Raymond D. Aller, MD
Michael Weilert, MD
IBM and Mayo collaborating to customize patient treatment plans
In an innovative cross-industry partnership, IBM and the Mayo Clinic are channeling their know-how into developing customized treatment plans for patients.
The goal of the multi-year project between IBM and the Mayo Clinic, which kicked off in August, is to put patient research data from Mayo and other institutions to greater use in patient care by linking that data with new types of medical information.
The partners have completed the first phase of the project, which involved integrating 4.4 million patient records from Mayo’s various databases into a single repository, and are deploying the second. Phase two involves integrating new types of data, beginning with data from genetic testing and moving to research data from proteomics and molecular modeling.
The second phase focuses on developing higher levels of mining capabilities to compare the data from patients with similar disease characteristics and a similar genomic makeup, explains Drew Flaada, who is directing the collaboration for IBM. Part of this work has involved designing automated methods for querying and interpreting the unstructured online text in physicians’ transcribed notes.
"It took a lot of intelligence from IBM and the Mayo research to understand what text analysis techniques we can apply and then how we can apply the right medical knowledge against that to make sure we mine the information effectively," Flaada says.
The second phase also involves integrating patient information from Mayo’s two other facilities, in Scottsdale, Ariz., and Jacksonville, Fla., and continuing to add other types of data, including public records, such as the database of the National Cancer Institute.
While the initial focus of the project is to integrate research, the ultimate goal is to use genetic information to develop highly customized treatment plans for patients. The Mayo Clinic will use IBM’s Blue Gene supercomputer for that aspect of the project involving molecular modeling for disease research.
"The ultimate goal is to have evidence-based, customized patient care, [with] obviously lots of steps along the way," says Mayo internist Nina Schwenk, MD, who heads Mayo’s information technology committee and directs the project for the clinic. The project includes several levels of safeguards to comply with HIPAA regulations and protect patient privacy, says Dr. Schwenk, who predicts the partnership will directly benefit patient care in five to 10 years.
The project is already producing promising results. In a test case for phase one, a five-person research staff took 13 months to comb through Mayo’s various databases for a representative group of participants for a research project. After Mayo’s patient records were integrated, the project’s primary investigator inputted the study’s parameters into a computer tool called Data Discovery and Query Builder. "An answer came back for a representative study group in 16 seconds," says Flaada, who believes this capability will change the way researchers approach their work. "I think you’re going to see a lot more flexibility and creativity on the researcher’s part about, ’Well, what if this parameter were different? What kind of answers would I get back?’"
The seeds of the IBM-Mayo project were planted in December 2001, says Flaada, during a conversation at a cocktail party attended by IBM and Mayo employees who discussed the clinic’s information technology needs for genetics research. That conversation led to more formal discussions, and the plan took off.
The partnership makes perfect sense, says Flaada, noting that Rochester, Minn., houses Mayo’s headquarters as well as a major IBM development laboratory.
So far, Flaada says, there have been no major technical problems. Rather, the challenge has been "trying to get everything to interface together correctly to provide information in a way that a physician can deal with it rather than a computer scientist."
Interactive Web site focuses on antimicrobial susceptibility
Roche is supporting Susceptibility
Central.com, an online resource designed to provide microbiologists, physicians,
and other health care professionals with data showing national and regional
antimicrobial susceptibility trends and information about antimicrobial susceptibility
surveillance programs. The Web site offers antibiogram software that institutions
can download and use to collect data from their own facilities to establish
local trends and update treatment recommendations. It also provides links to
other Web sites that offer information about antimicrobial susceptibility, susceptibility
testing, and antimicrobial resistance, as well as a calendar of events that
address antimicrobial susceptibility. The site includes national and regional
antimicrobial susceptibility data from the Surveillance Network Database USA
and the Antimicrobial Resistance Management Program.
Silicon Genetics upgrades gene expression software
Silicon Genetics is marketing GeneSpring 7 and Signet 5, the latest versions of the company’s software for analyzing gene expression information. The new releases offer extensive analysis capabilities and robust application programming interface architecture for developing customized research workflows.
Among the enhancements to GeneSpring 7 are learning algorithms that use prior knowledge of sample classes to identify unknown samples; project-based organization to organize data and analysis results from multiple technology platforms or multiple species into projects; a new Genome Wizard for genome import and management; new algorithms to find associations and correlations between parameters and expression profiles to help determine causative relationships; and Java APIfor extending GeneSpring with custom analyses or visualizations.
Signet 5 now provides informatics departments with a series of interfaces that allow administrators to build a custom informatics work space inside Signet, a scalable platform for managing and mining expression data in a work group or enterprise setting. The improvements to Signet 5 include the Signet SOAPAPI, which allows access to many of the data types within Signet; RemoteServer Native Java plug-in API, which gives GeneSpring users an interface to applications deployed on a Signet server farm; project-based navigation tools to search and sort data according to work groups or experiment type; and modifiable work spaces that allow Signet users to designate personal areas where data can be edited and analyzed.
ITC to integrate blood gas analyzer with data-management system
International Technidyne Corp. has announced plans to integrate its IRMATrupoint analyzer for point-of-care blood gas monitoring with Medical Automation Systems’ RALS-Plus data-management system.
With RALS-Plus, IRMA Trupoint users will be able to automatically manage, report, and electronically transfer patient-monitoring data for blood gases, electrolytes, and blood chem istries to the RALS-Plus database and the laboratory information system.
Digital pathology service platform
Trestle Holdings Inc. has released ePathNet.net, an online data-management and virtual microscopy slide viewing service for pathology laboratories.
With ePathNet.net, pathologists send Trestle their prepared tissue samples on glass slides, which are converted to high-resolution digital slides that can be manipulated and viewed using the system’s online virtual microscopy functions. The digital slides can be stored on Trestle’s secure servers and are accessible via the Internet for search, review, downloading, and analysis.
EPathNet.net offers numerous data mining functions, such as indexing, annotation, searching, and image processing. It is a pay-as-you-go service.
Upcoming Lab InfoTech summit
The Pathology Education Consortium’s Lab InfoTech Summit 2005 will be held March 2-4 in Las Vegas.
The conference focuses on clinical laboratory information technology and is designed for pathologists, medical technologists, laboratory information system managers, lab business managers, and other laboratory professionals. Approximately 30 vendors are expected to exhibit at the meeting.
For more information, go to www.labinfotech.org/LIS2005/index.php.
Wyndgate Technologies, a division of Global Med Technologies, has signed agreements for its SafeTrace Tx advanced transfusion management system with the MetroHealth System, Cleveland, and the University of North Carolina Hospitals, Chapel Hill.