Significant privacy issues arise when medical data is shared and stored in cloud computing settings, where third-party service providers provide resources. Current privacy protection methods include privacy by policy, cryptography, and statistics. However, different segments of medical data require tailored privacy solutions. A robust solution for medical database sharing in cloud technology should support various data access paradigms with multiple privacy benefits. To address this, we introduce HDLFLSTM_KeyGen for medical data privacy preservation. Initially, input data undergoes a privacy preservation phase, employing functionalities like the Kronecker product, encryption, and secret key generation with trust factors for security. The secret key is generated using HDLFLSTM, which integrates Hierarchical Deep Learning for Text classification (HDLTex) and Deep Long-Short Term Memory (DLSTM).