Raquel Romano Selected As Recipient of Alvarez Post-Doctoral Fellowship
May 10, 2004
By Jon Bashor
Raquel Romano, who joined the Imaging and Informatics Group in the Computational Research Division (CRD) in January 2004, has been selected as the next Luis W. Alvarez Post-Doctoral Fellow in Computational Science.
The fellowship, sponsored by CRD and the NERSC Center Division, was established to encourage the development and application of tools to advance scientific research. Here on a one-year (extendable) post-doc research appointment, Romano’s research interests are in the area of applying machine learning and geometric tools from computer vision toward the discovery of patterns, events, features and anomalies in scientific data.
At Berkeley Lab, she is working to develop computer vision and image processing tools to help biologists in the Life Sciences Division better understand why some cells become cancerous. “We can trace the process of DNA damage by looking at how proteins are expressed in the cell,” Romano said.
Typically, biologists “eyeball” or visually scan images of cells treated with radiation, looking for colors, patterns, and other features indicating protein expression. The process is both time-consuming and subject to human error.
“The field of computer vision has a whole body of work to offer, allowing scientists to analyze the data quickly and consistently,” Romano said. “In fact, novel computational routines can allow biologists to quantify properties in ways that we, as humans, don’t know how to do. We can automatically quantify meaningful information, such as how much, when, and where in the cell a protein is expressed.”
Not only do the computational tools look at the images in a more detailed way, she said, but they also allow massive numbers of images to be analyzed quickly, with the important features identified, defined, and classified in a consistent manner.
After earning her bachelor’s degree in mathematics from Harvard, Romano earned both her master’s and Ph.D. in computer science at the Massachusetts Institute of Technology. During these years, she also worked as a research assistant in an impressive list of laboratories, including MIT Artificial Intelligence Laboratory, French National Institute for Research in Computer Science and Control (INRIA), IBM T.J. Watson Research Center, and AT&T Bell Laboratories.
She also found time during her doctorate studies to organize a grad student hockey team. “My advisor was Canadian, so he was behind me 100 percent,” Romano said with a smile. While practicing five nights a week on the ice provided a welcome break from her studies, she said that getting away from staring at her monitor often gives a new perspective.
“One time, I was trying to understand how you could interpret motion from a video, and as I watched my shadow on the ice, I realized that shadows can be another cue for motion,” said Romano, who now plays in a Bay Area women’s hockey league. “I didn’t end up pursuing that approach, but new ideas often come to me when I’m playing sports or going for a run.”
Finding new ways to look at and interpret things is a recurring theme for Romano. Her graduate work focused on novel geometric methods for recovering 3D camera positions from collections of images and on statistical methods for distinguishing multiple motions from video sequences. When she finished her degree in computer science, she didn’t want to continue working only in that field. “I wanted to take a step back before I said my research. Computers are tools, and what’s fascinating to me on an intellectual level is the power of computation for problem-solving,” she said. “On a personal level, it really matters to me what this science is being applied to.”
She began looking for opportunities and, through the grapevine, heard about the post-doc position in the Imaging and Informatics Group. “I wanted to use the valuable aspects of research in my area and apply them to important problems in another field,” she said. “I think where two disciplines intersect is where the most exciting research is found.”
Coming to Berkeley was also a homecoming of sorts. Her mother was born and raised here as her grandfather, Marvin D. Martin, was an engineer here in the early days of the Lab, then moved to Livermore when that lab was established in 1952.
“He gave me my first computer and taught me BASIC,” she said, “but he wasn’t able to tell us much about his classified work at the Livermore or Oak Ridge labs.”
She had also spent a summer in the Bay Area while working on her Ph.D. Her brother, who was living in Berkeley, urged her to come out and do something different. She agreed and spent a summer teaching middle school math to disadvantaged students in Marin City and the Canal area of San Rafael. “They would ask things like ‘Why do we care if something is a quadratic equation?’ and I really had to stop and think about the answer, about how shapes in nature can be expressed mathematically.”
Adapting to new situations is something Romano is accustomed to. Although born in the U.S., her family lived in Mexico until she was six, when they moved to Ohio. There, since she spoke Spanish, many people assumed she was from Spain. “It was quite a culture shock,” she said. While Berkeley can prove overwhelming for some newcomers, Romano is quickly immersing herself in LBNL.
“I find the Lab fascinating,” she said. “There are so many people working on different kinds of research, and being able to chat with them and hear about their work is very enriching. I hear ideas that get me thinking about new ways of looking at problems in my own research. In a sense, I see myself as a perpetual student.”
About Computing Sciences at Berkeley Lab
High performance computing plays a critical role in scientific discovery. Researchers increasingly rely on advances in computer science, mathematics, computational science, data science, and large-scale computing and networking to increase our understanding of ourselves, our planet, and our universe. Berkeley Lab’s Computing Sciences Area researches, develops, and deploys new foundations, tools, and technologies to meet these needs and to advance research across a broad range of scientific disciplines.