A Novel Intelligent Hybrid Optimized Analytics and Streaming Engine for Medical Big Data

Area of Science:

• Health Informatics
• Artificial Intelligence
• Data Science

Background:

• The exponential growth of medical data from sensing devices necessitates advanced analytics for effective healthcare monitoring.
• Traditional machine learning models face challenges in processing large-scale, real-time medical data streams.
• There is a growing need for improved methods in healthcare big data analytics to extract insights and predict disease trends.

Purpose of the Study:

• To propose a novel deep learning framework for efficient processing and analysis of healthcare big data.
• To enhance medical data analytics and monitoring using a combination of electronic record search engine algorithm and optimized LSTM.
• To address the challenges in applying learning models to big/medical data streams.

Main Methods:

• Development of the electronic record search engine algorithm (ERSEA), a deep learning model.
• Integration of a firefly optimized long short-term memory (LSTM) model for enhanced data analysis.
• Experimentation using Apache Spark with diverse medical respiratory datasets.

Main Results:

• The proposed ERSEA and firefly-optimized LSTM model achieved high performance metrics.
• Accuracy, sensitivity, and specificity reached 94%, 93.5%, and 94% respectively for datasets < 5 GB.
• For datasets > 5 GB, the model demonstrated 94% accuracy, 92% sensitivity, and 93% specificity.

Conclusions:

• The proposed deep learning framework offers extraordinary performance in healthcare big data analytics and monitoring.
• The ERSEA combined with firefly-optimized LSTM effectively handles large volumes of real-time medical data.
• This approach represents a significant advancement in leveraging AI for medical data processing and disease prediction.