UCSF teams with Microsoft, others to fast-track AI in healthcare: 5 details

© Copyright ASC COMMUNICATIONS 2020. Intrigued in LINKING to or REPRINTING this material? View our policies by click on this link.

University of California San Francisco teamed Microsoft Azure, Intel and Fortanix to develop a private computing platform that will accelerate the development and recognition of scientific algorithms.
5 things to know:
1. The platform will have a “zero-trust” environment to protect copyright related to an algorithm in addition to protect healthcare data privacy.
2. The UC San Francisco Center for Digital Health Innovations proprietary BeeKeeperAI will provide workflows for more efficient information access and orchestration throughout numerous information service providers. The platform will also count on Fortanix Confidential Cloud Computing Enclave Manager, Software Guard Extensions from Intel and Microsoft Azures confidential cloud computing infrastructure.
3. When comparing the outcomes of other algorithms, the platform will utilize a clinical-grade algorithm that recognizes the need for blood transfusion for trauma clients in the emergency situation department as a referral standard.
4. The partnership plans to utilize HIPAA-protected information for algorithm designers and scientists to conduct multisite recognition in future phases of the platform. The collaborations goal is to support multisite clinical trials to accelerate regulated AI service development.
5. The information sets will remain in control of the health care organization performing the trial in a secure enclave within Azure cloud. The data is processed in a 2nd enclave, which links to a third enclave holding the algorithm.
” While we have actually been very successful in creating clinical-grade AI algorithms that can safely run at the point of care, such as instantly identifying lethal conditions on X-rays, the work was time expensive and consuming,” stated Michael Blum, MD, associate vice chancellor for informatics, executive director of CDHI and professor of medicine at UCSF. “Much of the cost and cost was driven by the information acquisition, preparation, and annotation activities. With this new technology, we expect to noticeably minimize the time and expense, while also attending to data security concerns.”
More posts on health IT: Maryland hospital reports mini catastrophe due to power failure, electrical issueEpic partners with Lyft: 5 things to knowThe most significant tech investments on leaders minds: 3 hospital officers weigh in

The platform will also rely on Fortanix Confidential Cloud Computing Enclave Manager, Software Guard Extensions from Intel and Microsoft Azures private cloud computing facilities.
The partnership plans to utilize HIPAA-protected information for algorithm designers and scientists to carry out multisite recognition in future stages of the platform. The information is processed in a second enclave, which links to a 3rd enclave holding the algorithm.
” While we have actually been really effective in developing clinical-grade AI algorithms that can safely operate at the point of care, such as right away identifying lethal conditions on X-rays, the work was time expensive and consuming,” stated Michael Blum, MD, associate vice chancellor for informatics, executive director of CDHI and professor of medicine at UCSF.