Introduction
My project was tied closely to the data collected for the other studies underway in Dr. Solovey's lab. Using functional near-infrared spectroscopy (fNIRS), a type of neuroimaging technology, the lab was able to collect brain data to supplement observable measurement when conducting experiments. My job was to help build an interface to process, analyze, and display the collected brain data from the various studies in a clear, intuitive interface.
fNIRS Background
The technology that the lab uses to collect brain data is called functional near-infrared spectroscopy (fNIRS). A type of neuroimaging technology, fNIRS is portable, noninvasive, and provides relatively little background noise in its readings. Each piece of machinery is equipped with fiber-optic probes that send infrared light into the brain, at wavelengths of 690 and 830 nanometers. The light scatters around the brain, and then exits, and is measured by detectors on the fNIRS tool. However, some of the infrared light is absorbed by the oxygen in the blood in that area of the brain. From this, we can figure out the amount of oxygenated and deoxygenated blood in that part of the brain, which can be used as an estimate of the activity in that area of the brain.
fNIRS Application in the Lab
There are two primary experiments underway in Drexel's AIR Lab. The first deals with changes in cognitive state and driving. Because fNIRS is so portable and has relatively little background noise, the subject wearing it can perform tasks without compromising the brain data. Therefore, fNIRS is a convenient tool for measuring the cognitive state of participants driving or simulating driving. In this study, participants are presented with tasks to perform while driving, and changes in their brain activity are recorded. This is an important study, as it is critical to be able to understand the cognitive workload of drivers in order to ensure that the vehicle is driven safely.
The second study is looking into intelligent tutoring systems, with the goal of improving STEM education in particular. Through the use of fNIRS technology, the lab can better grasp whether a student is in fact learning the material. Currently, the only measurements that are used to determine this are number of questions gotten correct, time it takes to answer a question, whether the participant asked for a hint and how long they waited before asking for one, etc. By including brain data, the lab can gain a clearer image of how deeply the participant is learning, and whether or not they are attempting to "cheat" the system (by immediately asking for a hint without attempting the project, for example).
The second study is looking into intelligent tutoring systems, with the goal of improving STEM education in particular. Through the use of fNIRS technology, the lab can better grasp whether a student is in fact learning the material. Currently, the only measurements that are used to determine this are number of questions gotten correct, time it takes to answer a question, whether the participant asked for a hint and how long they waited before asking for one, etc. By including brain data, the lab can gain a clearer image of how deeply the participant is learning, and whether or not they are attempting to "cheat" the system (by immediately asking for a hint without attempting the project, for example).
The Dashboard Project
These are merely two examples of many current and potential applications of fNIRS technology. In order to make use of the brain data, however, whoever is conducting the study has to process, analyze, and interpret the fNIRS data. There is software that exists that can do this; however the best current tool is difficult to use, particularly for those who have no experience with it. It is not designed with usability in mind. The Dashboard Project - the project that I worked on for 7 weeks - intends to develop a new interface for processing and analyzing fNIRS data. This will be one that keeps software engineering practices in mind, and better serves not only our lab, but additionally anyone who wants to use it for similar purposes.
The basic database structure and Dashboard graphical user interface (GUI) had been constructed last summer by Paritosh, the lab's previous STAR student. My work would consist of researching similar types of tools to determine what aspects we wanted to incorporate, learning javafx and related tools (such as SceneBuilder), building a Launcher Application that prompted the user to connect to a database, and connecting that class to an existing class, RawFileParse, which parses the given data. More work had to be done structuring the database as well. At the end of my time in the lab, there was still a lot of work to be done. Most of my work had to do with getting the start-up processes of the application working - specifying a database, creating a project, parsing and loading data. The ensuing steps are where the data processing and analyzing will begin.
In the future, Dr. Solovey would also like to adapt the tool that we create to process fNIRS data in real-time. This means that while a participant is wearing the fNIRS device, data would constantly be streaming into the Dashboard and being processes as it was received. That is a goal further down the line, however. First, we must get a working version of the Dashboard without the real-time processing capabilities.
The basic database structure and Dashboard graphical user interface (GUI) had been constructed last summer by Paritosh, the lab's previous STAR student. My work would consist of researching similar types of tools to determine what aspects we wanted to incorporate, learning javafx and related tools (such as SceneBuilder), building a Launcher Application that prompted the user to connect to a database, and connecting that class to an existing class, RawFileParse, which parses the given data. More work had to be done structuring the database as well. At the end of my time in the lab, there was still a lot of work to be done. Most of my work had to do with getting the start-up processes of the application working - specifying a database, creating a project, parsing and loading data. The ensuing steps are where the data processing and analyzing will begin.
In the future, Dr. Solovey would also like to adapt the tool that we create to process fNIRS data in real-time. This means that while a participant is wearing the fNIRS device, data would constantly be streaming into the Dashboard and being processes as it was received. That is a goal further down the line, however. First, we must get a working version of the Dashboard without the real-time processing capabilities.
More Information
If you would like to know more about the Driving & fNIRS project, see the link below! It will direct you to the page on the AIR Lab website regarding this project, equipped with a project synopsis and related publications.
If you would like to know more about the Intelligent Tutoring Systems project, you can learn more from the button below! This will direct you to the page on the AIR Lab website related to the Tutoring System project. It too has a project synopsis and related publications, including a link to Shelby's website, which she created for her CREU internship.
Finally, if you would like to know more about the Dashboard project and my work in the lab, you can refer either to my weekly blog, outlining my time in the lab, or my final report. Links to both resources are provided below!
| |||