Enhanced Consumer Healthcare Data Protection Through AI-Driven TinyML and Privacy-Preserving Techniques

In the recent digital landscape, securing healthcare data stored on personal devices has become imperative due to increasing cyberattacks. Healthcare data inside an organization is often vulnerable to security breaches and requires privacy-preserving mechanisms to ensure secure storage and sharing across various platforms. This research proposes a novel method by integrating TinyML (machine learning) with federated learning (FL) and differential privacy (DP) to ensure the security and privacy of health-related data on resource-constrained edge devices. The proposed method with TinyML looks at patient data, like ECG readings and reports of unusual heartbeats, right on local edge devices that have limited resources, enabled immediate privacy checks while using minimal computing power. Federated learning trains the model on local edge devices by ensuring sensitive data remains on user devices, which are accessed centrally. Finally, differential privacy techniques strengthen security by additionally adding noise to data for safeguarding against malicious attacks without compromising the data’s accessibility. The proposed model achieves a high accuracy of 99%, significantly outperforming the traditional models, like decision trees at 76%, random forests at 98.1%, and federated learning with deep Q at 95%. The proposed methodology provides a scalable and efficient solution for real-time healthcare data sharing by ensuring data privacy and security across healthcare data source environments. The analysis of experimental results validates the model’s efficiency in securing healthcare data, with implications for broader applications in secure, privacy-preserving medical data security analysis.