The same day that IBM’s Watson was trouncing human contestants on the US television game show Jeopardy!
, the IT giant revealed the next evolutionary step in the supercomputer’s development. A joint press release drafted with Nuance Communications (Burlington, MA, USA; www.nuance.com
) announced a research agreement to explore, develop and commercialise the Watson computing system’s advanced analytics capabilities in the healthcare industry.
Nuance is a leader in speech recognition, medical transcription and clinical documentation software, and an IBM Preferred Business Partner since 2009. This expertise, which IBM (Armonk, NY, USA; www.ibm.com
) intends to incorporate into Watson’s skill set, will provide medical terminology speech recognition and Clinical Language Understanding (CLU) solutions for diagnosis support and treatment of patients. CLU enables physicians to dictate clinical notes directly into electronic health records (EHRs). Watson’s ability to analyse the meaning and context of human-language EHR entries and quickly process information to find precise answers can assist physicians and nurses in unlocking important knowledge and facts buried within immense volumes of data. This diagnostics support can offer answers that may not have been considered, and may help doctors validate their own ideas or hypotheses.
“Combining our analytics expertise with the experience and technology of Nuance, we can transform the way that healthcare professionals accomplish everyday tasks by enabling them to work smarter and more efficiently,” says Dr. John E. Kelly III, Senior Vice President and Director of IBM Research. “This initiative demonstrates how we plan to apply Watson’s capabilities into new areas, such as healthcare.”
“The combination of Nuance’s speech recognition and existing Clinical Language Understanding solutions with the power of IBM’s Watson technology will introduce unmatched clinical information and analytic technological advancements for healthcare,” says Paul Ricci, Chairman and CEO of Nuance. “The initiative represents a logical step in Nuance’s evolution, one that expands our capabilities from recognising what was said to understanding the intent and providing guidance.”
Underpinning this synergistic leap forward in computer capabilities are recent advances in several computer science disciplines from universities working with IBM on the Watson project. Massachusetts Institute of Technology, University of Texas at Austin, University of Southern California, Rensselaer Polytechnic Institute, University at Albany (UAlbany), University of Trento (Trento, Italy), and University of Massachusetts Amherst each contributed aspects of the solution for this immensely complex challenge.
Carnegie Mellon University (CMU) is working with IBM on the development of a first-of-its-kind open architecture that enables researchers to efficiently collaborate on underlying question answering (QA) capabilities and then apply their solutions to IBM’s Watson system. The Open Advancement of Question Answering Initiative (OAQA) architecture, under development by IBM and CMU researchers led by Professor Eric Nyberg, CMU School of Computer Science, offers the ability to rapidly customise and adapt the QA system to deliver higher levels of performance on new domains, new data, new languages and different question types and styles.
OAQA is an open architecture, providing “a foundation for effective collaboration among researchers to accelerate the science of automatic question answering. The initiative aims to develop common metrics, architectures, experimental methodologies (and) tools ... to facilitate the collaborative advancement of the state-of-the-art,” IBM says in a press release. The OAQA architecture supports and standardises simultaneous research experiments, and its open-source software serves as a platform to enable participants to contribute to QA enhancements.
Working with this OAQA architecture is IBM’s DeepQA architecture, which refines and extends open-domain question answering to allow a computer to potentially draw from the entire breadth of human knowledge as it formulates its response. IBM says that “Watson uses massively parallel processing to simultaneously and instantly understand complex questions . . . that require the system to consider huge volumes and varieties of natural-language text to gather and then deeply analyse and score supporting or refuting evidence. The system then decides how confident it is in the answer. This approach marries advanced machine learning and statistical techniques with the latest in natural-language processing to result in human-like precision and speed, huge breadth and accurate confidence determination.”
Interactive QA capability for sustained investigation is under development by Professor Tomek Strzalkowski and his team of researchers at UAlbany. This enables a computer to remember the entire interaction, rather than treating every question as a first question. It not only simulates a real dialogue; it also enables the computer to give better answers more quickly, since the computer remembers the context of the conversation. IBM is working with UAlbany to integrate this capability into the Watson system as it is intended to become the engine underlying numerous healthcare, commercial, industrial and government applications.
IBM and Nuance say that introduction of a commercial offering from the collaboration is 18 to 24 months away. That introduction may prove to be a seminal moment in medical history, but it is only a first step. Next, and nearing development, is the integration of Watson technology with automated clinical laboratory, digital image management, and other current healthcare applications. While science fiction’s long-prophesied robodoc is not yet here, it is undeniably one step closer.