Hadoop-Based Intelligent Medical System and Method Thereof

Intelligent Telemedicine System Based on Hadoop MapReduce Parallel Analysis of Wearable Sensor Data

This technology is a medical system where body-worn sensors and coordinator sensors bundle blood glucose, blood pressure, pulse, body temperature, heart rate, and respiratory rate into a single packet for transmission. An intelligent building system then organizes this data into user-specific sequence files for parallel analysis using the Hadoop Distributed File System and MapReduce.

The proliferation of IoT and Wireless Body Area Networks (WBAN) has led to a surge in biometric data from connected medical devices, which existing processing structures struggle to store and analyze in real time. There has also been a lack of healthcare-specific big data architectures capable of handling both real-time and offline data to rapidly identify patient health anomalies.

This technology utilizes a master node and multiple parallel data nodes to process patient data divided by device identifier ranges, generating intermediate results for decision-making when readings fall outside normal ranges. This enables efficient processing of massive sensor data while providing feedback to users anytime, anywhere. Applicable to remote patient monitoring, nursing facility management, and chronic disease management platforms, it allows for the prompt detection of critical signs in multiple patients.

‍

‍

Key Features:

• A data collection unit consisting of multiple body-worn sensors, coordinator sensors, and basic medical devices to collect medical and activity information
• A coordinator sensor that adds a user ID packet header to sensor readings and transmits multiple readings, such as blood glucose, blood pressure, and pulse, as a single packet
• A collection unit that extracts global unique device identifiers, user IDs, and sensor measurements to generate user-specific sequence files
• A Hadoop processing unit where a master node and parallel data nodes analyze health readings via MapReduce and generate intermediate results when values deviate from the normal range

‍

This invention was developed with support from the Small and Medium Business Administration for resolving technical difficulties in large-capacity image analysis for film inspection equipment.

‍