AI Based Secure Analytics of Clinical Data in Cloud Environment: Towards Smart Cities and Healthcare

Recently, health care data comprises of a huge number of information which is regarded as a challenging one for manual maintenance. Owing to the development of big data in the healthcare and biomedical communities, the study of accurate medical data aids the recognition of early-stage disease prediction. As there were several techniques employed for the classification of disease, there were some limitations like low prediction and accuracy rate. To overcome this, deep learning-based classifier is presented. An Artificial intelligence scheme for disease prediction and privacy preservation using Identity based dynamic distributed Honey pot algorithm is proposed in this work for cloud security. Initially, the input medical dataset is preprocessed using normalization technique in which the missing values are replaced and the unwanted data are removed. Whale Optimization based passive clustering is employed for clustering huge data. Multi-scale grasshopper optimization is employed for the process of optimization to get best fitness value. Then the feature extraction and using Robust Shearlet based Feature Extraction algorithm. The classifier is responsible for predicting the disease and for this a Modified Long Short-Term Memory-Convolutional Neural Network (MLSTM-CNN) based classifier is used which provides high accuracy of prediction. Then the data are stored in cloud server or maintenance and monitoring purpose. It is essential to preserve the personal heath record from cloud attack. So as to satisfy this privacy reservation scheme cryptographic techniques are employed in this work. The PHR maintenance is done initially using Identity based dynamic distributed Honey pot algorithm for encryption. Finally, the performance analysis is carried out and the comparative analysis of proposed and existing techniques is done to prove the effectiveness of proposed scheme.