While often overlooked, itchiness is a symptom that provides utility in diagnosing a number of conditions, including lymphoma, liver failure, renal failure, and atopic dermatitis. When considering that excessive itch can also be just plain unbearable for the sufferer, it makes sense that capturing objective measurements of itching activity would be useful. The problem is that there just are not many convenient and accurate ways to capture this type of data.
A multidisciplinary team of researchers has just published a paper in Science Advances presenting a device that is designed to meet this currently unmet need. It is a soft, flexible, low-profile sensor that can capture vibration and motion signatures from the wearer and provide clinical-grade data quality. The non-invasive ADAM (ADvanced Acousto-Mechanic) sensor is completely wireless and can operate for seven days on a single charge. The tiny device is equipped with a Bluetooth Low Energy radio, a rechargeable battery, a millimeter-scale 1600 Hz three-axis accelerometer, and supporting hardware.
The hand-mounted device (📷: K. Chun et al.)
ADAM collects data indicative of scratching through a combination of motion and acousto-mechanic signals. This combination of sensing strategies allows the device to detect scratching that involves wrist movements, or more subtle vibrations of the fingers or finger tips generated when contacting a surface, without interference from ambient noise.
The team turned to a machine learning algorithm called a random forest classifier to determine which signal patterns correlate with itching, and which do not. This classifier was trained using data from the sensor paired with direct observations of camera recordings of patients with atopic dermatitis. The scratching data was recorded from fifteen various locations on the body, and six non-scratching activities (e.g. texting on a cellphone, waving a hand in the air) were also included.
Representative signals (📷: K. Chun et al.)
In a validation study, ADAM was found to detect itching activities with an overall accuracy of 99.0%, with a sensitivity of 84.3% and a specificity of 99.3%. While these results are very impressive, it must be noted that only eleven participants were included in the study due to the high amount of labor required to manually review the camera footage.
The researchers suggest that ADAM could be used to assist clinicians in diagnosing certain medical conditions, and would also have utility in objectively evaluating the effectiveness of anti-itch treatments.
Next up, they are testing out a new application of ADAM, by placing an additional sensor on the chest. By doing so, they hope to gain a better understanding of how itching impacts sleep quality.