PhD graduate Nicole Sultanum is the recipient of the 2022 Bill Buxton Dissertation Award for her research on text visualization research to support health-care workers navigating patient records. (Photo: Supplied)
In her doctoral dissertation, Nicole Sultanum explores using data visualization strategies to help health-care professionals navigate large and complex collections of text in patient records.
The recent PhD graduate from the University of Toronto’s Department of Computer Science has been named the 2022 recipient of the Bill Buxton Dissertation Award from the Canadian Human-Computer Communications Society (CHCCS) for her thesis, “Text-centric Visual Approaches to Support Clinical Overview of Medical Text.”
The national award is presented at the annual Graphics Interface conference for an outstanding doctoral dissertation completed at a Canadian university in the field of Human-Computer Interaction (HCI). It is named in honour of Bill Buxton, a Canadian computer scientist and HCI pioneer, who is also an adjunct professor at the University of Toronto’s Department of Computer Science.
“Receiving this award means a lot to me,” says Sultanum. “Past recipients of this award are people I look up to and who moved on to do great things in our field. I’m truly humbled to be listed among such accomplished and well-respected researchers in our community, and I just hope I can keep up with the good work they are doing.”
In observing the complex work being carried out by patient-facing health-care professionals in clinics and reading in between the lines of terse, objective medical records, Sultanum recognized the very personal stories in a patient’s file — the way they struggled with their illness, the ups and downs of the illness trajectory, how it affected their families and the diligent efforts of the professionals involved.
To make health-care work more efficient and comprehensive, and to pay homage to those stories, Sultanum uses computational tools to develop meaningful visualizations for pattern recognition and insight discoveries. She explores how data visualization can benefit from a more text-centric approach to help health-care workers peruse text portions of a patient’s record with ease, while still leveraging the power of visual representations.
Taking an iterative, user-centred approach, Sultanum created interactive visualization prototypes: MedStory, Doccurate, and ChartWalk.
In her thesis, Sultanum’s major contributions include leveraging a balanced mix of data visualization summaries and text-based views to solve the issue of navigation within large text collections; a semi-automated curation mechanism that helps users filter and organize information of interest in large collections of text; and a list of principled design requirements and innovative interaction and visualization techniques for black-box automation.
These prototypes were validated by health-care professionals and Sultanum’s findings underscore the importance of text content being visible front and centre, while also informing new ways in which information could be summarized and organized for better overview.
“The thesis contributes highly original results which have redefined how to build visual analytics solutions aimed to enable the understanding of very large, unstructured text corpora aided by natural language processing,” notes the CHCCS.
Sultanum was co-supervised by Department of Computer Science Assistant Professor Fanny Chevalier and Professor Michael Brudno who praise her “boundless passion” for her research.
“We are very proud to see Nicole’s exceptional research be recognized by the Bill Buxton award,” says Chevalier. “Nicole’s eagerness to solve hard and important problems is inspiring. With her natural curiosity and creativity, combined with exemplary scientific rigour, she produced contributions which will undoubtedly make a strong impact. It is exciting to see her achievements be acknowledged by the outstanding doctoral dissertation in the field of Human-Computer Interaction; and we look forward to seeing her next achievements as a senior researcher at Tableau Research.”
Beyond her time at U of T, Sultanum has her sights set on continuing to make significant contributions in HCI.
“My thesis was about helping health-care professionals navigate large and complex collections of medical text using data visualization,” she says. “In my current role, I continue to explore ways to make text more accessible and compelling for people in domains beyond medicine. A large portion of the world’s unstructured data is estimated to be text, so I’d say there’s plenty of work ahead of me!”