Matthew T. Latourette, Michigan State University; James E. Siebert; Michael J. Potchen, MD; Kenneth L. Marable; Robert J. Barto; Colleen A. Hammond, RT(R); Anthony Muyepa, MS

Malaria causes the deaths of 3-4 million children per year and threatens half of the world’s population.[1] As part of a new NIH-funded malaria study,[2] the program establishes a magnetic resonance imaging research facility located within a region suffering maximal incidence of pediatric cerebral malaria (CM). Michigan State University (MSU) collaborates with Queen Elizabeth Central Hospital (QECH) and University of Malawi College of Medicine (UNIMA) in Blantyre, Malawi to apply neurological MR methods in basic and clinical CM research. The research program requires daily transmission of studies Malawi MR from to MSU for quantitative post processing, and clinical research interpretation. Other supporting IT functions (i.e., clinical information systems operation, research data collection, IP telephony and conferencing, email) are also required. The under-developed location presents extraordinary challenges to the project implementation and execution.

This presentation will describe the implementation of DICOM networking over an intercontinental satellite-based Internet service, various technical challenges, achieved performance metrics including end-to-end DICOM image transmission rates, and IT related matters. There is scant information in the literature or Internet content concerning satellite-based DICOM networking, which is characterized by its intrinsic long latency.

Queen Elizabeth Central Hospital was selected as the research site because of its location in a P. falciparum-endemic area and because of its long-standing relationship with MSU. A permanent magnet based GE Signa Ovation 0.35T MR scanner was selected for the project. A free-standing building housing the MR scanner was constructed on hospital grounds.

Intercontinental Satellite-Based Network

Figure 1 depicts the implementation of the intercontinental satellite-based data networking between MSU and QECH. Connectivity within the MRI facility was implemented using a switched Ethernet local area network (LAN). Imaging studies performed on the MR scanner are archived locally to recordable compact disc media for long term storage and transmitted to the DICOM router, which caches the data and forwards it to a stand-alone DICOM image viewing workstation for primary diagnostic reading.

Malawi has no fiber connection to the Internet. Remote Internet connectivity was established via C-band wireless relay from a Very Small Aperture Terminal (VSAT) satellite ground station through a geosynchronous satellite to the Internet service provider’s (ISP) Cambridge, United Kingdom facility. The Service Level Agreement with the satellite ISP specifies a minimum of 98.5% uptime. A reciprocal agreement with UNIMA provides limited network redundancy by sharing an existing UNIMA-owned satellite. A bandwidth manager server facilitates failover bandwidth-sharing during a downtime scenario and ensures timely delivery of MRI data by configurable prioritization of network traffic types. A hybrid router/firewall provides network address translation and DHCP services for the LAN and excludes unauthorized traffic to protect the LAN from malicious agents. A satellite router performs sophisticated encoding and modulation/demodulation functions for efficient data transport and provides important network protocol acceleration strategies to minimize apparent link latency.

Automated encrypted end-to-end transmission of imaging studies is managed by the DICOM routers deployed at both the MRI facility and MSU. The DICOM routers provide caching and queuing of studies, handle transmission errors robustly, and optimize utilization of the available 384 kbps single channel per carrier (SCPC) satellite bandwidth. At MSU, the Research PACS supports image post-processing and secondary/research interpretation of research studies.

Financial considerations limited the satellite link speed that was contracted. Symmetrical 384 kbps service costing ~$3,600 per month was used. Preliminary testing of the link speed using ftp showed data transport between the UK and Malawi running between 33 KB/s and 22 KB/s, ~69% of maximum channel capacity. Detailed metrics of DICOM image transport speed between Malawi and MSU will be reported.

Clinical Study Interpretation and Research Data Collection IS

A custom web-based software titled NeuroInterp was developed to record neurologic MR study findings in a structured manner and generate clinical MR to study interpretation reports. A large hierarchical classification scheme was developed to capture findings by iteratively specifying additional levels of detail through a graphical user interface wizard. NeuroInterp also supports a triple-reader discrepant findings resolution process, basic patient registration, and technical note-taking functionalities.

The better-than-anticipated performance of the satellite Internet service allowed for the implementation of the NeuroInterp and registration software using ASP.NET 2.0 and Microsoft SQL Server Express. This approach significantly improved functionality, error handling, and overall software stability over the originally planned architecture of local Active Server Pages and Microsoft Access. NeuroInterp presently comprises 234 captured data elements across 17 screen pages.

After installation on-site, connection speed and quality were sufficient to use web-based remote administration software, deployed on all PCs at the site, to allow some configuration and maintenance tasks to be performed from MSU. UNIMA provides IT support at the Malawi site.

Successfully establishing a collaborative research MR center in an underdeveloped country requires surmounting logistical, political, and regulatory complications; limited local technical resources; and human research protection challenges. NIH proposal reviewers doubted success.

Configuration of the PCs required at the Malawi site was done prospectively at MSU. In order to minimize the complexity and time delays of technology export/import compliance, PCs were packed in their original containers and shipped by air freight to the site one month in advance. To speed clearance through customs and avoid paying duty taxes, each box contained a copy of the original purchase invoice, an excerpt of the grant application detailing the purpose of the equipment, a customs declaration form, and an executed letter of donation.

MSU technical personnel traveled to the site to coordinate the project launch. Satellite installation was partially completed but stalled, waiting on the arrival of parts. DICOM connectivity was established and tested between the modality, viewing workstation, and DICOM router using a temporary network configuration. After completion of the satellite Internet service, the availability of skilled IT staff at UNIMA was a crucial factor in successfully completing configuration of the facility network.

In the developing world, the whole job is subcontracted to access labor and resources in the locale of the job site because of inefficiencies associated with travel. More than one level of subcontracting is common. This practice resolves travel problems, but complicates project management. Email communication was critically important as a coordination tool, especially for communications with staff members at MSU. The expense of international calls and synchronization issues associated with time zone differences make telephone communication undesirable.

This project established an MRI center to support pediatric cerebral malaria research in sub-Saharan Africa. A satellite networking infrastructure was implemented to achieve practical intercontinental transport of DICOM image data. Custom project software was developed for structured generation of clinical neuroimaging reports and research data collection.

NIH/NIAID Grant 2 R01 AI034969-10A1 provided the principal funding for this project.

Dr. Terrie E. Taylor (PI) provided much oversight for the project aspects reported herein.

Support was provided in part by GE Healthcare.

1. SEAR malaria incidence. WHO, 1999.
2. NIH/NIAID Grant 2 R01 AI034969-10A1