The Ohio State University’s Translational Data Analytics Institute has partnered with the Center for Clinical and Translational Science, Department of Biomedical Informatics in the College of Medicine, and Kirwan Institute for the Study of Race and Ethnicity to support five interdisciplinary pilot research teams that are creating artificial intelligence technology for biomedical use and incorporating social impact concerns into data science curriculum. The partners’ unique combination of funding and in-kind resources is enabling faculty teams that together represent seven colleges and 11 different disciplines.
The pilot awards are the outcomes of a new approach to research grants created by TDAI in which teams engage in ideation workshops and are guided to leverage areas of overlap among their proposals in order to maximize resources and impact. Similarly, the funding entities’ collaboration to pool resources allows them to support more promising projects that align with their individual priorities. As a result, the awards afford recipients access to resources at a lower cost than if they were simply granted funds.
“In a sense it’s an interdisciplinary approach to enabling interdisciplinary research,” said TDAI Faculty Director Tanya Berger-Wolf, PhD. “By joining forces, TDAI, CCTS, BMI and the Kirwan Institute can support more game-changing work across the board than we otherwise could by ourselves.”
AI and bioimaging
Four of the teams receiving awards are developing artificial intelligence and machine learning tools and workflow processes for clinicians to use to analyze medical images. In addition to innovating efficient, low-cost alternatives to current technologies for diagnosing and treating conditions in four different medical specialties, the teams are leveraging their combined aims to compete for extramural funding to support future cutting-edge work at a much larger scale.
“Bringing engineers together with experts from diverse medical domains allows each of these teams to develop solutions from medical practitioners’ perspective as well from the engineering side,” said CCTS Director Rebecca Jackson, MD, principal investigator of Ohio State’s NCATS Clinical and Translational Science Award. “The ability to assemble this range of high-caliber expertise on one campus is Ohio State’s great strength,” said BMI Chair Lang Li, PhD.
Zaibo Li (Medicine) is leading the design and development of deep learning systems to help predict chemotherapy benefits and the risk of tumor recurrence for the most common type of breast cancers. In addition to improving patient outcomes, the project will provide a more affordable alternative to a current test that is prohibitively expensive, improve understanding of the relationship between genetic changes and histopathologic image features, and bridge mechanistic understanding of tumor biology with clinical experience. The team includes Anil Parwani, (Medicine) and Raghu Machiraju (Engineering/Medicine).
José Javier Otero (Medicine) is leading an effort to improve and standardize testing and diagnoses for brain tumor patients worldwide by validating new workflows that use digital imaging, immunohistochemical tests, open source computing platforms and machine learning algorithms. Also on the team are Catherine Czeisler (Medicine) and Raghu Machiraju (Engineering/Medicine).
Rajiv Ramnath (Engineering) is leading a team developing human-machine systems to facilitate image-driven pathology and optometry, and is leading proposal development for an extramural grant. Other members of the team are Jian Chen, Eric Fosler-Lussier, D.K. Panda, Srinivasan Parthasarathy and Hari Subramoni (Engineering); Raghu Machiraju (Engineering/Medicine); Anil Parwani (Medicine); and Nathan Doble and Dean VanNasdale (Optometry).
Matthew Tong (Medicine) is leading a study to leverage a large cardio magnetic resonance (CMR) registry database to assess accuracy, cost-effectiveness, and downstream testing in CMR vs. non-CMR based approaches for three common cardiovascular disease states and explore machine learning to refine prognostication leveraging deep phenotyping and large datasets. Co-investigators are Daniel Addison (Medicine), TDAI core faculty member Ping Zhang (Engineering/Medicine), Orlando P. Simonetti (Medicine) and Richard Gumina (Medicine).
Social, Political and Ethical Dimensions of Data Science
A team led by Inés Valdez (Political Science) and Mark Moritz (Anthropology) is creating curriculum and tools for training data science students, researchers and professionals that consider how big data and algorithms might amplify pre-existing inequalities, discriminatory practices and power disparities. The team includes TDAI core faculty member Dennis Hirsch (Law), Dana Howard (Medicine/Philosophy), Samantha Krening (Engineering), Samuel Malloy (Public Policy) and Srinivasan Parthasarathy (Engineering).
“As data science technology like artificial intelligence is adopted in every aspect of our society, ensuring that the people creating it are vigilant about its social implications is paramount,” said Berger-Wolf. “Something as seemingly straightforward as choosing data for training new tools can lead to biased outcomes that cause irreparable damage to people’s lives.”
In light of its new funding program’s successful first year, TDAI is engaging additional resource partners for its 2020-21 call for pilot research proposals, which will address topics including racial justice, the environment and art and their intersections with data analytics.
“We have such a wealth of talent and imagination at Ohio State,” Berger-Wolf said. “Our job at TDAI is to help that talent come together in ways that strengthen the impact of our work and amplify its ability to help the world and the people who live in it.”