BATS 

Development of a Biosignal Analysis Toolkit System (BATS) and Pipeline for Polytrauma Research

Method: 

• Construction of a data analysis pipeline to assess a person’s psychological changes based on noninvasively collected behavioral, psychological, and neurological signals. 

• The analysis pipeline is an aggregation, enhancement, and integration of leading open-source toolkits focused on fNIRS and speech data that includes MNE-NIRS, MNE-Python, PyAudioAnalysis, and a set of infinite impulse response (IIR) filters; 

• Analysis pipeline isbe hosted in an AWS cloud and used by WPI researchers and labs.

Research Responsibilities: 

• Conducted in-depth research of over ten packages and tool kits; 

• Reviewed functional and legal documentation; 

• Interviewed other labs the use these tools; 

• Analyzed, when necessary, the actual code to confirm functionality; 

• Wrote scripts and code in Python and Java to improve and adjust functionality to meet project and platform needs;

• Assisted with integration into AWS platform ; 

• Collaborated with team to develop and present a poster to the leadership of multiple cross-discipline departments

This project was completed as the Master Qualifying Project (MQP), a graduation requirement for WPI.
The final report can be read here.

Abstract: 

Polytrauma is a serious health condition that arises following traumatic events. Research into new technologies for diagnosing polytrauma is essential to improving health outcomes for patients. Technology development in this domain would be aided by better research tools that allow for simultaneous analysis of traditional evaluations and novel biosignals. There are currently no analysis pipelines designed for polytrauma-specific research. Based on previous research and models, we developed a web application designed specifically for speech and neural data analysis relating to polytrauma research. It allows users to make a secure account, store and analyze study data and results. The user study showed a favorable response to our system and inspired ideas for future improvement to support various users.