Yarra Framework

Duration:2014 - current
Technologies:C++, Qt, Wt, Flask, SQLite
Collaborators:Roy Wiggins

The Yarra Framework is a collection of open-source software tools developed to simplify the process of developing, evaluating, and translating novel magnetic resonance imaging (MRI) reconstruction techniques into clinical practice. It is provided free-of-charge for non-commercial purpose under the GPLv3 license. The tools can be used synergistically together, or they can be used individually as stand-alone components.

The “Yarra Raw-Data Service” (RDS) offers automatic collection of raw k-space data. It runs as background process on the host computers of Siemens MRI scanners and performs a nightly raw-data export of requested scan protocols into project folders on a network share or USB drive. This enables creating extensive libraries of raw data from routine exams without the need to access the scanners during clinical work hours. For example, it has been used to generate the public fastMRI datasets released jointly by NYU & Facebook, which provided a foundation for the development of the first AI-based MRI reconstruction techniques.

“Yarra Archive Search” (YAS) can be used to index large amounts of collected cases and provides a convenient web interface that allows searching for raw-data files by project, sequence, or patient information.

“YarraServer” can be used to automate offline reconstruction of scans on external Linux servers. It enables integration of reconstruction prototypes into the clinical workflow with minimal effort. Cases can be easily submitted from MRI scanners using the “Yarra Offline-Reconstruction Task” (ORT) client, and from workstations using the “Yarra Stand-Alone Client” (SAC) or the “Yarra WebClient”. Reconstruction prototypes can be implemented using any programming language running under Linux, ranging from Matlab over Python to C++, or using frameworks like BART, Gadgetron, and PyTorch. Yarra servers can be controlled and monitored using the “Yarra WebGUI”, which includes an interface for configuring reconstruction settings. It also provides a packaging system for simple installation of additional processing modules. Yarra has an integrated load-balancing and routing mechanism for distributing jobs across multiple servers. In addition, the “Yarra Router” software can be used for more advanced load balancing when using a cluster of servers. To ensure reliability, Yarra has an email notification system and can be monitored with the “Yarra LogServer” software. The status of reconstruction jobs can furthermore be queried with the “YarraTech” web interface, which aggregates status data from all servers and summarizes the information for clinical staff.

This software infrastructure has enabled to apply the GRASP DCE-MRI technique in over 200,000 clinical patient exams to date.

As additional functionality, the Yarra Framework can also be used to automatically collect clinical workflow data through the “Yarra LogServer” extension. Collected workflow data can be analyzed with common Business Intelligence (BI) software and utilized to make informed decisions for improving the clinical efficiency and examination quality.

More information