Medical Imaging · Information Technology · Artificial Intelligence
Dyania Health Collaborates with NVIDIA to Enhance AI Clinical Reasoning in Health Systems
November 18, 2026 · News Release
Source: Cision PR Newswire

Dyania Health has announced a collaboration with NVIDIA to implement AI clinical reasoning within health systems, utilizing their internal infrastructure on a broad scale. Dyania's Synapsis AI analyzes a patient's complete medical history, encompassing years of notes, operative reports, imaging reports, and lab results, to address clinical inquiries based on the chart evidence. This approach leverages NVIDIA’s accelerated computing to perform these analyses securely, without transferring patient records beyond the health system.
Enhancing Clinical Insights
Health systems have accumulated extensive clinical data over decades, yet much of it is challenging to analyze due to its narrative format. Manual chart reviews are time-consuming and limit the scalability of generating registries, outcomes studies, and predictive models. Dyania's Synapsis AI is designed to operate like a specialist reviewing a patient chart—a longitudinal approach that considers all relevant documents to ascertain patient conditions and treatment timelines.
The collaboration combines Dyania’s clinical reasoning capabilities with NVIDIA's high-performance computing resources, allowing for comprehensive analyses of patient records while maintaining security. Synapsis AI operates within the health system's confines on NVIDIA infrastructure, employing tools like NVIDIA TensorRT and the NeMo framework for training clinical models.
Initial Study on Aortic Valve Replacement
The first project under this partnership involved a study at Cleveland Clinic analyzing 7,812 patients who underwent aortic valve replacement between 2005 and 2021. The research aimed to determine the longevity of replacement valves and the potential need for subsequent procedures. Using Synapsis AI within Cleveland Clinic’s environment, the researchers addressed 88 clinical questions for each patient, evaluating parameters such as valve characteristics, comorbidities, medication use, and follow-up results.
The predictive machine learning model developed can forecast valve failure and the risk of re-intervention, highlighting that the implanted valve type is the chief factor influencing deterioration rates. Additional follow-up data, including heart ultrasounds, significantly improved predictive accuracy for future interventions. The model provides estimates of individual patient risks at three, five, and ten years, updating with each new ultrasound.
Future Implications
The heart valve study exemplifies the potential of this collaboration to advance healthcare delivery. The technology aims to empower health systems to gather insights from their patient experiences, leading to improved treatment strategies and personalized care recommendations. Ongoing advancements in Synapsis AI will enhance the accuracy of clinical decision-making, facilitating a shift towards more precision-driven healthcare approaches.
Dyania Health aims to ensure that clinicians do not overlook critical patient information buried within extensive medical records, offering tools to identify care needs and optimize treatment pathways within the walls of the health system.





