New Center Brings Top Researchers to UT School of Health Information Sciences

In addition to the recent recruitment of two world-renowned informatics leaders, Vimla Patel, PhD, DSc and Edward Shortliffe, MD, PhD, The University of Texas School of Health Information Sciences at Houston (SHIS) is adding other researchers who will each contribute expertise in their own area of research. All will play an important role in the success of the school’s proposed Center for Cognitive Informatics and Decision Making (CCIDM).

“This is an exciting time for our school,” said Jack Smith, MD, PhD, Dean of The University of Texas School of Health Information Sciences, “We are pleased to announce the formation of this new Center as well as the prestigious caliber of the team that comes along with it.”

The center will conduct basic, translational and applied research in medical decision-making and medical error, and develop advanced cognitive informatics tools for clinicians and patients. Four areas of research will be addressed; medical errors, distributed cognition, human-centered design of health information systems, and simulation and modeling of teamwork. The areas of medicine that will be most impacted by these initial four foci are emergency medicine, intensive care and preventative medicine.

Collaboration between the Center and institutions such as Memorial Hermann Healthcare System, UT M.D. Anderson Cancer Center, Rice University, the Gulf Coast Consortia, The Institute for Rehabilitation and Research or TIRR, and the Mental Sciences Institute will help to support this effort.

Cognitive informatics applies components of neuroscience, psychology, linguistics and artificial intelligence to study the human brain and mind, specifically internal information processing mechanisms and natural intelligence. It is important to the implementation of health information technology, a focus of the $19 billion government stimulus package. Health IT has the potential to improve quality care, increase patient safety, and reduce cost, but with this comes some great challenges.

“The Center will bring together leading researchers in cognitive science, computer science, and medical science and will establish SHIS as the world leader in cognitive biomedical informatics,” said Jiajie Zhang, PhD, Dr. Doris L. Ross Professor, associate dean of research at The University of Texas School of Health Information Sciences, who will co-direct the Center with Patel.

In addition to Patel and Shortliffe, current SHIS faculty, including Zhang and Associate Professors Hongbin Wang, PhD and Todd Johnson, PhD, associate dean for academic affairs, The UT Medical School at Houston faculty, postdoctoral fellows, and dozens of masters and doctoral students will all play a role in the Center’s success. Furthermore, several researchers have been recruited to fill positions pertinent to bringing together all aspects of the Center.

Trevor Cohen, MBChB, PhD

An expert in distributional semantics and cognition and error recovery, Trevor Cohen, MBChB, PhD, will lend his expertise to the Center when he joins the school as an assistant professor later this year.

Cohen comes to Houston from Arizona State University where he was an assistant professor in biomedical informatics and worked closely with Patel and the Center for Decision Making and Cognition that she directed. Cohen first met Patel when his interest in psychiatry, cognitive research, and understanding the minds of physicians led him to Columbia University where he became one of her doctoral students.

Originally from South Africa, Cohen trained to become a physician at the University of Cape Town. He remained in the country for his residency, where he spent much of his time working in psychiatry in a relatively rural and under-resourced area.

“Our patient records consisted of single sheets of paper which were filed by date of admission,” said Cohen, “So in order to establish a past medical history we’d need to consult the admissions log book and try to determine all the past admissions for a particular patient. This is something of a problem in a situation in which patients are psychotic and unable to give a good account of themselves, frequently use pseudonyms and recall their age in relation to historic events such as floods. It certainly made me acutely aware of the need for biomedical informatics.”

Cohen’s primary research interest is in empirical distributional semantics, which is the study of what computers can learn from the written word and how this learning relates to human cognition. 

“There’s a rich tradition in cognitive science that considers human knowledge to be organized as a network of associations between concepts,” said Cohen, “that is to say, when presented with the word ‘cat’, associated words such as ‘mouse’ spring to mind.”

The computer models used in this work are able to learn such associations from electronic text.

“In a biomedical context,” said Cohen, “as there’s much more information available in electronic form than any human could digest over several lifetimes, the question arises of how these human-like associations the computer model has learned might be used to enhance the ability of scientists to conduct research.”

Cohen believes, answering this question requires not only the development of new models that are able to recognize previously unknown connections between terms, but also the study of how scientists are able to use these associations to answer research questions of their own.

Additionally, Cohen is interested in how humans detect and recover from error, in particular medical error. Within the culture of medicine, he says the traditional approach to error involves assigning blame to a single individual. However, Cohen believes this fails to address the complexity of the system within which medical error occurs.

“A big part of Cohen’s contribution to the Center,” said Zhang, “will be his extensive knowledge in teamwork and social networking in the critical care environment.”

Cohen, whose work in this area began at Columbia University, will be working to optimize the efficiency of clinical workflow, and improve understanding of team decision-making.

“There’s been a lot of attention on patient safety as of late and as a society we tend to focus on isolated headline-grabbing incidents in which a particular patient has been harmed,” said Cohen, “ However, no system as dynamic and complex as a critical care unit can realistically expect to eliminate all error, so it is at least of equal importance to study the ways in which clinical teams contribute toward patient safety by detecting and correcting potential error.”

Amy Franklin, PhD

Cognitive linguistics expert Amy Franklin, PhD, will be joining the school later this year as an assistant professor. Most recently a postdoctoral fellow at Rice University, Franklin’s research goal is to better understand the connection between gesture, language and the mind.  Her interests explore this connection, including the capacity to develop linguistic structure in gesture systems, studies investigating the expression of meaning across verbal and non-verbal modalities, and the role of relationships in language learning.

“She’s the perfect person to help us build the bridge between cognitive linguistics and biomedical informatics,” said Zhang, “This is the next fundamental problem for our area of research and she will be here to help us solve it.”

As a cognitive linguist, Franklin uses language as a window to the mind. She examines how language performance reflects thinking and how ideas are coordinated between individuals.

“Cognitive linguistics fits well into biomedical informatics with a shared emphasis on the representation, processing and communication of information,” said Franklin, “I hope to contribute to our understanding of embodied thinking and its application in health care informatics.”

When she joins the school and the new Center, she will be working on a range of projects from studies of decision-making and distributed cognition to experiments exploring the use of gesture in medical records.

The addition of cognitive linguistics to biomedical informatics will advance the goal of reducing medical errors and improving care through a better understanding of the human thought processes including the dynamics of group and human-computer interactions.

“For example,” said Franklin, “when a team collaborates in the care of a patient a great deal of information is exchanged and negotiated both verbally and non-verbally. By improving our understanding of how people come to agree on a shared representation of a patient’s condition through the language and gestures they produce, we can create decision support systems that more accurately reflect thinking and facilitate group interactions.”

A Houston-native, Franklin received her undergraduate degree in Psychology from the University of Houston. She then went on to receive her MA and PhD in Linguistics and Psychology from the University of Chicago. She has authored and co-authored numerous papers and has been invited to present her work at conferences worldwide. Franklin has received research support from the Keck Foundation, National Science Foundation, Department of Defense, and the National Institute of Health.

by Rachel Bailey, School of Health Information Sciences