Managing the AI avalanche

In a recent blog, we introduced you to the challenges introduced by the recent massive growth in AI radiology applications. Today we’d like to focus on the

post image

In a recent blog, we introduced you to the challenges introduced by the recent massive growth in AI radiology applications. Today we’d like to focus on the positives and how AI is uniquely capable of helping make radiology departments faster and better.

Radiology AI is maturing Today’s AI algorithms typically focus on providing value by increasing revenue and/or accuracy, or decreasing costs, read-times and/or turnaround times. With these improvements, downstream value is captured within a radiology department’s referring community through faster times to treatment, quantification and/or illustration that assists in patient education. As these algorithms increase in maturity, both their positive and negative predictive values continue to climb, all but eliminating the “alert fatigue” that plagued legacy products in the early years. However, the amount of time for an average-size hospital or imaging center to evaluate, select, contract, fund, set up and deploy a given solution could range between six months and two years. This leads to the challenges faced by radiology and IT departments in selecting which AI algorithm to start with, or which one to evaluate next. Choose wisely, choose a platform approach In today’s market, capital and IT resources are limited. And, with long budgeting cycles, hospital administrators and radiology leadership most choose wisely. With its docker-container approach, Blackford’s platform can help streamline the selection of AI algorithms. With more than 35 regulatory cleared AI algorithms to choose from, Blackford can also deploy algorithms that are pending regulatory clearance, and can support internally-built AI algorithms for research purposes. Blackford’s contracting approach cuts through the red-tape of the procurement process, providing a single agreement that covers almost all approved applications, without requiring a new set of legal agreements each time an algorithm gets deployed. From an IT perspective, adding applications is incredibly straightforward. Once the platform is installed, the vast majority of deployments only require an account to provide access to the relevant DICOM data – Blackford takes care of the rest. Additionally, once up and running, the update process for the platform and algorithms is straightforward and requires minimal local IT administrative support. Ultimately, Blackford’s single dedicated platform transforms the deployment, orchestration, management and support of multiple imaging applications, providing hospitals and imaging centers with the ability to manage a broad portfolio of applications while being responsive to future application demands.