Cleveland Clinic Tests Dyania's Synapsis AI on 7,800 Cases

Key Takeaways

1Dyania Health's Synapsis AI analyzed over 7,800 aortic valve replacement records at the Cleveland Clinic between 2005 and 2021

2The tool operates on-site, using NVIDIA GPUs, and produces traceable clinical signals while leaving the final decision to clinicians

3The method is presented as transferable to patient identification for clinical trials, reducing care gaps, and creating registries

Why it matters : Automating the analysis of medical records allows for the processing of large cohorts while preserving data privacy and medical oversight.

Cleveland Clinic has mobilized Synapsis AI from Dyania Health to analyze over 7,800 aortic valve replacement records between 2005 and 2021 and build a predictive model for 3, 5, and 10 years. The tool operates within the hospital on NVIDIA GPUs and leaves the decision-making to clinicians. Dyania claims traceable signals and time savings in record review.

At Cleveland Clinic, a Study of 7,800 Cases Leveraged by AI

Cleveland Clinic relied on Synapsis AI to support a study involving more than 7,800 individuals who underwent aortic valve replacement between 2005 and 2021. Synapsis AI examined each medical record, also incorporating echocardiograms performed after the procedure, which allowed for the development of a model to anticipate valve degradation and the risk of re-intervention three, five, and ten years post-surgery. According to the institution, the AI platform completed the record review for all patients in the cohort, a task that would have required several years of manual work for clinicians. Dr. Lars Svensson, head of the Institute of Cardiology, Vascular, and Thoracic at Cleveland Clinic, believes that using these technologies to gather and analyze information to predict the lifespan of implanted heart valves represents a promising step toward better outcomes and more personalized care for patients.

A Tooled Clinical Reasoning, with Traceability and Medical Oversight

Synapsis AI acts as a clinical reasoning engine that sifts through notes, reports, and lab results, linking information that may be years apart. The system produces traceable clinical signals back to their source in the record, indicating which patients require special attention and why. Clinicians review this evidence and retain control over all decisions, with the tool highlighting clues that the physician must confirm and utilize. This automated record processing frees clinical teams from manual review, allowing them to dedicate that time to patient care.

Le brief IA que lisent les pros

Le brief IA que les pros lisent chaque soir

Les 7 actus IA du jour, décryptées en 5 min. Gratuit.

Inclus dès l'inscription : notre sélection des meilleurs guides & comparatifs IA.

Choisissez votre rythme

Gratuit · Pas de spam · Désabonnement en 1 clic

On-Site Processing and NVIDIA Stack to Preserve Data

Synapsis AI is deployed within the firewall of the facilities, on NVIDIA-accelerated computing. All records from a healthcare system can thus be processed locally, without external transfer or data extraction outside the hospital. Dyania develops and trains its clinical models using NVIDIA NeMo open-source libraries, then optimizes them with NVIDIA TensorRT and TensorRT-LLM to enhance speed and performance on GPU servers installed in hospitals.

Extended Uses Beyond Cardiology

AI is already finding its place in the medical field, ranging from connecting patients for clinical trials to assisting radiologists, hospital teams, and patients at home, all while keeping clinicians at the center of decision-making and leveraging the computational power of NVIDIA technologies. Beyond cardiology, this same method can be used to identify patients meeting clinical trial criteria, address gaps in care in cardiology and oncology, and create clinical registries when relevant information is found in the medical record. Hospitals have decades of patient history, mostly in free text form, which remains complex for thorough human review, something Synapsis AI aims to automate under medical oversight.