Restoring private autism dataset from sanitized database using an optimized key produced from enhanced combined...

Area of Science:

• Computer Science
• Medical Informatics
• Data Security

Background:

• Timely identification of autism spectrum disorder (ASD) is vital for child development.
• Sharing sensitive autism data for diagnosis raises significant security and privacy concerns.
• Existing anonymization methods struggle with accurate data restoration and preventing leakage.

Purpose of the Study:

• To present a novel approach for improved data restoration of sanitized sensitive autism datasets.
• To enhance the security and privacy of autism data during transmission and storage.
• To address the deficiencies in accuracy associated with conventional data restoration processes.

Main Methods:

• Utilized an optimal key generated by the Enhanced Combined Particle Swarm Optimization-Grey Wolf Optimizer (PSO-GWO) framework.
• Applied the generated key for both sanitization (concealing data) and restoration (recovering data).
• Employed the same optimal key to improve the accuracy of original data recovery from sanitized datasets.

Main Results:

• Achieved highly competitive accuracies in autism data restoration experiments, reaching up to 99.90%.
• Demonstrated superior performance compared to existing meta-heuristic algorithms across various datasets.
• Outperformed other methods specifically on the 30-month autism children dataset.

Conclusions:

• The proposed method significantly enhances the security and privacy of autism data restoration.
• The Enhanced Combined PSO-GWO framework effectively generates optimal keys for robust data protection.
• This approach offers a promising solution for accurate and secure handling of sensitive autism-related information.