A smart and secured blockchain for children's health monitoring using hybrid encryption and adaptive machine learning techniques

Health care monitoring of the children is needed to prevent the children from any diseases during the child's growth. The mental and physical conditions monitoring of children at the age of 5-10 is necessary and it is the parent's duty. Nowadays, healthcare systems utilize the Internet of Things (IoT) for many purposes. The actuators and sensors are used in the IoT for providing treatments as well as monitoring the patient's health. For managing and sharing electronic medical records, blockchain plays an important role in securing the data. To track and monitor the patient's physiological conditions, IoT-based systems are useful. However, these approaches have some drawbacks. Privacy leakage and rollback attacks are common problems that occur in Permissioned blockchains, even if they provide better scalability and throughput. Due to the structure of blockchain, the data-retrieving process in healthcare monitoring systems is difficult. To solve these problems, we implemented a new machine learning-aided technique to monitor and secure the children's health data in the blockchain environment. Initially, the children's healthcare data are accumulated from the public data resources that are involved in the Hybrid Encryption Technique (HET). In the HET, techniques such as Elliptical curve cryptography (ECC) and Attribute-Based Encryption (ABE) are integrated. After performing the encryption method effectively, the encrypted data are stored in the blockchain. Further, to perform the health monitoring, the doctor decrypted the data utilizing the same HET. In this approach, to monitor the children's health data, the Hybrid Adaptive Machine Learning (HAML) approach is adopted, which is the combination of both Conditional Random Fields (CRF) and Multi-Layer Perceptron (MLP). Here, parameters present in this approach are optimized with the support of the Modified Binary Battle Royale Optimizer (MBBRO) algorithm. In the end, the comparative estimation is performed with certain performance criteria to justify the effectiveness of the developed approach.