Brent J. Liu

Trends in PACS image storage and archive.

PACS is widely used in hospitals and is considered a mission critical system for around-the-clock daily clinical operation. Scheduled or unscheduled downtime of the main PACS archive storage or server could potentially cripple the entire PACS operation. This is especially the case in a filmless hospital environment. Therefore, in a downtime event, it is most desirable for users to have only a minimal performance impact without interruption of clinical data flow or loss of data and to have available historical PACS studies. This paper summarizes some of the developments in the design and implementation of a reliable PACS that insures maximum uptime for end users while preserving the integrity of the PACS data and making it available during downtime events. It also details strategy for developing proper clinical workflow contingency procedures when a scheduled downtime event to the main archive storage and server occurs. Specifically, the design and implementation of a fault-tolerant (FT) main archive server, the development of a FT back-up archive using an application service provider (ASP) model, and the clinical experiences while upgrading a main archive server and migrating the stored PACS data to new storage media will be discussed.

Fault-tolerant back-up archive using an ASP model for disaster recovery

A single point of failure in PACS during a disaster scenario is the main archive storage and server. When a major disaster occurs, it is possible to lose an entire hospitals PACS data. Few current PACS archives feature disaster recovery, but the design is limited at best. These drawbacks include the frequency with which the back-up is physically removed to an offsite facility, the operational costs associated to maintain the back-up, the ease-of-use to perform the backup consistently and efficiently, and the ease-of-use to perform the PACS image data recovery. This paper describes a novel approach towards a fault-tolerant solution for disaster recovery of short-term PACS image data using an Application Service Provider model for service. The ASP back-up archive provides instantaneous, automatic backup of acquired PACS image data and instantaneous recovery of stored PACS image data all at a low operational cost. A back-up archive server and RAID storage device is implemented offsite from the main PACS archive location. In the example of this particular hospital, it was determined that at least 2 months worth of PACS image exams were needed for back-up. Clinical data from a hospital PACS is sent to this ASP storage server in parallel to the exams being archived in the main server. A disaster scenario was simulated and the PACS exams were sent from the offsite ASP storage server back to the hospital PACS. Initially, connectivity between the main archive and the ASP storage server is established via a T-1 connection. In the future, other more cost-effective means of connectivity will be researched such as the Internet 2. A disaster scenario was initiated and the disaster recovery process using the ASP back-up archive server was success in repopulating the clinical PACS within a short period of time. The ASP back-up archive was able to recover two months of PACS image data for comparison studies with no complex operational procedures. Furthermore, no image data loss was encountered during the recovery.

Application development environment for advanced digital workstations

One remaining barrier to the clinical acceptance of electronic imaging and information systems is the difficulty in providing intuitive access to the information needed for a specific clinical task (such as reaching a diagnosis or tracking clinical progress). The purpose of this research was to create a development environment that enables the design and implementation of advanced digital imaging workstations. We used formal data and process modeling to identify the diagnostic and quantitative data that radiologists use and the tasks that they typically perform to make clinical decisions. We studied a diverse range of radiology applications, including diagnostic neuroradiology in an academic medical center, pediatric radiology in a childrens hospital, screening mammography in a breast cancer center, and thoracic radiology consultation for an oncology clinic. We used object- oriented analysis to develop software toolkits that enable a programmer to rapidly implement applications that closely match clinical tasks. The toolkits support browsing patient information, integrating patient images and reports, manipulating images, and making quantitative measurements on images. Collectively, we refer to these toolkits as the UCLA Digital ViewBox toolkit (ViewBox/Tk). We used the ViewBox/Tk to rapidly prototype and develop a number of diverse medical imaging applications. Our task-based toolkit approach enabled rapid and iterative prototyping of workstations that matched clinical tasks. The toolkit functionality and performance provided a hands-on feeling for manipulating images, and for accessing textual information and reports. The toolkits directly support a new concept for protocol based-reading of diagnostic studies. The design supports the implementation of network-based application services (e.g., prefetching, workflow management, and post-processing) that will facilitate the development of future clinical applications.

