Using AI to Quantify Skin Scratching Behavior

Leo Pharma, a Danish pharmaceutical company that focuses on treatment for skin conditions, is partnering with the Massachusetts Institute of Technology’s (MIT) Computer Science and Artificial Intelligence Laboratory (CSAIL) to quantify patients’ scratching habits through using a motion sensing device developed by CSAIL.

People with eczema, dry skin, rashes, psoriasis, and scabies have a tendency to compulsively scratch. However, scratching usually only worsens the condition and can provoke inflammation. In order to determine the severity of the scratching, dermatologists rely on the patient’s self-evaluation of their scratching habits. Yet, this kind of subjective assessment isn’t always reliable.

CSAIL’s Emerald system, a wireless technology that uses radio waves to identify body movement and vital signs, could be the first device of its kind to accurately measure itch behavior. Emerald operates like a wifi router, and can track a person’s movement even if they are not in the same room as the device.

Emerald’s developers, professors Dina Katabi, PhD, and Fadel Adib, PhD, and research assistant Zachary Kabelac, PhD, have already received glowing recognition for the device and were invited to present the technology for President Obama in 2015. It has been touted as a safer and more effective alternative to wearable sensors. The more information Emerald absorbs, the smarter it becomes at identifying movements and behaviors like scratching.

“After seeing Dina present data on the Emerald system, it immediately spurred the idea that this could be used to help improve clinical trial design in dermatology,” said Troels Marstand, Chief Data Scientist at LEO Science and Tech Hub, Leo Pharma’s R&D unit in a press release. “This project shows that innovation in pharma can come from unexpected places and reinforces the need to look beyond your own area of expertise for potential solutions.”