the brain, including the anterior cerebral artery (ACA), middle cerebral artery (MCA), posterior cerebral artery (PCA), cerebellar, basilar, and watershed regions. By evaluating actual brain tissue injury rather than only vessel occlusion, the system offers a more comprehensive approach to stroke triage. This capability allows radiologists and emergency physicians to identify potential stroke cases earlier and move them to the front of the clinical review queue, helping ensure that patients receive timely diagnostic evaluation and treatment.
Clinical Validation Demonstrates High Sensitivity in Detecting Infarcts
According to validation data submitted to the FDA, Harrison.ai’s AI system demonstrated up to 89.2% sensitivity on thin CT slices and 85.7% sensitivity on thicker slice imaging, while maintaining strong specificity across multiple evaluation thresholds. The validation studies used advanced imaging confirmation as the reference standard, ensuring that the algorithm was tested against clinically confirmed infarcts rather than only those visible on standard CT scans. This rigorous validation approach allowed researchers to measure the system’s ability to identify subtle brain tissue damage that may be difficult for clinicians to detect during early stroke assessment. The FDA submission also included comparisons with existing stroke AI systems currently used in radiology workflows. One widely deployed comparator designed for large vessel occlusion triage on non-contrast CT scans reported sensitivity of 63.5%, while Harrison.ai’s technology demonstrated greater than 80% sensitivity and specificity across multiple operating points for identifying ischemic tissue injury across all six vascular territories. This broader diagnostic coverage means the technology can detect more stroke cases that might otherwise go unrecognized in early imaging, potentially improving patient outcomes.
Advertisement
Expanding AI Coverage Across Brain Imaging Workflows
With the addition of acute infarct triage, Harrison.ai now provides eight FDA-cleared findings for non-contrast CT brain examinations, covering both hemorrhagic and ischemic stroke detection within a single imaging platform. Overall, the company has secured 13 FDA-cleared radiology indications supported by nine independent regulatory submissions, demonstrating a strong track record in developing clinically validated AI tools for medical imaging. The technology also received Breakthrough Device Designation from the FDA, highlighting its potential to significantly improve patient care in time-critical neurological emergencies. Harrison.ai’s AI solutions are already deployed across more than 1,000 healthcare sites globally, supporting over 3,400 clinicians in diagnosing medical conditions and improving patient outcomes. The company reports that its technologies have impacted more than seven million patients worldwide, with widespread adoption across hospitals, healthcare systems, and emergency departments. By integrating AI tools directly into clinical imaging workflows, Harrison.ai aims to enhance clinician efficiency while enabling earlier detection of life-threatening conditions such as stroke.
Stroke remains one of the leading causes of death and disability worldwide, and rapid diagnosis is critical to improving survival and recovery outcomes. Because brain cells begin to die within minutes after blood flow interruption, faster identification of ischemic stroke can significantly influence treatment success. Harrison.ai’s latest FDA clearance highlights the growing role of artificial intelligence in medical imaging, enabling healthcare providers to identify neurological emergencies faster, prioritize critical cases more effectively, and ultimately improve patient care.
z
Source: Harrison.ai press release
