Whole Slide Image to DICOM Conversion as Event-Driven Cloud Infrastructure

Abstract

The Digital Imaging and Communication in Medicine (DICOM) specification is increasingly being adopted in digital pathology to promote data standardization and interoperability. Efficient conversion of proprietary file formats into the DICOM standard format is a key requirement for institutional adoption of DICOM, necessary to ensure compatibility with existing scanners, microscopes, and data archives. Here, we present a cloud computing architecture for DICOM conversion, leveraging an event-driven microservices framework hosted in a serverless computing environment in Google Cloud to enable efficient DICOM conversion at scales ranging from individual images to institutional-scale datasets. In our experiments, employing a microservices-based approach substantially reduced runtime to process a batch of images relative to parallel and serial processing. This work demonstrates the importance of designing scalable systems for enabling enterprise-level adoption of digital pathology workflows, and provides a blueprint for using a microservice architecture to enable efficient DICOM conversion.