Blockchain-Based Framework for Secure Transfer of Medical Data for Deep Learning Models

Abstract

The current healthcare system is primarily dependent on EHRs (Electronic Health Records). Due to the use of these systems an enormous amount of data is generated. Although, EHR systems give patients control over their medical data. The current technology is having a hard time ensuring the owner's data privacy and security. According to the Analysis of cybersecurity breaches in healthcare systems, the amount of healthcare attacks reached 45 million individuals in 2021. Additionally with the help of these records, especially the medical images. Accurate diagnoses of diseases like tumors and cancers are crucial for medical researchers. Since most research is conducted on ML models to enhance the accuracy of the diagnoses of disease, it is very important to have the right quality and quantity of data to improve the efficiency, classification, and prediction rates of the ML model. Data is currently stored in centralized systems which affects the quality and cost of data and research. A centralized system is also more likely to fail effecting the reliability of the system. This is why our research suggests blockchain implementation in medical imaging for the improvement of data security and interoperability. It gives us a distributed network that stores transaction information in each block or node, this improves the quality, quantity, and interoperability of data which helps in further ML research. Therefore, this research presents an effective deep learning model based on a Blockchain to secure the transfer network as well as to accurately diagnose brain tumors