Circadian rhythm estimation by core body temperature filtered with simultaneously recorded physiological data

In field measurements, monitoring of core body temperature is influenced by physical activities; therefore, the estimation of circadian rhythm from the data may not be exact. The purpose of this study is to design a core body temperature filter in order to reduce artifacts induced by physical activities using simultaneously recorded physiological data such as heart rate data. The effects of physical activities on core body temperature and heart rate are assessed through three experiments. Based on the above knowledge, a core body temperature filter was designed. The filter removes part of rectal temperature data as artifact when heart rate rises above a predetermined threshold. As a result, most of the spike-like noise was removed and the filtered temperature data showed sinusoidal variation more than the unfiltered data. The mesor of the estimated rhythm significantly decreased. This filtering method can provide more precise information about circadian rhythm, especially in field measurements.