NVIDIA Research: First Privacy-Preserving Federated Learning System for Medical Imaging

Показать описание

#NVResearch #Healthcare #FederatedLearning

Рекомендации по теме

Комментарии

In machine learning the medication image are just provided with labels sick or not without knowing about the patient or where he is or gender or age. So I don't see yet the privacy issue. However the data stored in hospitals contains patients identification information alongside with the medical image. So making a query on those servers seems to be less secure than the traditional method for which Hospital just provides the required data without patients identification information. I suppose I have to read this paper to have insight into what they are saying. If no it seems like make patients data available for hackers

bediosoro

This is just an excuse to centralise large amounts of data in one place, instead of placing it in a small pile. The reality is that all data relevant to the hospital that requires your records will have your data, even if its not replicated in the medical imaging system. So in this way, its just one more place to have your data stored, but because its all in a large pile, the motivation to leak it or expose it is so much more. Now, privacy compliance is easier, because it should be mastered in one location, not 1000s. The real story here is the AI smarts that can read imaging data and return analysis that is better than a Doctor viewing a recorded image.

icurafuse

NVIDIA Research: First Privacy-Preserving Federated Learning System for Medical Imaging

NVIDIA Research: First Privacy-Preserving Federated Learning System for Medical Imaging

Federated Learning Research using NVIDIA Flare on AWS

Federated Learning for Medical AI and Triaging COVID-19 Patients

NVIDIA Clara Federated Learning

Privacy Preserving Federated Learning⎟FL workshop

Privacy Preserving Federated Depression Detection From Multisource Mobile Health Data

Privacy-preserving Machine Learning (Federated Learning): What, Why, and How? - Part2v4

ID 38: Privacy-preserving patient clustering for personalized federated learning

Federated Learning EXPLAINED (Tutorial + Research + Frameworks)

An Introduction to Scaleout: Federated Learning and Privacy-Preserving Technology in 15 Minutes

Privacy Preserving AI (Andrew Trask) | MIT Deep Learning Series

CryptGPU: Fast Privacy-Preserving Machine Learning on the GPU

Tutorial: Decentralized and Privacy-Preserving ML via TensorFlow Federated

Unlocking the Potential of ML in Healthcare in a Privacy-Preserved Way with Federated Learning

Privacy-preserving Machine Learning (Federated Learning): What, Why, and How? - Part3v1

Privacy-preserving Machine Learning (Federated Learning): What, Why, and How? - Part1v2

Privacy Preserving Machine Learning

NDSS 2021 6C-1 POSEIDON: Privacy-Preserving Federated Neural Network Learning

Federated Learning for Healthcare AI: NVIDIA and Rhino Health Accelerate Research Collaborations

FedVault: Efficient Gradient Outlier Detection for Byzantine-Resilient and Privacy-Preserving FedML

Protect Privacy in a Data-Driven World: Privacy-Preserving Machine Learning

“Learn Everything, Know Nothing: The Medical Applications of Privacy-Preserving AI”

Privacy-preserving Machine Learning (Federated Learning): What, Why, and How? - Part3v2

NVIDIA FLARE 101: Overview and Getting Started