LONI Pipeline
LONI has created an intelligent and interactive distributed visual programming environment, the LONI Pipeline, for building and executing data analysis protocols utilizing existing processing and visualization tools and data sets.
The LONI Pipeline facilitates neuroscience research by providing an intelligent interface for defining scientifically and algorithmically correct protocols. The intelligent interface will be completely extensible and tunable for user- and task specific data processing. Furthermore it is being adapted for several other applications and projects. We recently published the results of using the LONI Pipeline to implement a Brain Extraction Meta-Algorithm (BEMA). The LONI Pipeline enabled us to execute several extraction algorithms and a registration procedure in parallel, and to then combine the results in an intelligent fashion to obtain improved results over any of the individual algorithms.
Taking advantage of this has allowed us to make other dramatic improvements (which are included in this v3 release). These new improvements include improved security, grid computing integration, automated process management, automated resource management, advanced debugging, and fault tolerance, and a new feature rich graphical user interface (GUI). Secure communication is now an integral part of the Pipeline. All messages between a server and a client are encrypted using an AES key, which is established via the Diffie-Hellman algorithm. This will aid researchers in staying compliant with HIPAA and local IRB rules.
Grid computing integration via Sun's Grid Engine allows labs with large but disparate computational resources to make full use of those resources.
Automated process management schedules jobs according to type, user priority, resource availability, etc. To facilitate the submission and execution of compute jobs in this heterogeneous compute environment, SUN's Grid Engine (SGE) is used to virtualize the resources above into a compute service. A grid layer sits atop the compute resources and submits jobs to available resources according to user-defined criteria such as CPU type, processor count, etc. The laboratory has successfully integrated the latest version of the LONI Pipeline with SGE using SUN's Java DRMAA bindings. The bindings allow jobs to be submitted natively from the LONI Pipeline to the grid without the need for external scripts. Furthermore, the LONI Pipeline can directly control the grid with DRMAA, significantly increasing the operating environment's versatility and efficacy, and improving overall end-user experience.
The Pipeline GUI has been redesigned in response to user feedback. Usability testing has provided a wealth of knowledge in usability design, and we have applied it toward this new version. These improvements include the highlighting of connections between selected modules to make them more visible, quick access to zooming and panning actions to navigate pipelines that do not easily fit into a user's window, and an ability to apply the same action to multiple modules simultaneously to prevent monotonous actions for users. The current release of the LONI Pipeline is available for use by researchers at the LONI Software site.
