Measurement Concordance of Body Temperature Between YONO Earbud and Oral Thermometer

YONO earbud is a personal wearable thermometer that measures in-ear ambient temperature during sleep. This report presents the analysis of body temperature (BT) collected from 3 females, and 2 males, test users age 24 to 40, over 6-10 consecutive times. The analysis found a very high concordance of measured BT by YONO earbud and traditional oral thermometers in both cross-sectional correlation and longitudinal patterns.

 

Test User 1 (Male) and User 2 (Female):

As marked in the graph, we used CVS oral thermometer data. 

Frequency: every 5 minutes
Number of data points: 8
Room temperature: 77.58F

Test User 3 (Male) and User 4 (Female):

As marked in the graph, we used different oral thermometer data.
Frequency: every 10 minutes
Number of data points: 6
Room temperature: 77.58F

Test User 5 (Female):

Frequency: every 5 minutes
Number of data points: 10
Room temperature: 77.58F

All graphs, above, from the 5 test users, clearly indicate the high concordance of body temperature values measured by YONO thermometer and a traditional oral thermometer. Such graphic visualization may be quantified precisely by numerical statistics given as follows. In the following table, the sample mean and standard deviation of the paired difference of BT, over multiple times, is listed. It is evident that the BT from YONO earbud exhibits a steady, lower measure, and such difference, between two devices, is virtually constant. Since we are looking for the change of body temperature in each cycle, it’s not necessary to calibrate YONO data to match with traditional, under-tongue temperature.

To quantify the agreement between YONO thermometer and a traditional oral thermometer, the Pearson Correlation Coefficient was calculated for the cross-sectional dependence between the body temperature values measured from YONO thermometer and a traditional oral thermometer. The estimated coefficient was 0.93, which indicates a very strong correlation between the two devices. Furthermore, to examine the concordant pattern of longitudinal BT curves measured by the two devices, we applied the Distance Correlation Coefficient [Székely. 2009], which was estimated as 0.90 by the data from the test users. Such a high distance correlation coefficient indicates that the two devices provided similar temporal patterns of BT measurements, which is essentially the key feature that allows us to use YONO to predict the timing of ovulation, as done by using an oral thermometer. 

In conclusion, in this preliminary data analysis, we have demonstrated that YONO thermometer, and a traditional oral thermometer, exhibited a very high concordance measured BT by YONO earbud and traditional oral thermometer in both cross-sectional correlation and longitudinal pattern. Thus, for measuring Basal Body Temperature data (BBT), YONO is a reliable surrogate for the oral thermometer.

Unlike traditional thermometers, which only collect one data point over a complicated process, YONO collects about 100 data points of continuous body temperature during sleep. The amount of data makes the fertility algorithm more accurate and can also provide insights of sleeping quality.

References: Székely, G. J., Rizzo, M. L. (2009). "Brownian distance covariance", Annals of Applied Statistics, 3/4 1233- 1303