● Covered by 1 source · 1 reportMedium impact1 neutral
Google Cloud and MLCommons use Confidential Computing for privacy-preserving medical AI evaluation
Key points
- Google Cloud partners with MLCommons MedPerf for medical AI evaluation.
- Confidential Computing creates secure clean rooms for benchmarking.
- Evaluates AI models on real-world data without exposing patient privacy.
- Applied to Federated Tumor Segmentation (FeTS) for brain tumor research.
Privacy-First AI Evaluation in Medicine
Google Cloud is collaborating with MLCommons through the MedPerf initiative to develop and implement privacy-preserving methods for evaluating medical AI models. This partnership utilizes Confidential Computing to establish secure environments for benchmarking AI models using real-world patient data.
The approach addresses the challenge of validating AI tools on diverse datasets while protecting patient privacy. It ensures that neither the model code nor the patient data is visible to any party, including the hospital, research institution, other participants, or Google, during the evaluation process.
MedPerf and Confidential Computing
MLCommons, a global community, launched MedPerf in 2023 as an open-source platform to standardize medical AI evaluation through federated evaluation. Google Cloud's Confidential Space, which uses hardware-isolated Trusted Execution Environments (TEEs), is central to this initiative.
These TEEs encrypt in-use memory and harden the operating system. For compute-intensive medical AI benchmarking, Confidential VMs extend to GPUs, running on Google Cloud's A3 machine series with NVIDIA H100 GPUs, combining Intel TDX and NVIDIA Confidential Computing technologies.
Secure Data Handling and Verification
Before any patient data is introduced into the workload, the system provides cryptographic proof. This proof verifies that only approved code is running on genuine Confidential Computing hardware and that the environment has been properly secured. This mechanism ensures the integrity and confidentiality of both the AI models and the sensitive patient data.
Impact on Brain Tumor Research
This technology is currently being applied to the Federated Tumor Segmentation (FeTS) initiative, which focuses on brain tumor research. Brain tumors, such as glioblastomas, are rare, making it difficult for individual hospitals to gather sufficient data for accurate AI model training. The federated approach, enabled by privacy-preserving evaluation, helps overcome these data scarcity challenges and addresses issues where models trained in one hospital may not perform effectively in another.
✨ This summary was generated by AI from the outlets' reporting listed below. It is not independently verified and may contain errors — check the original sources. How BrevFeed works →
The daily brief
One email each morning: the day's tech stories, clustered across outlets and summarized. No account needed.
One email a day. Unsubscribe in one click, any time.
Today's brief
Spend a few minutes, get the whole day. Every topic's top stories in one hands-free rundown — listen, watch, or read the transcript.
▶ Play today's brief
New every morning, and the back catalogue is archived by date.
Reporting from
Google Cloud is collaborating with MLCommons' MedPerf initiative to enable privacy-preserving evaluation of medical AI models using Confidential Computing. This partnership allows AI models to be benchmarked on diverse, real-world patient data within secure environments, addressing the challenge of evaluating AI without direct access to sensitive information. The technology is already being applied to brain tumor research through the Federated Tumor Segmentation (FeTS) initiative, helping to overcome data scarcity and model generalization issues.