Securing Clinical AI Pipelines: From DICOM Ingestion to Model Inference
3 partsA 3-part technical monograph examining end-to-end threat vectors in hospital AI deployments: defending 3D volumetric vision models against adversarial perturbations, sandboxing untrusted inference runtimes with eBPF, and preventing gradient leakage in federated training.
00Adversarial Robustness in 3D Volumetric Medical Imaging: Defending Vision Transformers Against Imperceptible Perturbations
01Memory-Safe Sandboxing & eBPF Runtime Monitoring for Untrusted Clinical AI & Agent Execution
02Privacy-Preserving Federated Learning in Hospital Networks: Mitigating Gradient Inversion Attacks