Fault-tolerant PACS server

Failure of a PACS archive server could cripple an entire PACS operation. Last year we demonstrated that it was possible to design a fault-tolerant (FT) server with 99.999% uptime. The FT design was based on a triple modular redundancy with a simple majority vote to automatically detect and mask a faulty module. The purpose of this presentation is to report on its continuous developments in integrating with external mass storage devices, and to delineate laboratory failover experiments. An FT PACS Simulator with generic PACS software has been used in the experiment. To simulate a PACS clinical operation, image examinations are transmitted continuously from the modality simulator to the DICOM gateway and then to the FT PACS server and workstations. The hardware failures in network, FT server module, disk, RAID, and DLT are manually induced to observe the failover recovery of the FT PACS to resume its normal data flow. We then test and evaluate the FT PACS server in its reliability, functionality, and performance.

Fault-tolerant PACS server design

A single point of hardware failure in PACS is at the PACS controller, or the main archive server. When it occurs, it renders the entire PACS inoperable and crippled until the problem is diagnosed and resolved. Many current PACS do not have a fault-tolerant design or adequate back-up system for the main archive server due to several issues including cost. Several large scale PACs utilize the Tandem or cluster design but are very costly and have not been critically tested for their degree of fault tolerance. This paper describes a novel, portable, and scalable fault-tolerant PACS controller design that is affordable for most PACS implementations.

Educational RIS/PACS simulator

Many educational courses have been designed for training radiologists and allied healthcare providers to operate PACS workstations. There are yet tools available for educational training of PACS concepts and workflow analysis. We have designed and implemented a RIS/PACS Simulator for this purpose. The RIS/PACS Simulator consists of six key components simulating a typical clinical RIS/PACS: RIS simulator, acquisition modality Simulator (AMS), DICOM gateway, PACS controller (UNIX-based), clinical viewing workstation, and network infrastructure with a 100mbits/sec Ethernet switch connecting to all these components. A generic RIS and a generic DICOM compliant PACS software package are used to simulate normal clinical data flow. Using this simulator, trainees can: 1. Observe clinical RIS/PACS operation, component by component 2. Trace image flow through each component 3. Identify PACS data flow bottle neck 4. Induce failure in a component to observe its impact on the PACS workflow and operation RIS/PACS simulator is a valuable tool for participants to gain knowledge of the complexity of RIS/PACS data flow with hands-on experience. As a stand-alone system, it also becomes a good test bed for evaluation of medical imaging applications without interrupting clinical workflow.

Clinical experiences with an ASP model backup archive for PACS images

Last year we presented a Fault-Tolerant Backup Archive using an Application Service Provider (ASP) model for disaster recovery. The purpose of this paper is to update and provide clinical experiences related towards implementing the ASP model archive solution for short-term backup of clinical PACS image data as well as possible applications other than disaster recovery. The ASP backup archive provides instantaneous, automatic backup of acquired PACS image data and instantaneous recovery of stored PACS image data all at a low operational cost and with little human intervention. This solution can be used for a variety of scheduled and unscheduled downtimes that occur on the main PACS archive. A backup archive server with hierarchical storage was implemented offsite from the main PACS archive location. Clinical data from a hospital PACS is sent to this ASP storage server in parallel to the exams being archived in the main server. Initially, connectivity between the main archive and the ASP storage server is established via a T-1 connection. In the future, other more cost-effective means of connectivity will be researched such as the Internet 2. We have integrated the ASP model backup archive with a clinical PACS at Saint Johns Health Center and has been operational for over 6 months. Pitfalls encountered during integration with a live clinical PACS and the impact to clinical workflow will be discussed. In addition, estimations of the cost of establishing such a solution as well as the cost charged to the users will be included. Clinical downtime scenarios, such as a scheduled mandatory downtime and an unscheduled downtime due to a disaster event to the main archive, were simulated and the PACS exams were sent successfully from the offsite ASP storage server back to the hospital PACS in less than 1 day. The ASP backup archive was able to recover PACS image data for comparison studies with no complex operational procedures. Furthermore, no image data loss was encountered during the recovery. During any clinical downtime scenario, the ASP backup archive server can repopulate a clinical PACS quickly with the majority of studies available for comparison during the interim until the main PACS archive is fully recovered.

