Assessfy Pvt. Ltd Advanced 5 milestones 100 marks

Privacy-Preserving Federated Learning for Medical Imaging

Target year: BE Sem 7-8 (Major Project Phase-I/II) AICTE: 6 credits · ~150 hrs Bloom: Create / Evaluate MU CBCS: IT801 / ITDLO8021 BE Project

Real-world project · AICTE-aligned · AI-graded · Audit-ready certificate

Enroll as Student Mentor this project

Milestones

Available mentors

Enrolled students

Core skills

About this project

Implement Federated Learning for chest X-ray classification (pneumonia vs normal) across 3 simulated hospitals — model weights are aggregated without raw data ever leaving each hospital. Add differential privacy noise. Compare model accuracy vs centralized training. Validate privacy guarantees.

Course Learning Outcomes (CLOs):
CLO1: Apply federated-learning algorithms (FedAvg, FedProx).
CLO2: Implement differential-privacy noise + epsilon-budgeting.
CLO3: Quantify privacy-vs-utility trade-off empirically.
CLO4: Design a system that simulates a real multi-hospital deployment.
CLO5: Communicate privacy guarantees to non-experts.

Industry/societal relevance: India's DPDPA + healthcare AI (Practo, HealthifyMe, Tata 1mg, Niramai) make privacy-preserving ML a hiring filter for next-gen healthcare-AI engineers.

Milestones

1. FL Architecture + Setup

15 marks 18d

Design 3-hospital simulation in Docker. Flower server + 3 clients. Each client gets disjoint X-ray subset.

2. Centralized Baseline

20 marks 14d

Train ResNet-18 on combined data centrally. Report test AUC. This is the upper bound.

3. FedAvg Federation

25 marks 21d

Implement FedAvg with 10 rounds × 3 clients. Compare accuracy vs centralized. Document gap.

4. Differential Privacy

20 marks 21d

Add DP-SGD via Opacus. Sweep ε ∈ {1, 3, 10}. Plot privacy-vs-utility curve.

5. Final Report + IEEE Paper

20 marks 26d

Defensive analysis: membership-inference attack. 14-page IEEE paper + oral defense.

Open internships using this project -->

Skills you'll learn

Federated Learning (FedAvgFedProx)PyTorch + Flower frameworkDifferential Privacy (Opacus)Medical Image ClassificationPrivacy + Healthcare EthicsReproducible Research

Tools used

Python 3.11PyTorch 2.xFlower (federated learning framework)Opacus (DP-SGD)Kaggle Chest X-Ray datasetDocker Compose (simulate 3 hospitals)GitHubGPU access

Prerequisites

Deep Learning; Distributed Systems; Python advanced; intro to privacy/cryptography helpful

Available mentors

No mentors have signed up for this project yet.

Be the first to mentor

You'll earn — Certificate (PDF)

AICTE-aligned Project Completion Certificate

A formal, audit-ready PDF certificate issued by Assessfy + your institute on successful completion. Includes AICTE credit hours, your evaluator's signature, and a QR code for third-party verification.

AICTE-aligned

Certificate of Project Completion

This is to certify that

has successfully completed the project

Privacy-Preserving Federated Learning for Medical Imaging

Scan to Verify

Date

Signature

Auto-issued on completion QR-verifiable

You'll earn — Digital Badge

Shareable LinkedIn / Resume Skill Badge

A compact, verifiable Open-Badges-2.0-compliant digital credential. Add to your LinkedIn profile, GitHub README, or resume in one click. Recruiters can validate authenticity via a unique URL.

Advanced

Privacy-Preserving Federated Learning…

Assessfy

Auto-issued on completion One-click LinkedIn add