IBM, Mayo Clinic Unveil Medical Data Initiative For Open Source Developers

IBM and the Mayo Clinic have partnered to offer the open source development community a crack at their new natural language processing (NLP) initiative, the Open Health Natural Language Processing Consortium.

The hope, according to IBM, is that advancements in NLP will help speed the adoption of electronic medical records (EMR) in health-care systems around the world. IBM and Mayo on Thursday debuted a new Web site, www.ohnlp.org, where open source developers can access their tools.

NLP is a specialized area of computer science that in a health-care setting enables the searching and indexing of information from patient-clinician interactions. That is, with fully realized NLP, a doctor would be able to enter search terms into an EMR and call up the exact information he or she needs to help diagnose a patient's condition without having to thumb through reams of paper records or cumbersome computer files.

That doctor could access information on studies that had been done on a particular condition, or previous diagnoses that have been made, to assist his or her diagnosis and ongoing disease research.

The idea is that doctors won't have to rely solely on their own clinical experience to make decisions, IBM said. They have patient history, case studies and the latest information at their disposal, and they can "confer," so to speak, with other physicians to make better diagnoses.

"This is a natural extension of the Mayo-IBM partnership," said Dan Pelino, general manager of IBM's Global Healthcare & Life group, in a Channelweb.com interview. "We're looking for areas where we can bring research capabilities and domain expertise."

The plan is one of several new health-care initiatives IBM is said to be unveiling in line with next week's Healthcare Information and Management Systems Society (HIMSS) conference in Chicago.

IBM and Mayo have released their two clinical text NLP technologies -- one on clinical notes for physician/patient encounters and one on pathology reports for tissue findings -- into public domain to more than 2,000 open source developers and researchers worldwide.

Pelino called these types of opportunities "tremendous" for IBM's partner community within health care.

"If they can continue to build out solution sets that take advantage of these standards-based procedures and open source protocols, they're going to help solve these problems," Pelino said. "There is tremendous opportunity for creating value."

"We are inviting our international colleagues to help continue development of these valuable tools," added Christopher Chute, a bioinformatics expert at the Mayo Clinic and a senior consultant on the consortium, in a statement. "By making it an open source initiative, we hope to enable wide use of these NLP tools so medical advancements can happen faster and more efficiently."

Security risks over the privacy of patient data are still a big deterrent to making EMRs truly interoperable, Pelino said.

"The more we do this around standards, the easier it is to build secure, private, bulletproof infrastructure," he said. "As you build off of standards-based protocols to exchange information, you can also build skill sets to ensure you have a certain level of privacy. The more we can put out into the open domain to build that level of privacy and security, the better we can help patients."

Pelino said he hoped IBM and Mayo's push for open source development helps the IT community give open source a fair shake when it comes to software development in health care.

"The way that this collaboration will help health-care reform is to make this information readily available," he said. "If IBM and Mayo say this helps solve a problem and we put it into an open source environment, we're hoping people will take it seriously."

Tweet

   

More