Privacy-Centric Multi-Class Detection of COVID 19 Through Breathing Sounds and Chest X-Ray Images: Blockchain and Optimized Neural Networks

Abstract

The CoronaVirus Disease of 2019 (COVID19) pandemic poses a significant global challenge, with millions affected and millions of lives lost. This study presents a privacy-conscious technique to early identification of COVID19 using breathing sounds and chest X-ray images. Using Blockchain and optimized neural networks, the suggested solution provides data confidentiality and accuracy. Chest X-ray pictures are preprocessed, segmented, and feature extracted using modern algorithms. Breathing sounds are treated simultaneously using tri-gaussian filters and mel frequency cepstral coefficient features. A progressive split deformable field fusion module combines audio and visual characteristics. The proposed Dual Sampling dilated Pre-activation residual Attention convolution Neural Network (DSPANN) improves classification accuracy while decreasing computational complexity using augmented snake optimization. Furthermore, a privacy-focused blockchain-based encrypted crypto hash federated method is used for safe global model training. This complete method overcomes COVID-19 detection issues while simultaneously prioritizing data privacy in healthcare applications. The proposed framework exhibited recognition accuracy rates of 98%, specificity of 97.02%, and sensitivity of 98%.

Keywords

• Early detection
• COVID 19
• multimodal fusion
• deep learning
• breathing sounds
• chest X-ray images