Neural network estimation of eardrum temperature using six sensors integrated on a wristwatch-sized device

A novel system for the estimation of eardrum temperature, utilizing neural networks, based on the data acquired from six sensors integrated on a wristwatch-type wearable device has been successfully demonstrated. The conventional estimation method, which uses a heat balance model of the body, cannot be applied as it needs parameters, such as the thermal index of the body and the information on clothes, which cannot be measured by the wristwatch-type device. We have introduced sensors that measure environmental quantities as well as vital signals to experimentally acquire sensing data from various places. To improve the estimation accuracy, the time series data were inputted to a neural network, and the system was optimized by comparing the estimation accuracy while varying the measurement time and interval of each sensor. By applying an image processing method, such as Max pooling, the standard deviation of error and the maximum error between the measured and estimated eardrum temperatures were 0.08 °C and 0.16 °C, respectively.