SOW-Justificationblackford.docx

National Institutes for Health (NIH) have an active development program for AI algorithms, generally the outcomes of grant-supported projects by academic clinicians. These algorithms are typically developed and tested on a manually selected cohort of patients, collated and processed as part of the grant-supported work.

Although there is significant potential benefit to the wider patient population from the use of these algorithms to improve outcomes and efficiency in NIH’s more routine clinical practice, the complications of hosting and maintaining the algorithms precludes their use. As a result, algorithms representing significant investment are left orphaned at the conclusion of the trial / grant that funded their development. An opportunity exists for NIH to deploy these applications at scale within their existing systems across their broader clinical radiology workflow.

Understanding, that Blackford has a proprietary platform that allows organizations to deploy FDA cleared and commercially available AI algorithms and specialty applications NIH were keen to examine how this technology could be utilized. Through this process NIH explored how they can simultaneously manage the deployment, hosting, integration, and monitoring of multiple value creating applications by leveraging Blackford’s experience and technology.

Working together with Blackford, NIH will be able to deploy their algorithms and applications across their clinical radiology workflow via the Blackford Platform. The Platform:

· Hosts multiple AI and traditional algorithms to process images.

· Automatically determines which algorithms are relevant to run given anatomical and clinical indications available in the RIS order or DICOM headers.

· Allows manual routing of studies from modalities, routers, PACS, and any other DICOM entity to trigger specific algorithms.

· Integrates into the existing PACS and the rest of radiology IT architecture without noticeably impairing performance for Clinical users regardless of site study volume.

· Provides a web-based monitoring solution that allows users and support teams to assess performance and number of studies being processed by each algorithm.

· Deploys via onsite hardware, securely in the cloud, or a combination of both technologies.

Due to the nature of the agreement, custom engineering work is required from both sides to get applications and algorithms operating on the Platform effectively. Based on Blackford’s experience integrating their Platform Partners applications, they are uniquely qualified and experienced to support these efforts.

The Blackford Platform will allow NIH to realize the full value of the solutions they have proven in the research.