Feature Extraction Approach for Speaker Verification to Support Healthcare System Using Blockchain Security for Data...

Area of Science:

• Biometric systems
• Artificial Intelligence in Healthcare
• Data Security and Privacy

Background:

• Speech analysis offers a unique biometric modality combining physiological and behavioral traits for remote transactions.
• Predicting human emotions from speech is complex, influenced by internal behavior and vocal characteristics.
• Existing healthcare systems face challenges in data privacy and security, especially with sensitive patient information.

Purpose of the Study:

• To develop a deep learning approach for emotion prediction from speech to enhance intelligent healthcare systems.
• To implement a Blockchain-based system for secure and privacy-preserving sharing of healthcare data.
• To analyze the effectiveness of feature removal techniques in optimizing emotion prediction accuracy.

Main Methods:

• Utilizing deep learning models for behavior prediction based on vocal track, voice flow, and frequency.
• Employing Blockchain technology to ensure secure and decentralized storage and distribution of healthcare data.
• Evaluating various feature removal strategies to identify optimal parameters for accuracy and error reduction.

Main Results:

• The proposed model demonstrated increased accuracy in evidence collection and minimized data loss and authentication issues.
• Feature extraction based on text validation was found to enhance the sustainability of the healthcare system.
• Different feature removal approaches showed varying performance based on voice specimen variations.

Conclusions:

• The integration of deep learning for emotion recognition and Blockchain for data security offers a robust solution for modern healthcare.
• The research highlights the potential of speech biometrics in creating more intelligent and secure healthcare platforms.
• The findings support the development of privacy-preserving, secure, and sustainable healthcare systems through advanced technological integration.