By Harry To
The Center of Excellence in Wireless and Information Technology at Stony Brook University
hosted its 5th annual Hack@CEWIT “hackathon” featuring student-made inventions, Feb. 26-28.
Usually this showcase takes place in person, but due to the COVID-19 pandemic this
year’s event was hosted online. In place of the usual format, the over-200 competitors
communicated through Zoom or Discord.
Satya Sharma, executive director of CEWIT, emphasized the abnormal circumstances weren’t
a problem.
“This year’s 5th annual Hack@CEWIT had over 200 registered undergrad and graduate hackers from across the U.S.,” he said.
“And though it was held virtually due to the pandemic, it did not diminish the quality
of projects submitted by these bright and motivated students. It’s opportunities like
this hackathon that builds confidence in their creativity and grows their entrepreneurial
spirit.”
According to Sharma, this year’s theme, Innovating Through the Pandemic, reminds people
that though there are sudden and unknown challenges, they can seize the opportunities
those challenges create and harvest ideas never before imagined.
Students Mohammad Elbadry, 23 (left) and Aaron Gregory, 23 (right). Photo from event
A standout project was R-AGI: Radiology Artificial General Intelligence, created by Stony Brook University graduate students Mohammed Elbadry, Joshua Leeman
and Aaron Gregory.
“According to a survey, radiologists only have about 3-4 seconds to look over an X-ray
and determine if there are any anomalies,” said Elbadry, a Ph.D. student with over
20-plus hackathons under his belt. “They don’t have much time, so if they had an AI
that could help them that would be very useful.”
The limited time for scanning X-rays may result in a higher frequency of errors or
discrepancies, with some studies citing an average 3% to 5% error rate, he said. That’s
about 40 million radiologist errors every year, mistakes that could potentially cost
hundreds of lives.
With the problem in mind, the team of three went to work to create AI that would offer
a solution — a program that automatically scans X-rays and detects anomalies. This
is something that could save not only time, but human lives.
By using an existing dataset of labelled X-rays, the team trained its AI to detect
the presence of pneumonia as well as its specific manifestation. The AI then labels
and informs the user of any further anomalies.
The SBU team ended up with an impressive showing, including Top-Tier Graduate Best
in Show and Best Healthcare Innovation.
Another award winning project was DarkWebSherlock, created by Andrew Zeoli, Colin Hamill, Donald Finlayson and Ian Costa from Johnson & Wales University, Providence, R.I.
The sale of personal information on the dark web, a hidden part of the internet accessible
through the TOR Browser, is a problem that has persisted for years, and DarkWebSherlock
aims to create a solution.
The program allows users to scan through online marketplaces on the dark web to see
if their data is up for sale anywhere.
This enables victims to be proactive by updating their passwords or changing their
credit card numbers to better secure their information.
Costa said the program will be an invaluable asset. “Searching for usernames on the
dark web is something our team does on a daily basis,” he said. “Our project will
save valuable time for investigators and with some extra work will become a staple
tool for dark web investigations.”
DarkWebSherlock won Top-Tier: Undergrad Best in Show.
Another award-winning project, Vaccine-Finder, aims to help speed up COVID-19 vaccine distribution for 65-year-old-plus vaccine
seekers.
The interface allows the elderly, also people with disabilities, to plug in their
zip codes and view the appointment availability of the COVID-19 vaccine.
Joshua Muckey started this project only recently, and it won Best Pandemic Innovation.
In all, the event hosted 15 projects, many of which showcased student ingenuity in
the face of the COVID-19 pandemic.
“This year is a reminder of why innovation is key to our success and our survival
as a region, as a state and as a society,” said Marc Alessi, a judge for the event,
CEO of SynchroPET and executive director of Tesla Science Center. “This weekend’s
hackathon at Stony Brook University’s CEWIT center is an example of bringing together
emerging innovators from very diverse backgrounds for the purpose of celebrating and
practicing innovation in its most raw form. This is essential to foster an environment
of innovation.”
All of the participants’ projects can be found online here.