Healthcare service quality has been improved by integrating the Internet of Things (IoT) with conventional medical networks. In contrast, device-mounted sensors and wearables employed in Healthcare Systems (HS) monitoring and data transmission ongoing over unprotected open channels to adjacent devices. The effectiveness of operations is being improved by the link among IoT devices and computers, yet it allows attackers to commit a variety of cyber-attacks that could jeopardize patients under vital observation. Using Deep-BiLSTM in a healthcare IoT system for Secure Data Transmission is presented in this paper. In particular, the first unique there is a suggested blockchain design that assure data security and reliability transfer through the use of the Zero Knowledge Proof (ZKP) mechanism. The validated data is then utilized to create a deep learning framework for identification of intrusions in the HS network. A Bidirectional Long Short-Term Memory (BiLSTM) and Deep Convolutional Neural Network (DCNN) are integrated to create a highly efficient intrusion detection method. experiments using two sources of open data (CICIDS-2017 and ToN-IoT) has been used to compare the proposed method with 96% better performance. The suggested BD-BiLSTM methodology has 98% precision, accuracy, recall, and F1 score, which is pretty high when compared to other approaches. of BDL-SMDTA, PBDL and GWMNN.