| Abstrakt: |
With the help of pervasive computing, human living has changed into a smarter way using the developments in IoMT, telecommunication technologies, and wearable sensors for ensuring improved healthcare services. IoMT is comprised of certain potentiality for the revolution in the healthcare industry. IoMT is associated with caregivers, healthcare providers, patients, and wearable sensors with software and ICT. The healthcare industry is also a well-known expanding market that has huge demands. It ensures the potential services towards the patients and also provides its contributions to the profits of the health sector. According to the technical advancements, a healthcare system must be developed based on decision-making capacity. Numerous researchers have also focused on involving cognitive behavior in IoT technology. Thus, in this paper, a new smart healthcare system with the help of IoT devices is suggested. Initially, the data is collected from IoMT devices, which are fed to further processing. Secondly, the data pre-processing is carried out to remove the corrupted data and for removing the noise from the data. Thirdly, the features are collected from the pre-processed data through wavelet entropy computation, and deep features are gathered using CNN. Fourthly, both extracted wavelet entropy features and deep features have undergone an adaptive fusion process using an improved meta-heuristic algorithm, thus termed adaptive wavelet entropy deep feature fusion. Finally, the classification is performed through I-RNN to get the disease-related outcomes, where the weight of RNN is optimized using a new MVS-AVOA. Through the evaluation, the performance analysis of the proposed MVS-AVOA-RNN has 41.5% better than Naive Bayes, 26.8% better than SRU, 18.3% superior to LSTM, and 5.4% enriched than RNN. Thus, the obtained result reveals that the proposed optimized RNN with an advanced feature set supersedes the aforementioned techniques. [ABSTRACT FROM AUTHOR] |
Full text from SpringerLink