A new cognitive computing project that enables more natural interaction between physicians, data and electronic medical records.
After a year-long research collaboration with faculty, physicians and students at Cleveland Clinic Lerner College of Medicine of Case Western Reserve University, IBM Research has unveiled two cognitive computing technologies that can be used by Watson, and are expected to help physicians make more informed and accurate decisions faster and to cull new insights from electronic medical records (EMR). The projects known as “WatsonPaths” and “Watson EMR Assistant” will create technologies that can be used by Watson in the domain of medicine.
With the WatsonPaths project, IBM scientists have trained the system to interact with medical domain experts in a way that’s more natural for them, enabling the user to more easily understand the structured and unstructured data sources the system consulted and the path it took in offering an option. The Watson EMR Assistant project aims to enable physicians to uncover key information from patients’ medical records, in order to help improve the quality and efficiency of care.
"Through our research collaboration with Cleveland Clinic, we've been able to significantly advance technologies that Watson can leverage to handle more and more complex problems in real time and partner with medical experts in a much more intuitive fashion. These are breakthrough technologies intended to assist future versions of Watson products." said Eric Brown, IBM Research Director of Watson Technologies.
- Read the press release
- Michael Barborak discusses the future of Watson and how computers will interact naturally with people
Explore the WatsonPaths interface
- Scenario analysis For WatsonPaths, in the background we use Watson's question-answering capabilities like a pick to chip away at a complex scenario and break it down into comprehensible pieces. Here we are looking at such a scenario and a few factors identified for Watson to begin exploring.
- Solution In this hypothetical case, when questions are generated for Watson about the scenario. WatsonPaths can begin to make relevant inferences. By combining these inferences into a graph, it creates a network in which support for the hypotheses can be propagated and shows to the user to help teach critical reasoning skills.
- Solution exploration Answers without explanation are unsatisfying and that's why WatsonPaths is programmed to report its results with confidence (meaning the found lots of supporting evidence) along with evidence (typically passages from documents). Since every connection in the WatsonPaths solution is based on having asked Watson a question, there will be some evidence to support it. Here, the user is choosing a connection to explore, again to help teach the user critical reasoning skills.
- Viewing evidence The WatsonPaths solution graph is based on Watson doing thousands of searches across tens of millions of pages of text and considering thousands of candidate options. The WatsonPaths solution graph highlights the most significant results of that massive analysis. Here, the user is exploring evidence behind a single answer to a single question to help the physician or student learn critical reasoning skills.
- The response WatsonPaths provides responses to questions. WatsonPaths uses that capability to analyze scenarios and support or refute hypotheses and the "response" is supported through the evidence it presents. And if you disagree with Watson? Let it know. The learning teaching tool is a cognitive computing application where each interaction by the user is an opportunity for WatsonPaths to learn and improve.
Next
Show carousel 1Show carousel 2Show carousel 3
Using WatsonPaths to support clinical reasoning
The WatsonPaths project explores a complex scenario and draws conclusions much like people do in real life. When presented with a medical case, it extracts statements based on the knowledge it has learned from being trained by medical doctors and from medical literature.
Using Watson’s question-answering abilities, WatsonPaths can examine the scenario from many angles, working its way through chains of evidence – pulling from reference materials, clinical guidelines and medical journals in real-time – and drawing inferences to support or refute a set of hypotheses. This ability to map medical evidence allows medical professionals to consider new factors that may help them to create additional differential diagnosis and treatment options.
As medical experts interact with WatsonPaths, the system will use machine-learning to improve and scale the ingestion of medical information. WatsonPaths incorporates feedback from the physician who can drill down into the medical text to decide if certain chains of evidence are more important, provide additional insights and information, and weigh which paths of inferences the physician determines lead to the strongest conclusions. Through this collaboration loop, WatsonPaths compares its actions with that of the medical expert so the system can get “smarter”.
Once ready, WatsonPaths will be available to Cleveland Clinic faculty and students as part of their problem-based learning curriculum and in clinical lab simulations.
- Read an interview with Dr. Neil Mehta, Associate Professor of Medicine and Director, Education Technology at Cleveland Clinic Lerner College of Medicine of Case Western Reserve University
“WatsonPaths is designed to augment the problem-based learning methods that Cleveland Clinic medical students employ in the classroom. The vision is for WatsonPaths to act as a useful guide for students to arrive at the most likely and least likely answers to real clinical problems, but in a classroom setting.”
-J. Eric Jelovsek, MD, MMEd, Director of the Cleveland Clinic Multidisciplinary Simulation Center
Watson EMR Assistant: Unlocking the promise of electronic medical records
IBM and Cleveland Clinic are also using Watson to explore how to navigate and process EMRs to unlock hidden insights within the data, with the goal of helping physicians make more informed and accurate decisions about patient care.
The massive amount of health data within EMRs alone presents tremendous value in transforming clinical decision making, but can also be difficult to absorb. For example, analyzing a single patient’s EMR can be the equivalent of going through 100MB of structured and unstructured data, in the form of plain text that can span a lifetime of clinical notes, lab results and medication history.
Watson’s natural language processing capabilities allows it to process an EMR with a deep semantic understanding of the content and can help medical practitioners quickly and efficiently sift through the massive amounts of complex and disparate data and better make sense of it all. Working with de-identified EMR data provided by Cleveland Clinic, and with direction from Cleveland Clinic physicians, the goal of the Watson EMR Assistant research project is to develop technologies that will be to collate key details in the past medical history and present to the physician a problem list of clinical concerns that may require care and treatment, highlight key lab results and medications that correlate with the problem list, and classify important events throughout the patient’s care presented within a chronological timeline.
Smart Machines: IBM's Watson and the Era of Cognitive Computiing
By John E. Kelly III and Steve Hamm
- Download a free chapter
- Purchase the book from Columbia University Press Use code SMART to receive a 30% discount.
- Purchase the book from Amazon
Learn more about Watson and healthcare
- Putting Watson to work
- Why cognitive systems?
- WatsonPaths demo (no audio) (00:01:07)
- Analytics in healthcare
- Explore smarter healthcare solutions
