Smartphones as Sleep Duration Sensors: Validation of the iSenseSleep Algorithm

TY - JOUR

T1 - Smartphones as Sleep Duration Sensors

T2 - Validation of the iSenseSleep Algorithm

AU - Ciman, Matteo

AU - Wac, Katarzyna

N1 - ©Matteo Ciman, Katarzyna Wac. Originally published in JMIR Mhealth and Uhealth (http://mhealth.jmir.org), 21.05.2019.

PY - 2019/5/21

Y1 - 2019/5/21

N2 - BACKGROUND: Smartphones are becoming increasingly ubiquitous every day; they are becoming more assimilated into our everyday life, being the last thing used before going to sleep and the first one after waking up. This strong correlation between our lifestyle choices and smartphone interaction patterns enables us to use them as sensors for sleep duration assessment to understand individuals' lifestyle and sleep patterns.OBJECTIVES: The objective of this study was to estimate sleep duration based on the analysis of the users' ON-OFF interaction with their smartphone alone using the iSenseSleep algorithm.METHODS: We used smartwatch sleep assessment data as the ground truth. Results were acquired with 14 different subjects collecting smartwatch and smartphone interaction data for up to 6 months each.RESULTS: Results showed that based on the smartphone ON-OFF patterns, individual's sleep duration can be estimated with an average error of 7% (24/343) [SD 4% (17/343)] min of the total duration), enabling an estimate of sleep start and wake-up times as well as sleep deprivation patterns.CONCLUSIONS: It is possible to estimate sleep duration patterns using only data related to smartphone screen interaction.

AB - BACKGROUND: Smartphones are becoming increasingly ubiquitous every day; they are becoming more assimilated into our everyday life, being the last thing used before going to sleep and the first one after waking up. This strong correlation between our lifestyle choices and smartphone interaction patterns enables us to use them as sensors for sleep duration assessment to understand individuals' lifestyle and sleep patterns.OBJECTIVES: The objective of this study was to estimate sleep duration based on the analysis of the users' ON-OFF interaction with their smartphone alone using the iSenseSleep algorithm.METHODS: We used smartwatch sleep assessment data as the ground truth. Results were acquired with 14 different subjects collecting smartwatch and smartphone interaction data for up to 6 months each.RESULTS: Results showed that based on the smartphone ON-OFF patterns, individual's sleep duration can be estimated with an average error of 7% (24/343) [SD 4% (17/343)] min of the total duration), enabling an estimate of sleep start and wake-up times as well as sleep deprivation patterns.CONCLUSIONS: It is possible to estimate sleep duration patterns using only data related to smartphone screen interaction.

U2 - 10.2196/11930

DO - 10.2196/11930

M3 - Journal article

C2 - 31115341

SN - 2291-5222

VL - 7

SP - e11930

JO - JMIR mHealth and uHealth

JF - JMIR mHealth and uHealth

IS - 5

ER -