Medical and Healthcare

Evolving at a Rapid Rate

The medical and healthcare industries are evolving at a rapid rate. This change is being fostered not only by rapid technological advancements in the areas of digital imaging, laboratory and clinical systems, patient record management, and so on, but also by the development and evolution of the many industry standards (e.g. DICOM, HL7) and by regulatory forces (HIPAA).

A common underpinning to all of this is the management of collected data at the "micro" level in individual systems or hospital units, and potentially at the "macro" level as patient data is aggregated and managed across a hospital for diagnostic purposes.

Data Collection

There are many different units within a hospital where patient diagnostic data is collected. These include laboratory tests, digital imaging, and bedside monitoring. In each case, specialized hardware is responsible for collecting the data, but software systems are responsible for capturing, storing, analyzing and managing the data. It is in the storage of this data that Versant can play a significant role.

In each case, there is a fair amount of complexity that is represented in the data that is collected. Sequences of tests must be linked together in terms of time or sample. Related images must be linked together. Clinical systems (bedside monitoring) data must be stored in terms of time within a particular waveform, as well as relative to other waveforms that may be collected simultaneously. In each case, data must be linked to the individual patient, and perhaps in relation to each hospital stay.

Data of this complexity requires capabilities that are not typically available in a traditional data store. The Versant object database is such a product. Versant is a sixth generation database system combining the ability to directly manage the complex, graph structured data inherent in modern medical applications with direct access from today's most prevalent programming languages. From its inception, Versant has been architected to achieve the highest levels of performance and 24x7 reliability in concurrent, highly distributed environments. As a result, the Versant object database gives medical equipment manufacturers an off-the-shelf tool to rapidly and efficiently solve problems that have to date been difficult, or sometimes even impossible, to solve using older storage management technologies.

Example Applications

The following illustration shows a bedside monitoring application built around Versant. In this application, the patient monitors collect vital sign data and forward it to a data acquisition application. This application receives the data, displays it o?n a real-time monitor at a nurses' station, and stores the data o?n a patient-by-patient basis in Versant, keeping it for a pre-determined amount of time. A remote version of the monitoring application can retrieve data from Versant in near real-time as well as provide the ability to drill down into previously collected data.

In a similar manner, Versant can be used to manage diagnostic imaging data. The following illustration shows how this might be architected.

In this case, images from various modalities are acquired and forwarded to a Data Acquisition application. The Data Acquisition application stores metadata associated with the images, as well as thumbnail versions of the images if appropriate. Note that while it is possible to store the images themselves in Versant, developers typically do not choose to do so because there is little benefit: images are large, static, and not capable of being queried directly.

The Image Data Server performs two basic functions. o?ne is to distribute images to viewing stations that request them. The other is to manage the repository, migrating older images from disk to tertiary media such as DLT (digital linear tape) for archival purposes and then updating Versant with the new location of the image. If a Viewing Station requests an image that has been previously archived, the Image Data Server manages retrieval of the image from DLT back to disk, updates Versant, and then forwards the image to the Viewing Station as it normally would.

This final example shows how Versant might be used in a hospital setting.

In this case, Versant's enJin™ product is used to provide persistence of data within a commercial application server. The power of this architecture comes from Versant's ability to model and manage complex data. The complexity comes from the need to aggregate common data from various sources (i.e. patient data from hospital information systems, laboratory systems, etc.), and presenting a unified view of the data to requesting applications around the hospital. Versant acts as a repository of this aggregated data, providing a powerful object model around which new applications might be built. Versant provides an event notification mechanism (essentially triggers) that can be used to proactively initiate processing when certain events - perhaps completion of a lab test - occur.

The Bottom Line

The bottom line is that Versant provides a robust, industrial strength object data management solution that allows developers to solve complex problems in ways they have never before been able. Versant stores objects natively as objects, allowing the application to take full advantage of the power of object-oriented development. It provides the ability to manage complex graphs of data and the ability to transparently navigate from one object to another - regardless of whether the object has been previously retrieved into memory from the database. Another advantage of using Versant is the fact that it supports deployment in low/no maintenance environments. When all of Versant's features and capabilities are taken into account, it is clear that Versant is extremely well suited to development and deployment in medical applications and hospital environments.

The following are some of the industry leaders that have selected Versant as the data management component for their sophisticated analytical applications and equipment:

• Amgen
• Applied Biosystems
• Beckman Coulter
• Becton Dickinson
• bioMerieux
• bionomics
• MDS Sciex
• Media Cybernetics
• Mount Sinai School of Medicine, Institute for Computational Biomedicine
• Siemens
• Tecan
• UCLA Multimedia Stream System Interdisciplinary Research Group
• Weill College of Medicine at Cornell University, Institute for Computational Biomedicine