Table 1.
Analysis of Literature Survey
| Reference Number | Method Name | Contribution | Merits | Demerits |
|---|---|---|---|---|
| 1. | DL-based feature detection | DL-based feature detection was proposed to detect the malware and the application’s behavioral analysis was done using different classifiers. | Malware Detection Accuracy was higher | Execution time was higher |
| 2. | Partitioned deep CNN | A partitioned deep CNN was proposed to learn the electrocardiogram features by using the classification of the electrocardiogram signal. | Better precision and sensitivity were attained | Compromising the incurring overhead during the healthcare analysis |
| 3. | DL-based Internet of Health Framework named, DeTrAs | DeTrAs were proposed to ensure the customized service via three phases. Sensory moment data were obtained using a recurrent neural network, followed by an ensemble technique used for tracking abnormality based on CNN, and the timestamp model` | Accuracy was said to be ensured | Early prediction remained unaddressed |
| 4. | Faster R-CNN | Faster-RCNN was designed for pandemic disease prediction. | Early detection was ensured with higher detection accuracy | Compromising the overhead incurred during the healthcare analysis |
| 5. | COVID-19 diagnostic techniques | COVID-19 diagnostic techniques were tested with pertinent adverse samples based on DL algorithms. | An efficient performance was attained. | Computational overhead was higher. |
| 6. | IoT-enabled technologies | The advantage of IoT-enabled technologies was used for energy saving, to smoothen the association between human and smart healthcare systems to a significant feasible magnitude. | Better precision was attained | The error rate was higher. |
| 7. | IoMT with Product Lifecycle Management (PLM) | IoMT with PLM introduces for regulating information transfer between devices in an accurate manner. | Better sensitivity was attained | Computational overhead was higher. |
| 8. | An IoT-based framework | The framework was introduced to identify the coronavirus suspects in the early stage. | Accuracy was said to be ensured. | Detection time was higher |
| 9. | Multi-Task Gaussian Process (MTGP) model | MTGP is used for efficient prediction of the COVID-19 epidemic. | Minimizing the overall impact of infection | Prediction accuracy was minimal. |
| 10. | Generalize Approximate Reasoning-based Intelligence Control (GARLIC) | GARLIC acquires information about the patient from IoT devices by using regression rules. | Early disease identification was attained. | The error rate was higher. |
| 11. | GFB-CNN | GFB-CNN uses Grey Filter Bayesian Convolution Neural Network in real-time analysis to address the qualities of services. | Time and overhead were addressed | The error rate was higher |
| 12. | Cognitive healthcare framework | A cognitive healthcare framework was designed with healthcare smart devices and sent to a cognitive model for further processing. DL was applied for decision-making. | Higher Accuracy was attained. | Computational overhead was higher. |
| 13. | Automatic heart disease analysis | The ensemble DL models are used to perform the automatic heart disease analysis. | The prediction accuracy was improved | Prediction time was higher. |
| 14. | Smart healthcare monitoring system | A smart healthcare system was introduced to observe a patient’s critical health signs and room information. | The error rate was minimal. | However, computational overhead was higher. |
| 15. | Smart healthcare monitoring model | A smart healthcare monitoring model was presented to identify the priority patient treatments with the aid of a smart healthcare model. | Computation overhead was reduced | Disease identification accuracy was lower |
| 16. | Emerging IoT methods | Emerging IoT methods in smart healthcare were discussed to analyze 9561 articles in IoT smart health | Emerging IoT methods supported panoramic knowledge. | The classification rate was not attained |
| 17. | Secure remote health monitoring model | A secure remote health monitoring model was introduced to discover the disease’s diagnosis at an early stage. | Secure data communication was achieved by using the designed framework | Data security level was not higher |
| 18. | Security and privacy problems | Security and privacy problems were discussed with five technical aspects. | Transmission efficiency was improved | Associated standards and technical specifications need to enhance in healthcare. |
| 19. | Hybrid architecture | The architecture was designed for IoT healthcare for analyzing the fundus images process | Image quality was improved | A hybrid architecture was not focused. |
| 20. | DL IoHT-driven technique | IoT framework was discussed to identify cervical cancer. | Accuracy was improved | Diagnosis of critical diseases was not performed |
| 21. | DL-based techniques | DL-based techniques were introduced for BTC | Performance was enhanced | Computational overhead was not reduced |
| 22. | ML technique named DLMNN | DLMNN was developed to identify the HD | The security level was higher and time was lower | The error rate was not minimized |
| 23. | Enhanced DL-assisted CNN | CNN method was introduced to enhance the patient’s prognostics of heart disease | Patient accuracy and reliability were increased | Precision was not enhanced by using advanced artificial intelligence. |
| 24. | Smart healthcare monitoring framework | A smart healthcare monitoring framework was introduced by using ensemble DL and feature fusion techniques to improve heart disease prediction accuracy. | Classification performance was attained | Processing time was higher |