Modelling for wearable Healthcare Product

Abstract

In the human body, all cells have their own clock called circadian clock, which decides the time to energies, rest or perform tasks. Aligning our personal schedule with the circadian clock helps optimize our energy and improve our immune system. An important step of aligning the clock is sleeping on time. A good sleep at night is essential to recharge the body and enables the body to recover from the strains in the day and wake up refreshed and energized. Moreover, studies have found that sleep has a very integral role in very vital processes of humans like regulation of emotion, reinforcing the memories in the mind which in turn improves learning. However, many people suffer from sleep deficiency due to a myriad of factors like sleep disorders, medical conditions, mental health, poor lifestyle etc. The effects can be detrimental and result in chronic health conditions. Though the medical labs offer precise ways of assessing sleep, polysomnography, it is often laborious and expensive. Hence researchers are pursuing alternative methods to track sleep. The popular alternatives include but are not limited to the various embedded devices that monitor day-to-day activity and sleep using various sensors, viz. altimetry sensor, piezoelectric sensors, heart rate monitors, and temperature monitors etc. However, every person has a different lifestyle and sleeping patterns, which cannot be catered to with a generic solution/algorithm. In such scenarios, artificial intelligence solutions can be leveraged to understand the behavior of each user and customize the analysis that best fits their routine and recommend life-improving suggestions. In this project, we model actigraphy data to help consumers monitor their day-to-day sleep and summaries their daily scores. The results show that we successfully generated the sleep score.