Getting a Handle on Medical Images

Enterprise Storage Forum content and product recommendations are editorially independent. We may make money when you click on links to our partners. Learn More.

Data is king at Bartron Medical Imaging, and Kazeon’s Information Server is helping to lower storage costs and ease management while allowing users to search and act on information generated by 2-D medical image processing.

Before settling on Kazeon, Bartron Medical, which develops advanced image analysis and data mining for biomedical applications, had searched for a year for a product that could organize and search image data generated by its 2-D and future 3-D Medical Segmentation System (Med-Seg).

Today, Bartron Medical is using the Kazeon IS1200-ECS system in conjunction with a storage solution from Network Appliance to store and search medical imaging data.

“We needed an off-the-shelf tool that could handle the searches that need to take place on the images and in medical records,” said Fitz Walker, president and CEO at seven-year old Bartron Medical, located in Largo, Maryland.

Using NASA-pioneered technology that is licensed and adapted by Bartron Medical, the company’s Med-Seg system goes beyond current medical imaging technologies such as CT scans and MRIs to look deeper into the images and bring out properties the eye can’t see.

“We believe that this technology can be applied to clinical problems and will help lead to early detection and early treatment,” said Walker.

Med-Seg is currently in the latter stages of FDA approval, with commercialization to follow.

Managing 10GB Images

Bartron Medical’s Med-Seg imaging device utilizes the two-dimensional version of NASA’s Goddard Space Flight Center’s Recursive Hierarchical Segmentation Software (RHSEG), which the company licensed in 2002. Med-Seg was designed to analyze digital X-rays, soft tissue scans, mammograms, ultrasounds, MRI images and CT scans for the diagnosis and management of diseases.

RHSEG partitions an image into related sections or an entire region formed by aggregated feature values. Powered by parallel processing computer clusters, clinicians are provided with rapid, sensitive and precise analysis, as each image pixel is treated separately. The result, according to Walker, is an accurate graphical representation of the imagery data with fine resolution of detail and minimal distortion.

Processing a single 2-D image generates more than 10 gigabytes of data that can be used for data analysis. Multiplied by more than 1,000 images, the amount of storage required grows rapidly.

Bartron is also working on a new version of Med-Seg that will include 3-D processing. In 2005, the company signed a cooperative research and development agreement with NASA to enable the joint development of the 3-D version on Goddard Space Flight Center’s RHSEG.

“We’re beginning to work on how to analyze 3-D segmented data off of an MRI, which generates incredible amounts of data,” said Walker. He said that a single 3-D slice may generate 15 MB of data. That multiplied by 128 slices per image results in huge quantities of image data.

Success of the Med-Seg system is heavily dependent on finding a solution for storing and searching the data. Today, Bartron has a computing cluster and Kazeon device in its R&D lab in Connecticut and recently installed a redundant system at its incubator site in Maryland.

When Bartron began working on the 2-D Med-Seg system, the company manually put data in a folder using server-attached storage. “We processed over 5 million segmentations,” said Walker.

Then the search for a storage and search solution began. A team at Bartron explored options that had to meet a set of criteria. The box had to:

  • Hold up under production and provide longevity;
  • Minimize power consumption;
  • Provide fast access time to the drives;
  • Allow for virtual drives; and
  • Interface to the company’s Linux platform.

At the end of the day, Walker brought in a Kazeon/NetApp solution. “We worked with Kazeon, who came in and installed the system,” he said.

The company’s tech team then put the system through the ropes, testing it for performance and durability. A proof of concept test included running the system with 4 TB of data, then 160 TB of data, and finally, 240 TB of data.

“We expect that the amount of data the system will process will go even higher once we get units in the field,” said Walker. After just one year, Bartron estimates that its solution will grow to stream 10,000 images per week.

Kazeon’s IS1200-ECS features include discovery, classification, searching, reporting and action. The system, according to Kazeon, is capable of indexing terabytes of information representing hundreds of millions of files and e-mails and culling the large data sets down for review, secure storage or litigation hold.

Not only did the solution meet the company’s criteria, but according to Walker, it was also cost effective.

Up And Running

The way Med-Seg works is to draw images from medical centers that install a server at their facility. The image data is collected in the server and then sent to Bartron Medical, where the data is processed by the company’s imaging application. The data is then sent back to the medical center for diagnosis. Customers are offered the option of storing the image files long-term at Bartron’s data center.

Bartron expects to have a minimum of 100 units in the field once the Med-Seg System receives FDA approval. System cost is about $100 million, or $50 million for the average cluster, not including data storage, according to Walker.

Today, Bartron has systems installed at hospitals such as the University of Connecticut Health Center, New York University Medical Center, Yale-New Haven Medical Center and the University of Maryland Medical Center.

Solution benefits include improved customer service to hospitals; improved productivity through easy access to files; reduced storage costs by 30 percent by archiving infrequently accessed files to optical media; and management improvement of up to three hours per week.

To date, the system has passed proof of concept with no problems. Walker expects to work with a technical team from Kazeon to meet the company’s objectives for a commercial product.

Back To Enterprise Storage Forum

Lynn Haber
Lynn Haber
Lynn works as an editor-at-large meeting with new and established channel partners in the IT and telecommunications ecosphere. She works to understand what motivates partners, vendors, and distributors as they navigate growth in a dynamic industry. Lynn has worked as a business and technology writer for magazines, journals, and online media, and has authored hundreds of articles on technology, communications, business management, employment and careers, industry, and products.

Get the Free Newsletter!

Subscribe to Cloud Insider for top news, trends, and analysis.

Latest Articles

15 Software Defined Storage Best Practices

Software Defined Storage (SDS) enables the use of commodity storage hardware. Learn 15 best practices for SDS implementation.

What is Fibre Channel over Ethernet (FCoE)?

Fibre Channel Over Ethernet (FCoE) is the encapsulation and transmission of Fibre Channel (FC) frames over enhanced Ethernet networks, combining the advantages of Ethernet...

9 Types of Computer Memory Defined (With Use Cases)

Computer memory is a term for all of the types of data storage technology that a computer may use. Learn more about the X types of computer memory.