Provost’s MQP Award–winning project

See how WPI students put theory into practice through projects.

The Big Idea:

Scientists utilize Functional Near-Infrared Spectroscopy (fNIRS) to discover how much of the brain is activated during certain tasks. For this Provost’s MQP Award–winning project, fNIRS data was studied to detect issues such as broken sensors or head motion, which must be factored into research conclusions. Algorithms were developed that can accurately predict motion artifacts (head movement) in fNIRS brain signals; and a software platform was created for users to easily apply these algorithms to their own data.


Petra Kumi

Advisors: Rodica Neamtu, Computer Science. Erin Solovey, Computer Science, Suzanne Weekes, Mathematical Sciences

How it works:

•  A set of labeled anomalies was created to determine the accuracy of methods

•  A measure of accuracy was determined

•  A previously published anomaly detection method was investigated

•  Three new anomaly detectors were created

•  Algorithms were gathered into a toolkit for researchers

What’s next:

To make these methods available to others, a digital anomaly detection platform was created where users can begin anomaly detection jobs, view jobs previously submitted, and visually analyze them.


Petra Kumi ’20 and Ally Salvino ’22 worked with associate dean Suzanne Weekes and the Office of Undergraduate Research to create the Virtual Undergraduate Research Showcase D2020.  Since COVID-19 shut down the physical campus in the spring, preventing sharing ideas in person, undergraduates from virtually all majors submitted slide and video presentations of their
MQPs and other research to this site.

Visit to witness Lehr und Kunst (Theory and Practice) unfold before you.

Click on this switch to toggle between day and night modes.