PACS archive upgrade and data migration: clinical experiences

Saint Johns Health Center PACS data volumes have increased dramatically since the hospital became filmless in April of 1999. This is due in part of continuous image accumulation, and the integration of a new multi-slice detector CT scanner into PACS. The original PACS archive would not be able to handle the distribution and archiving load and capacity in the near future. Furthermore, there is no secondary copy backup of all the archived PACS image data for disaster recovery purposes. The purpose of this paper is to present a clinical and technical process template to upgrade and expand the PACS archive, migrate existing PACs image data to the new archive, and provide a back-up and disaster recovery function not currently available. Discussion of the technical and clinical pitfalls and challenges involved in this process will be presented as well. The server hardware configuration was upgraded and a secondary backup implemented for disaster recovery. The upgrade includes new software versions, database reconfiguration, and installation of a new tape jukebox to replace the current MOD jukebox. Upon completion, all PACS image data from the original MOD jukebox was migrated to the new tape jukebox and verified. The migration was performed during clinical operation continuously in the background. Once the data migration was completed the MOD jukebox was removed. All newly acquired PACS exams are now archived to the new tape jukebox. All PACs image data residing on the original MOD jukebox have been successfully migrated into the new archive. In addition, a secondary backup of all PACS image data has been implemented for disaster recovery and has been verified using disaster scenario testing. No PACS image data was lost during the entire process and there was very little clinical impact during the entire upgrade and data migration. Some of the pitfalls and challenges during this upgrade process included hardware reconfiguration for the original archive server, clinical downtime involved with the upgrade, and data migration planning to minimize impact on clinical workflow. The impact was minimized with a downtime contingency plan.

Methodology for design of adaptive interfaces for diagnostic workstations with integrated images and reports

Diagnostic workstations have generally lacked acceptance due to awkward interfaces, poor usability and lack of clinical data integration. We developed a new methodology for the design and implementation of diagnostic workstations and applied the methodology in diagnostic neuroradiology. The methodology facilitated the objective design and evaluation of optimal diagnostic features, including the integration of images and reports, and the implementation of intelligent and adaptive graphical user interfaces. As a test of this new methodology, we developed and evaluated a neuroradiological diagnostic workstation. The general goals of diagnostic neuroradiologists were modeled and directly used in the design of the UCLA Digital ViewBox, an object-oriented toolkit for medical imaging workstations. For case-specific goals, an object-oriented protocol toolkit was developed for rapid development and integration of new protocols, modes, and tools. Each protocol defines a way to arrange and process data in order to accomplish diagnostic goals that are specific to anatomy (e.g., a spine protocol), or to a suspected pathology (e.g., a tumor protocol). Each protocol was divided into modes that represent diagnostic reading tasks. Each mode was further broken down into functions supporting that task. Via a data mediator engine, the workstation communicated with clinical data repositories, including the UCLA HIS, Clinical RIS/PACS and individual DICOM compatible scanners. The data mediator served to transparently integrate, retrieve, and cache image and report data. Task-oriented Reading protocols automatically present the appropriate diagnostic information and diagnostic tools to the radiologist. We describe a protocol toolkit that enables the rapid design and implementation of customized reading protocols. We also present an intelligent layer that enables the automatic presentation of the appropriate information. This new methodology for diagnostic workstation design led to an improved neuroradiology workstation. Future research will focus on developing protocols for other diagnostic specialties.

Automatic segmentation of bones from digital hand radiographs

The purpose of this paper is to develop a robust and accurate method that automatically segments phalangeal and epiphyseal bones from digital pediatric hand radiographs exhibiting various stages of growth. The algorithm uses an object-oriented approach comprising several stages beginning with the most general objects to be segmented, such as the outline of the hand from background, and proceeding in a succession of stages to the most specific object, such as a specific phalangeal bone from a digit of the hand. Each stage carries custom operators unique to the needs of that specific stage which will aid in more accurate results. The method is further aided by a knowledge base where all model contours and other information such as age, race, and sex, are stored. Shape models, 1-D wrist profiles, as well as an interpretation tree are used to map model and data contour segments. Shape analysis is performed using an arc-length orientation transform. The method is tested on close to 340 phalangeal and epiphyseal objects to be segmented from 17 cases of pediatric hand images obtained from our clinical PACS. Patient age ranges from 2 - 16 years. A pediatric radiologist preliminarily assessed the results of the object contours and were found to be accurate to within 95% for cases with non-fused bones and to within 85% for cases with fused bones. With accurate and robust results, the method can be applied toward areas such as the determination of bone age, the development of a normal hand atlas, and the characterization of many congenital and acquired growth diseases. Furthermore, this methods architecture can be applied to other image segmentation problems.