HIMSS: Medical imaging moves to the cloud
LAS VEGAS—Cloud-based image management can help deliver improved real-time access to medical images anywhere and on any device, Andrew J. Lee, PhD, chief scientist, Harris Healthcare Solutions in Melbourne, Fla., said Feb. 22 at the annual meeting of the Healthcare Information and Management Systems Society (HIMSS). 

Healthcare has not kept pace with consumer developments in imaging and information sharing. Patient images are still shared on film and CD, which results in challenges for patients and physicians, Lee said. These include:
  • Radiologists may diagnose patients with incomplete information because they cannot access all images;
  • Redundant imaging exams;
  • Physician can’t access images with the rest of medical record;
  • Patients are unnecessarily transferred from one emergency department to another;
  • Manual handling issues impede efficiency and safety; and
  • Patients can’t access or control images.
James F. Philbin, PhD, co-director of the center for biomedical & imaging informatics at Johns Hopkins Medicine in Baltimore, detailed image storage challenges at his facility. The average size of a CT exam has reached 1,000 slices, and average image size is 50 megabytes, he said. In addition, 30 percent of all hard disk storage is used for medical images and data are growing 30 percent annually.

IT is challenged by the increasing numbers and sizes of studies; increasing locations of enterprises; the need to link disparate IT systems and infrastructures; budget constraints and understaffed, undertrained and under-resourced departments.

Johns Hopkins envisions a new paradigm that aggregates images across enterprises and shares them at the point of care. The model employs cloud technology to connect locations securely and with zero footprint viewing, Lee said.

Benefits of the cloud include reduced costs; improved scalability; fewer technology worries; improved reliability and serviceability; high availability; automated data replication and mobility via remote visualization. However, Lee noted there are technical challenges, which include cloud or network outage, performance and security.

An array of technologies is addressing these challenges, Philbin said. These include the internet, virtualization, remote visualization, identity management, security and storage advances.

Virtualization and remote visualization are examples of enabling improvements, Philbin said. Virtualization not only lowers the total cost of ownership but also has dropped delivery time for new virtual clients from weeks to minutes. Remote visualization leverages rendering engines in the data center and then sends images to a zero or thin client or virtual desktop client.

Lee concluded with several lessons learned from the cloud-based imaging project at Johns Hopkins. “Zero-footprint clients are fast and efficient. Start early on the master person index and get the ADT feed working early. Networking is key and must be reliable.”