UMass scientist devising software tools to analyze mobile health data
AMHERST — Computer scientist Benjamin Marlin of the University of Massachusetts recently received a five-year, $536,527 National Science Foundation award to develop machine learning-based tools for analyzing complex, large-scale clinical and mobile health data.
Marlin’s project is designed to help health researchers handle what he calls a data revolution. “Electronic health records are seeing wide adoption across the United States and we’re starting to see the emergence of large stores of complex clinical data as a result,” Marlin says. “There’s significant interest in leveraging these data to enhance all kinds of clinical decision support tools with the hope that they can ultimately improve quality of care.”
Marlin’s research will also explore ways to analyze data from emerging wearable sensor systems that collect large volumes of continuous physiological measurements like respiration and electrocardiogram signals.
“Developing models and algorithms that can accurately and reliably detect activities like smoking from wearable sensor data has tremendous potential for use in behavioral science research as well as continuous health monitoring,” he notes.
The challenge with analyzing data from these sources is that they exhibit a number of complicating factors such as sparse and irregular sampling, incompleteness and noise.
“We’re not dealing with nice, clean data in these areas,” Marlin says. “The data are noisy, parts are missing due to sensors disconnecting or clinicians not recording measurements. The goal of this work is to design new machine learning-based data analysis tools that are significantly more robust and accurate.”