FLED-Block: Federated Learning Ensembled Deep Learning Blockchain Model for COVID-19 Prediction

Abstract

With the SARS-CoV-2's exponential growth, intelligent and constructive practice is required to diagnose the COVID-19. The rapid spread of the virus and the shortage of reliable testing models are considered major issues in detecting COVID-19. This problem remains the peak burden for clinicians. With the advent of artificial intelligence (AI) in image processing, the burden of diagnosing the COVID-19 cases has been reduced to acceptable thresholds. But traditional AI techniques often require centralized data storage and training for the predictive model development which increases the computational complexity. The real-world challenge is to exchange data globally across hospitals while also taking into account of the organizations' privacy concerns. Collaborative model development and privacy protection are critical considerations while training a global deep learning model. To address these challenges, this paper proposes a novel framework based on blockchain and the federated learning model. The federated learning model takes care of reduced complexity, and blockchain helps in distributed data with privacy maintained. More precisely, the proposed federated learning ensembled deep five learning blockchain model (FLED-Block) framework collects the data from the different medical healthcare centers, develops the model with the hybrid capsule learning network, and performs the prediction accurately, while preserving the privacy and shares among authorized persons. Extensive experimentation has been carried out using the lung CT images and compared the performance of the proposed model with the existing VGG-16 and 19, Alexnets, Resnets-50 and 100, Inception V3, Densenets-121, 119, and 150, Mobilenets, SegCaps in terms of accuracy (98.2%), precision (97.3%), recall (96.5%), specificity (33.5%), and F1-score (97%) in predicting the COVID-19 with effectively preserving the privacy of the data among the heterogeneous users.

Role of federated learning in healthcare systems: A survey

“…T ρ+T η T ρ+T η+F ρ+F η * 100 [10], [38], [13], [36], [58], [46], [47], [28], [77], [35], [7], [79], [64], [29], [1], [85], [9], [53], [25], [51], [66], [17], [45], [50] Sensitivity (Recall)…”

“…T ρ T ρ+F η [10], [38], [40], [47], [28], [64] , [35], [29], [85], [9], [25] [51], [17], [45], [50] Precision T ρ T ρ+F ρ [10], [40], [47], [28], [64], [35], [29], [85], [9], [25] [51], [17], [50] Specificity T η T η+F ρ [51], [17], [45] F1-Score 2 * P recision * Recall P recision+Recall [10], [40], [4], [47], [28], [64], [35], [29], [85], [9], [17] ROC-AUC Score b a f (x)dx [10], [35] , [9] PR-AUC Score b a f (x)dx [10], [36], [22] Root Mean Square Error (RMSE)…”

A Survey on Blockchain-Based Federated Learning: Categorization, Application and Analysis

“…Lakhan et al [125] designed a framework termed FL-BETS, which was a BCFL-enabled task scheduling framework, to identify fraud of data and protect privacy at a low resource cost. In [126], BCFL was used to diagnose COVID-19 while protecting patients' privacy, and it could also deal with heterogeneous data.…”

DMFL_Net: A Federated Learning-Based Framework for the Classification of COVID-19 from Multiple Chest Diseases Using X-rays

“…On the other hand, medical institutions, the owners of CXR images, are known to prefer training models on their data. This is because medical data are subject to stringent privacy requirements [ 9 , 10 , 11 ], which makes it difficult for medical institutions with fewer samples to train a model with the expected performance for detecting various chest diseases. General deep neural network (DNN) models are well known for general image classification; however, they frequently perform poorly with imbalanced datasets, as discussed in [ 6 ], and models trained on restricted samples typically have low generalization capabilities due to a lack of sample diversity [ 10 , 11 , 12 ].…”

“…However, the existing FL-based COVID-19 and chest disease detection algorithms have the propensity to employ the most prevalent techniques FedAvg [ 15 ], and maximize the number of objectives, including latency, energy use, and privacy. There have only been a few studies [ 5 , 6 , 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 ] that have looked at the real-world distribution characteristics of COVID-19 instances and built a one-of-a-kind central aggregation method. This strategy achieves a successful adjustment to the complex sample environment of COVID-19 and at the same time focuses on enhancing the accuracy and stability of the global model.…”