May 17, 2020

Will Twitter Predict a Flu Outbreak Before the CDC?

3 min
There are over 241 million monthly active users on Twitter.
As we enter the final weeks of the year, looking back we can note that this year had various devastating health outbreaks around the world. Ebola in 201...

As we enter the final weeks of the year, looking back we can note that this year had various devastating health outbreaks around the world. Ebola in 2014 remains to be the largest outbreak in history and the first in West Africa, Madagascar is currently under surveillance due to an outbreak of the Black Death, and China has reported multiple cases of human infection with avian influenza.

As the colder weather settles in, there is something else that comes with the holiday cheer: the flu. It’s not possible to predict what this flu season will be like, but if it’s anything like last year – with more than 105 flu-related deaths in children alone being reported – it could be epidemic.

Which brings us to our question, can Twitter track the flu and predict outbreaks faster than the Centers for Disease Control and Prevention (CDC)?

A Difference in Tracking Methods

Traditional flu tracking performed by the CDC relies on outpatient reporting and virological test results supplied by laboratories nationwide that confirms an outbreak within two weeks after they begin. The CDC does not track all cases, however.

Instead of enduring this labor-intensive and time-consuming approach, researchers can capture comments from people with the flu who are sending out status messages, providing daily reports.

Twitter agrees.

The social media site unveiled a new grant program earlier this year that will allow outside researchers to mine its stockpile of tweets, and Johns Hopkins is one example of an institute taking advantage of this form of flu tracking.

A team from Johns Hopkins and George Washington universities conducted a study to track flu-related tweets from New York City. The team concluded that Twitter data can accurately gauge the spread of flu at the local level, too.

Citing data from the 2012-2013 U.S. flu season, the research team reported on results they obtained by sifting through billions of tweets to identify flu infections and where flu patients were located.

“We found that we could do just as well in predicting flu trends in New York City as we did nationally,” said Mark Dredze, an assistant research professor of computer science at Johns Hopkins who supervised the research. “That’s critical because decisions about what to do during a flu epidemic are largely made at the local level.”

Success Found in Numbers

During last year’s severe flu season, the team members compared their national Twitter flu findings with data that the CDC collected from health care providers. They also compared their results with flu cases compiled by the New York City Department of Health and Mental Hygiene.

“Not only did our results track trends on the national level, but they also did so on the local level,” said David A. Broniatowski, lead author of the study. “It gives our system validity. It shows that we’re measuring what we say we’re measuring, that we’re tracking very useful information. And that localized data is valuable because the flu activity in, say, Boise, Idaho, may be quite different from the national flu trends.”

Twitter’s Data Grants program will give scholars access to its public and historical data for use in garnering helpful information on various topics. Techniques used to track flu trends via Twitter might also be applied to the study of subjects such as HIV-incidence and drug-related behaviors.

“The exciting results we’ve come up with so far bring up new questions that will require additional data that the Twitter grant program may enable us to work with,” said a graduate student on the team. “The more experiments we do with Twitter posts, the more proof I see that this is a great idea.” 

Share article

Jun 18, 2021

Skin Analytics wins NHSX award for AI skin cancer tool 

2 min
Skin Analytics uses AI to detect skin cancer and will be deployed across the NHS to ease patient backlogs

An artificial intelligence-driven tool that identifies skin cancers has received an award from NHSX, the NHS England and Department of Health and Social Care's initiative to bring technology into the UK's national health system. 

NHSX has granted the Artificial Intelligence in Health and Care Award to DERM, an AI solution that can identify 11 types of skin lesion. 

Developed by Skin Analytics, DERM analyses images of skin lesions using algorithms. Within primary care, Skin Analytics will be used as an additional tool to help doctors with their decision making. 

In secondary care, it enables AI telehealth hubs to support dermatologists with triage, directing patients to the right next step. This will help speed up diagnosis, and patients with benign skin lesions can be identified earlier, redirecting them away from dermatology departments that are at full capacity due to the COVID-19 backlog. 

Cancer Research has called the impact of the pandemic on cancer services "devastating", with a 42% drop in the number of people starting cancer treatment after screening. 

DERM is already in use at University Hospitals Birmingham and Mid and South Essex Health & Care Partnership, where it has led to a significant reduction in unnecessary referrals to hospital.

Now NHSX have granted it the Phase 4 AI in Health and Care Award, making DERM available to clinicians across the country. Overall this award makes £140 million available over four years to accelerate the use of artificial intelligence technologies which meet the aims of the NHS Long Term Plan.

Dr Lucy Thomas, Consultant Dermatologist at Chelsea & Westminster Hospital, said: “Skin Analytics’ receipt of this award is great news for the NHS and dermatology departments. It will allow us to gather real-world data to demonstrate the benefits of AI on patient pathways and workforce challenges. 

"Like many services, dermatology has severe backlogs due to the COVID-19 pandemic. This award couldn't have come at a better time to aid recovery and give us more time with the patients most in need of our help.”

Share article