How Does Research Inform Scientific Discovery?

Guest post by Charitie Martino, a research assistant and undergraduate student in physics at the University of South Florida, Tampa. Ms. Martino was a summer intern in the 2023 NLM DataWiz-IN Scholars Program at Indiana University.

This past summer I worked for the DataWiz-IN Scholars Training Program, funded by the NLM R25 Research Education Program, as a lab scholar in Dr. Hee-Tae Jung’s Health-Technology Laboratory at Indiana University. While there, I learned the importance of balancing two opposing approaches to research that are often seen as two sides of a deep chasm within academia. As Mark Fishman, a professor at the Harvard Department of Stem Cell and Regenerative Biology, said in a 2018 Harvard Gazette article, “The discoveries that lead to the creation of a new medicine do not usually originate in an experiment that sets out to make a drug. Rather, they have their origins in a study—or many studies—that seek to understand a biological or chemical process.”

But on the other hand, like most biomedical informatics work, there are tools based in problem-solving that have catapulted scientific knowledge: Just as the telescope heralded the chance discovery of the cosmic microwave background (CMB) and 3D printing meant we could “hold” the CMB, computer simulations, mathematical models, and statistical analyses have enabled computational neuroscientists to model grid cells in the brain and map three-dimensional space in the mind.

So should you love knowledge for its own sake and not necessarily as a blueprint to action? Or does direct clinical application and solving information problems deserve priority when making scientific discoveries?

With Dr. Jung’s guidance, I conducted a small-scale project that investigated the interaction between visual complexity and cognitive overload. Because his lab focuses on maximizing the quality of health care-related services for people with chronic conditions, this initial analysis will provide the foundational research needed to develop virtual reality therapies for individuals with traumatic brain injuries (TBI).

Though my interests in neuroscience have been more theoretical—for example, consciousness theories and brain entropy research—I appreciated the opportunity this program offered because of its basis in clinical and application research. I expected this summer research experience to tilt me in those directions or, at the very least, help inform how I study theoretical neuroscience in the lab so I would understand more clearly the implications it might have in the real world.

Generating a pilot study for health purposes with informatic implications required me to focus on finding one well-defined solution to a problem, starting with conceptualizing and defining visual complexity and then determining how it should be measured within the study. My mind naturally goes to more theoretical, foundational questions (How does the neuroplasticity of TBI patients differ than that of neuroplasticity in child development? How do their personal experiences shape hierarchical visual perception of color? Are universal responses more prevalent?), but because it was for the purpose of future therapy development, I anchored my scope and narrowed subsequent research questions.

As people with acquired brain injuries often suffer from cognitive impairments, this negatively affects their engagement in rehabilitation therapies. While virtual reality can immerse users in realistic activities of daily living and practice, too much visual stimuli can distract or cognitively overload people with TBI, which may delay therapeutic benefit.

My summer research internship taught me that the experimental design process in biomedical informatics involving the use of data can drive efforts to improve human health. Data science, in its usage of statistics, scientific computing, algorithms, and systems to extract meaningful knowledge and insights from noisy, unstructured data, hopefully will provide a solution to this challenge. As the purpose of the NLM R25 program is to bring exposure to and training in areas related to biomedical informatics and data science research, it will continue to shape my field of study, experimental and computational neuroscience, and my perspective.

Ultimately, while I do not see myself specifically pursuing clinical or data science, this experience has been invaluable in shaping my outlook for how disciplines within science need to collaborate to achieve a more cohesive whole.

Featured image attachment page