Toward ubiquitous healthcare services with a novel efficient cloud platform

Area of Science:

• Computer Science
• Medical Informatics
• Cloud Computing

Background:

• Ubiquitous healthcare services are increasingly vital due to the global aging population.
• Cloud computing offers suitable characteristics for healthcare but faces challenges with multimodal, heterogeneous, and nonstationary physiological signals.
• Persistent personalized services and high-concurrency online analysis for public healthcare present significant hurdles for general cloud platforms.

Purpose of the Study:

• To propose a private cloud platform architecture specifically designed for the demands of ubiquitous healthcare services.
• To overcome the limitations of general cloud platforms in handling complex physiological data and concurrent user requests.

Main Methods:

• A six-layer private cloud platform architecture was developed, utilizing a message queue as the central engine.
• A loosely coupled communication model with a publish/subscribe mechanism ensures layer independence.
• A plug-in algorithm framework facilitates adaptive access to massive semi-structured and unstructured medical data.

Main Results:

• The proposed private cloud platform architecture demonstrated robust, stable, and efficient performance.
• The platform successfully satisfied high concurrent requests essential for ubiquitous healthcare services.
• Adaptive data access mechanisms effectively managed diverse medical data formats.

Conclusions:

• The developed private cloud platform architecture is well-suited for ubiquitous healthcare applications, particularly those addressing the needs of an aging population.
• The architecture effectively handles multimodal, nonstationary physiological signals and supports high-concurrency analysis.
• The platform's design ensures efficient and adaptive processing of large-scale medical data, enhancing personalized healthcare delivery.