Transforming Healthcare Tech: Intelligent Document Processing with GenAI
Overview
Our client is a healthcare technology company that partners with hospitals and anesthesia practices to optimize operating room performance and maximize revenue. Its cloud-based solutions improve data collection, enhance clinical workflows, and provide powerful analytics to drive better decision-making, thus empowering healthcare teams to deliver superior patient experiences. With a mission to streamline administrative tasks, they empower healthcare teams to focus on delivering better care by offering solutions that reduce complexity and cost.
The company manages a large volume of medical records in various PDF formats, particularly from Epic systems. This posed challenges in extracting, enriching, and transforming the data into a standardized format. The goal of the project was to facilitate efficient reporting and analytics, ensuring that the standardized data could be seamlessly integrated and used across the client's systems.
Challenge
The company needed a solution to ingest Epic records from multiple hospital systems, private practices, and other healthcare providers within the Electronic Medical Record (EMR) system. These records came in varying formats, with inconsistencies in the type and amount of information included. Some were brief, while others spanned hundreds of pages, making manual review time-consuming and inefficient.
To address these challenges, the client required a scalable and automated solution for accurate data review. This led them to partner with JBS Dev for a GenAI solution to improve and automate the entire process.
The solution needed to support PDF records with varying formats and structures, identify and extract relevant data, apply the correct medical coding, and deliver the data in a standardized format for use by external systems.
Solution
JBS Dev developed an AI-powered solution to streamline medical document processing, leveraging the power of AWS services. The solution utilizes Amazon Bedrock hosted Large Language Models (LLMs) to extract information from the parsed PDFs and Epic formatted documents, intelligently handling various formatting types and standardizing the results into JSON format. Also, AWS Lambda was used to automate document processing workflows, triggering tasks, and storing output data in Amazon S3 for further analysis.
- These LLMs, accessed through Amazon Bedrock, processed the data to identify and standardize key information, such as medication names, dosages, administration methods (e.g., orally or via IV), and physician notes on drug usage and monitoring timelines
- Both the extracted PDF data and the standardized JSON files were securely stored in Amazon S3, ensuring availability for downstream processing or integration with other systems. Operational details, such as physician and nurse names, were also captured and integrated with the hospital's systems for holistic reporting and decision-making.
- AWS Lambda enabled containerization and dynamic scaling of the document processing services, optimizing resource management and operational efficiency. AWS CloudWatch and CloudTrail provided monitoring and logging, ensuring visibility into the report generation process, and facilitating quick resolution of any data extraction or processing issues.
This AI solution significantly enhanced real-time document retrieval, improving medical decision-making during and after surgery, all while ensuring compliance with industry data privacy regulations to protect sensitive medical information.
Results
The GenAI-powered solution, utilizing AWS Lambda and Amazon Bedrock, significantly enhanced the client’s data processing and operational performance.
By integrating Amazon Bedrock for Large Language Models (LLMs) and AWS Lambda for automated workflows, the solution efficiently extracts and standardizes critical data from diverse medical records, improving interoperability and data consistency across systems.
The AI-driven performance analysis and reporting enabled faster, data-driven decisions for hospitals and clinical teams. Real-time insights, powered by AWS Lambda’s dynamic scaling, improved report accuracy and timeliness, enhancing care delivery.
Automating the data review and correction process with AWS Lambda reduced manual effort and streamlined workflows, speeding up document processing and enabling quicker, informed clinical decisions, improving patient care.
The structured data, processed via AWS Lambda, Amazon S3, and Amazon RDS, provided precise insights for optimized resource allocation and streamlined operations across hospitals and practices. The system’s ability to scale with document volume fluctuations improved efficiency, reduced errors, and ensured consistent processes, boosting overall operational performance.
Conclusion
By partnering with JBS Dev, this healthcare technology company resolved a critical bottleneck, providing greater operational efficiency and ensuring the timely delivery of accurate data to empower informed decision-making.
