• GROW

    students-celebrating
  • DISCOVER

    lab-students
  • CREATE

    synchrony-students
  • Homepage
  •   /  
  • Headline
  •   /  
  • Senior Design Journey 2021: Using Lung Scans and Algorithms to Diagnose COVID-19, Part 2

Senior Design Journey 2021: Using Lung Scans and Algorithms to Diagnose COVID-19, Part 2

Because of the work of the team’s neural net, researchers and diagnosticians get an almost instantaneous confirmation of if a patient has COVID-19 or not. (Courtesy of the team)

 

By: Eli Freund, Editorial Communications Manager, UConn School of Engineering 

This article is part of a multi-part series on engineering students, and their journey through Senior Design. Click here to read part 1 of this article series.

A group of five University of Connecticut computer science students are on the precipice of completing a Senior Design project that could not only aid researchers studying lung disease—but could be a boon for diagnosticians when it comes to COVID-19.

The team of Jamey Calabrese, Jay Chandran, Everett Han, Yuwen Jin, and Adam Veilleux are working on creating a portal where anyone could input x-rays and scans of diseased lungs using a machine learning concept called a neural net, and get an almost instantaneous confirmation of if a patient has COVID-19 or not.

For the team, their decision to go after building this diagnostic platform comes from a solid batch of research that suggests that using lung scans is one of the best ways to quickly diagnose COVID-19. According to a study done by radiologists at Louisiana State University Health Sciences Center, where they identified common characteristics and compared their diagnosis to a concurrent COVID-19 PCR test, they found that they were able to predict a positive test nearly 84 percent of the time when those common characteristics were used to make a diagnosis.

More specifically, according to one of the radiologists, they found that “the presence of patchy and/or confluent, band-like ground glass opacity or consolidation in a peripheral and mid-to-lower lung zone distribution on a chest radiograph is highly suggestive of SARS-CoV-2 infection and should be used in conjunction with clinical judgment to make a diagnosis.”

Initially, the team intended to build a platform that diagnoses several type of lung diseases, but according to Chandran, the scope of the project pivoted several times.

“Originally the project was planned to be a web service where a person could upload a picture, and then have the model classify the picture into a ‘Covid’ or ‘Non-Covid’ category. Simple enough. However, the project was then changed to be a web service where we actually train models for the users themselves, as the users would upload entire datasets to make this all possible.

Considering the fact that datasets can often be thousands of images long, and making a model is not a task that can be done the same way every time, this project quickly proved to be impossible, or at least not in the realm of possibility given our resources and knowledge. So, we shifted back to our original idea, and cut a lot of features that we were originally working on such as database hosting and a login page.”

But for Chandran and the team, while there were setbacks, there were also multiple triumphs, which included some members learning some key real-world skills.

“I think that working on this project has definitely helped me develop skills useful in a real job setting. My portion of the project dealt with integration and front-end development, allowing me to get experience working with GitHub and React. GitHub is at the center of most companies for keeping track of version control and React is one of the hottest packages for web development. It has been a difficult but rewarding experience working on this project, and I hope that the skills I have learned will be useful to me in the future,” Chandran said.

With any major project like this, the team also learned some important lessons in time management and group communication. With one of the team members overseas and everyone virtual, at times, coming together became a challenge—but not one they couldn’t overcome.

“Our team has definitely run into some communication and management problems. One of our members is overseas, making it difficult to schedule meetings due to the time difference. It was also very difficult to tell how much progress people were making before our regular meetings. School definitely played a role in our disorganization, seemingly with at least one person busy with homework or exams each week. This made it very difficult for us to manage our workflow, however we gradually learned to work around each other’s schedules,” Han said.

With Senior Design Demonstration Day looming on April 28, the team says they’ll most likely come down to the wire, but will be there with a working demo when the day comes.

“We will be ready by Design Day. We have a product working and just have to tidy up what we did and make it ready for showcase,” Veilleux said.

Senior Design Demonstration Day will be held virtually again this year, on April 28. For more information on the Senior Design program, please click here.