title:: Transforming radiology workflows with clinical decision support powered by AWS
publisher:: AWS
organization:: ContextFlow, Amazon Web Services (AWS), Medical University of Vienna, Vienna General Hospital
domain:: Healthcare, Radiology, Medical Imaging, Data Analytics, Machine Learning, Cloud Computing
Info
Summary
This blog post discusses how the healthcare technology startup contextflow leverages Amazon Web Services (AWS) to power its clinical decision support system, SEARCH Lung CT. The software helps radiologists diagnose patients faster and more accurately by searching and retrieving historical diagnostic data to guide report writing. Key points:
- Radiology faces increasing workloads, data complexity, and radiologist shortage globally
- contextflow’s SEARCH Lung CT analyzes lung CT scans for interstitial lung diseases, COPD, and lung cancer screening
- Uses deep learning models to identify 19 disease patterns, provide measurements, and link to diagnosis information
- Searches a custom knowledge base of 22,600 lung CTs from 11,600 patients for similar cases to the current scan
- Leverages AWS services like Amazon EFS for data storage/sharing and Amazon EC2 for compute scalability
- In a study, reduced radiology report reading time by 31% and improved diagnostic accuracy when using the tool
Data Points
- Radiologists reviewed 15,000 images/day in 2010, tripled by 2018
- UK has shortage of 1,939 radiology consultants, costing £206 million/year for outsourcing
- contextflow founded in 2016 based on KHRESMOI European research project
- SEARCH Lung CT provides quantitative info like % lung volume affected by anomalies
- Indexes and processes data using AWS services like Amazon EFS and Amazon EC2
- Study at Medical University of Vienna and Vienna General Hospital showed benefits
- Presented results at European Congress of Radiology 2022 and upcoming publications