Artificial Intelligence Workflows

Common Challenges

Regardless of the deployment model, there are common challenges that exist when integrating AI algorithms into medical imaging workflows. To ensure maximum value is realized, it is essential that AI algorithm workflow augment and enhance existing clinical workflow by automatically:

• Receiving, fetching, anonymizing, and delivering current and relevant prior studies to the AI algorithms (video)
• Re-identifying AI algorithm results and delivering those results to the correct clinical IT system or archive
• Integrating algorithm results into master patient record and billing workflows

Laurel Bridge AI Workflow Suite

The Laurel Bridge AI Workflow Suite enhances the capability of healthcare providers and AI developers to integrate AI algorithms into their clinical workflows by automating key tasks and by leveraging functionality in our Compass Routing Workflow Manager, Navigator Imaging Retrieval Workflow Manager, and Waypoint Encounter & Modality Worklist Manager.

Our AI Workflow Suite:

• Integrates on-premises, cloud-based, and AI marketplace workflow deployments into existing clinical workflows
• Is HIPAA-compliant
• Adheres to DICOM standards
• Enables the delivery of AI algorithm results to a PACS, VNA, and EMR

On-Premises Workflow

Workflow tools, whether part of the local infrastructure or integrated into the AI deployment, are often required to inspect data attributes and perform normalization and transformation on-the-fly to accommodate the differing data models that exist in multi-vendor AI algorithm environments.

Cloud-Hosted Workflow

This model can be susceptible to connectivity and performance challenges where transfer volumes and/or object sizes are significant and often requires additional privacy and security considerations and failover processing/notifications. Also, ensuring secure and encrypted channels are established between local and cloud-hosted environments without the added complexity of managing separate virtual private network (VPN) connections is a priority.

AI Marketplace Workflow

An AI marketplace can be deployed on-premises, in the cloud, or as a hybrid model depending on site-specific preferences and the underlying requirements of the engaged algorithms. Because a common infrastructure is shared, there is the potential for data models to be aligned across hosted algorithms, which can reduce the complexity associated with data transformations and interoperability.

Benefits

The Laurel Bridge AI Workflow Suite provides the following benefits:

• Integrates study data between AI algorithms and existing clinical systems and workflows
• Offers seamless interoperability between local facilities and cloud-based AI algorithms

Supports standards-based interoperability with third-party applications and clinical systems

Whether you are working with radiology, cardiology, or ophthalmic applications, when anonymized study information is needed for training or research, or to train new machine learning algorithms, we can automate the secure data flow from any number of sources and automatically de-identify patient specific information as needed.