News

Facebook posts can predict and identify health conditions: Study

Views: 5.9K
Facebook posts

A study revealed by the researchers from University of Pennsylvania and Stony Brook University stated that Facebook posts can now predict and identify health conditions like diabetes, anxiety, depression and psychosis. They also proposed that language through web-based media posts with patient consent could be physical indicators of health.

They have analysed the entire Facebook post history of nearly 1,000 patients using an automated data collection technique.

To support their analysis, the researchers then built three models to examine their analytical power for the patients. With one model only analysing the Facebook post language, the second model was to analyse the demographics such as age and sex. Combining the first two models, they built the third model.

Getting the models into action, researchers looked into 21 different health conditions, which was better predictable from Facebook than by demographic analysis.

Commenting on the findings, Raina Merchant, an associate professor at University of Pennsylvania said, “This work is early, but our hope is that the insights gleaned from these posts could be used to better inform patients and providers about their health.”

“As social media posts are often about someone’s lifestyle choices and experiences or how they’re feeling, this information could provide additional information about disease management and exacerbation,” Merchant said.

Furthermore, the researchers laid down that signs of “drink” and “bottle” were shown to be more predictive of alcohol abuse. Also, it was also revealed that people who mentioned religious terms like ‘God’ or ‘pray’ in their Facebook posts were likely to be diabetic than others who didn’t use these terms.

Another researcher, Andrew Schwartz, an assistant professor at Stony Brook University said, “Our digital language captures powerful aspects of our lives that are likely quite different from what is captured through traditional medical data,” said “Many studies have now shown a link between language patterns and specific disease, such as language predictive of depression or language that gives insights into whether someone is living with cancer.”

“However, by looking across many medical conditions, we get a view of how conditions relate to each other, which can enable new applications of AI for medicine,” he said.

Shedding light on the findings, ,Merchant said, “For instance, if someone is trying to lose weight and needs help understanding their food choices and exercise regimens, having a healthcare provider review their social media record might give them more insight into their usual patterns in order to help improve them.”

A year ago, many members of this research team demonstrated that analysing Facebook posts could foresee an analysis of depression three months earlier than a diagnosis in the clinic.

Latest News

To Top