UCSF, Fortanix, Intel, and Microsoft Azure announced that they form a collaboration to establish an accurate computing platform with privacy-preserving analytics to accelerate the development and validation of clinical algorithms.
This platform will offer a “zero-trust” environment to protect the intellectual property of an algorithm and the privacy of healthcare data, while CDHI’s BeeKeeperAI will provide the workflows to enable more efficient data access, a transformation of data to several data providers.
Achieving regulatory approval for clinical artificial intelligence (AI) algorithms requires large and detailed clinical data to develop, optimize, algorithm models. These algorithms must be capable of performing across different patient populations, socioeconomic groups, geographic locations. With clinical-grade AI algorithms that can safely operate, such as immediately identifying life-threatening conditions on X-rays, the work was time-consuming and expensive.
This innovation associates all the computing capabilities of Fortanix Intel, and Microsoft Azure to regulate a proven clinical algorithm against a simulated data set. A clinical-grade algorithm will fastly identify blood transfusion in the Emergency Department which compares with valid results. They will also test whether the model or the data were vulnerable at all points.
The validity and security of AI algorithms is a major concern earlier to their implementation into clinical practice. This becomes a major barrier to whether the scaling algorithms can maximize the potential to detect disease, personalize treatment, and predict a patient’s response to their course of care.
This confidential computing technology protects the privacy of patient data by allowing a specific algorithm to interact with a specifically accurate data set which remains the same at all times. Then the data will be placed in a secure place within Azure confidential computing, powered by Intel SGX, and leveraging Fortanix cryptographic functions in which will validate the signature of the algorithm’s image. The data will be then processed in a separate part which securely connected to another enclave holding the algorithm, ensuring multiple parties can influence the system without needing to trust one another.
So, by bringing together these technologies creates an unpredictable opportunity to accelerate AI deployment in real-world environments. In the Future this innovation utilizes HIPAA-protected data within the context of a joined environment, which enables algorithm developers and researchers to conduct multi-site validations. The main vision in addition to validation supports multi-site clinical trials that will accelerate the development of regulated AI solutions.