Hamza Khalid

Academic

Career

Verbatim text

MedGnosis
MedGnosis is a decentralized platform that provides secure and privacy-preserving collaborative medical machine learning using the power of federated learning. The platform offers several features, such as:
• Privacy-Preserving: MedGnosis ensures that client data is always protected using advanced cryptographic techniques.
• Decentralized: The platform uses a peer-to-peer network to distribute the model training process across multiple clients, ensuring that no single entity has control over the model.
• Flexible: MedGnosis is highly customizable, allowing clients to select the type of data they want to contribute and the level of participation they are comfortable with.
• Secure Aggregation: The platform uses a secure aggregation protocol to ensure that the model is trained on aggregated data while preserving the privacy of each client.
We use a technique called Federated Learning to train machine learning models on data from multiple sources without centralizing the data in one place. This preserves the privacy of the data and ensures that it never leaves the client’s device. Additionally, MedGnosis uses a decentralized blockchain-based architecture to ensure that the data is not tampered with and that the model is not maliciously modified

MedGnosis
ABSTRACT
A Decentralized Privacy Preserving Federated Learning System For Medical Biagnosis Using Blockchain
ARCHITECTURE
Supervisor
Dr. Hina Binte Haq
Members
Hamza Khalid - I19-2011
Hamza Iftikhar - I19-2003
Technology Used:
Python, Tensorflow, React, Flower, Ethereum, IPFS, Ganache
Supervisor Name:
Dr. Hina Binte Haq
Group Members:
Hamza Khalid (I19 - 2011)
Hamza Iftikhar (I19 - 2003)

AI enrichment