Body weight-normalized Energy Expenditure estimation using combined activity and allometric scaling clustering

Abstract

Wearable sensors have great potential for accurate estimation of Energy Expenditure (EE) in daily life. Advances in wearable technology (miniaturization, lower costs), and machine learning techniques as well as recently developed self-monitoring movements, such as the Quantified Self, are facilitating mass adoption. However, EE estimations are affected by a person's body weight (BW). BW is a confounding variable preventing meaningful individual and group comparisons. In this paper we present a machine learning approach for BW normalization and activities clustering. In our approach to activity-specific EE modeling, we adopt a genetic algorithm-based clustering scheme, not only based on accelerometer (ACC) features, but also on allometric coefficients derived from 19 subjects performing a wide set of lifestyle and gym activities. We show that our approach supports making comparisons between individuals performing the same activities independently of BW, while maintaining accuracy in the EE estimate.

A Survey on Energy Expenditure Estimation Using Wearable Devices

“…Although they eventually settled their dispute 8 , this may have led to the end of Jawbone 9 . Apple is also selling millions of smartwatches 10 and protecting his knowledge using patents [116]. To win the commercial battle, current leaders (Fitbit and Apple Watch) are trying to seek US Food and Drug Administration (FDA) approval for their products 11 12 .…”

“…Lawn & Garden (08) Sports (15) Conditioning Exercise (02) Miscellaneous (09) Transportation ( 16) Dancing (03) Music Playing (10) Walking (17) Fishing & Hunting (04) Occupation (11) Water Activities (18) Home Activities (05)…”

Estimation of Body Size by Combining Genetic Algorithm with Human Body Model

“…It can easily solve the resource allocation problem with certain constraints, as shown in Figure 3. The improved GA [12] divides the initial population into multiple subpopulations, and assigns a sub-target to each subpopulation by individual selection techniques such as parallel selection. The integration of a single optimal solution to meet the overall goal constraints.…”

Combining wearable accelerometer and physiological data for activity and energy expenditure estimation

“…Normalization procedures do exist (e.g. kcal/kg), but do not take into account that EE during different activities is a↵ected di↵erently by body weight [3]. Pearson's correlation coe cient is used to quantify the predictive power of di↵erent signals in estimating EE, without being a↵ected by inter-individual di↵erences (average between subjects).…”