“We are thrilled to welcome one of the foremost data scientists to our team,” said LIO CEO Heather Morgan Shoemaker. “This is an all-woman, three-person team consisting of Dr. Dodel, Alanna Larson, Jing Zhao and will help us bring accurate, real-time, conversational translation to the entire world. We are thrilled to welcome Dr. Dodel to our team. Only one in four data scientists are women and we’ve hired one of the best to continue improving our proprietary core technology.”
Dodel is a woman obsessed with networks whether they be human, neural, or molecular. A German native, Dodel started her career developing graph theoretical patterns of whole brain functional connectivity in humans using fMRI data.
Between 2011 and 2013, while working as a research assistant professor at the Center for Complex Systems and Brain Sciences at Florida Atlantic University, she initiated and led a summer study abroad program “Methods in Brain Computer Interfaces” in which she took a group of students to Tübingen Germany where she did her undergraduate studies, to teach them machine learning and techniques that the students then applied to brain imaging.
At the time, Dodel also conducted research in a variety of other interdisciplinary projects. Dodel analyzed data from behavioral and brain imaging (fMRI and EEG) experiments and developed observer independent measures of team performance, she studied the relationship between topology and dynamics in neural networks using mathematical modeling, developed a conceptual framework of brain dynamics in the presence of degeneracy, and characterized gene networks related to epilepsy (RNASeq data).
“While the human brain is arguably the most complex system known to humans, progress in biology and medicine necessitates the leveraging of recent advances in artificial intelligence and computer power to build expert systems that assist human researchers with automated literature analysis and hypothesis generation,” Dodel said.
The goal of the Germany trip was for the students to analyze brain signals to help patients that were completely paralyzed due to diseases like ALS – so-called locked-in patients – to communicate with the world through their brain signals. Machine learning, Dodel said, was just an extension of her interests.
“I became interested in machine learning more than ten years ago,” Dodel said. “It’s particularly useful in conjunction with the network paradigm – studying units and their interactions – which can explain a lot of phenomena.”
Before joining LIO, Dodel was extracting information from biomedical articles in an attempt to create a hierarchical visualization of articles and topics that would make it easier for researchers in the biomedical field to keep up with a constant deluge of information.
“Machine learning can be used for all types of industries, yet many industries use machine learning in ways that I would consider unethical,” Dodel said. “I’m thrilled that I’m contributing to an ethical cause — improving communication.”
The first priority of Dodel’s team is to enhance LIO’s self-improving glossary. The self-improving glossary uses numerous bodies of text to modify complicated terms and phrases so that they’re correctly translated every time.
“I like getting to the essence of a problem by analyzing it from an abstract level, identifying the most promising approach and implementing it so that a version one can be built,” Dodel said.