A Secure Parallel Pattern Mining System for Medical Internet of Things.

Area of Science:

• Computer Science
• Data Science
• Medical Informatics

Background:

• The Internet of Medical Things (IoMT) generates vast amounts of data, posing significant challenges for effective big data exploration and pattern discovery.
• Existing methods struggle to efficiently analyze complex medical datasets while ensuring data privacy and security.

Purpose of the Study:

• To develop a novel, generic parallel pattern mining framework, MD-PPM, for big data exploration in the IoMT.
• To address the challenges of connectivity analysis and pattern discovery in large-scale medical data.
• To integrate robust privacy and security measures using blockchain technology.

Main Methods:

• Developed a multi-objective Decomposition for Parallel Pattern-Mining (MD-PPM) framework.
• Utilized a novel multi-objective k-means algorithm for medical data aggregation.
• Employed parallel pattern mining techniques leveraging GPU and MapReduce architectures.
• Integrated blockchain technology for enhanced data privacy and security.

Main Results:

• MD-PPM demonstrated high performance in terms of memory usage and computation time for sequential and graph pattern mining on large medical datasets.
• The framework achieved strong efficiency, accuracy, and feasibility compared to existing models.
• Successful integration of blockchain technology ensured complete privacy and security of sensitive medical data.

Conclusions:

• The proposed MD-PPM framework offers an efficient and accurate solution for big data exploration in the Internet of Medical Things.
• MD-PPM effectively addresses the challenges of pattern discovery and data privacy in complex medical data environments.
• The framework's performance and security features make it a valuable tool for advancing medical informatics and IoMT applications.