What does this research mean for the field?

A multi-model federated learning framework significantly improves the accuracy and efficiency of deep vein thrombosis segmentation while preserving data privacy. Novelty: ClaimNovelty.NOVEL_FINDING. Consensus alignment: ConsensusAlignment.NEUTRAL.

What question did this study set out to answer?

The aim is to enhance the accuracy of DVT segmentation through a federated learning framework while preserving data privacy.

February 28, 2026Scientific Reports0 citationsOpen Access

Privacy-aware deep vein thrombosis segmentation using a multi-model federated learning framework with the federated averaging algorithm

PBPavihaa Lakshmi BVellore Institute of Technology University NPN. S. Vishnu PriyaVellore Institute of Technology University

Key Points

The aim is to enhance the accuracy of DVT segmentation through a federated learning framework while preserving data privacy.
Developed a federated learning architecture utilizing the federated averaging algorithm.
Trained seven local models on non-IID CT images across three phases with increasing samples.
Evaluated model performance in accuracy, F1-score, Tversky loss, validation loss, and communication costs.
Accuracy increased from 0.9076 to 0.9603 across phases.
F1-score improved from 0.8889 to 0.9521.
Tversky loss decreased significantly from 0.1111 to 0.0479.
Validation loss reduced from 0.910 to 0.061, showing consistency in model performance.

Abstract

Deep Vein Thrombosis (DVT) is the formation of blood clots in the deep veins of the calf, requiring precise Computer Tomography (CT) scan segmentation for accurate diagnosis and treatment. We proposed and developed an efficient Federated Learning (FedL) architecture using the Federated Averaging (FedAvg) algorithm. Seven distinct local models were designed and trained on non-independent and identically distributed (Non-IID) CT images to maintain data privacy and security, enhancing DVT segmentation efficiency and accuracy. The global model was progressively improved by aggregating the local model’s weights using FedAvg algorithm. Our algorithm was evaluated in three phases using datasets of 1000, 2000, and 3000 samples to assess the global model’s performance. Phase 1 involved three clients, each with unique local models (Convolutional Neural Network (CNN), Sequential, and Semantic). While, Phase 2 expanded to five clients, incorporating additional models (U-Net and VGG Net-19). In Phase 3, scaled to seven clients with advanced models (Modified U-Net and Modified-Net). Empirical results across Phases 1–3 showed significant gains with increasing dataset size –attaining higher Accuracy (0. 9076 0. 9603) and F1-score (0. 8889 0. 9521), while Tversky Loss decreased to (0. 1111 0. 0479). Notably, our framework proved consistent improvement across all phases, achieving a reduction in validation loss from 0. 910 to 0. 061 and a communication cost increase from 14 MB to 3279 MB with increasing model scales. The average training time rose proportionally (7. 67 s 18, 702 s) while maintaining robust differential privacy preservation (1. 24) and improved client heterogeneity (0. 13\, \, 0. 51), demonstrating our framework’s scalability and stability across heterogeneous environments.

Demander à l'IA

Bookmark

View Full Paper