Applying Multivariate Segmentation Methods to Human Activity Recognition From Wearable Sensors' Data

Background

Time-resolved quantification of physical activity can contribute to both personalized medicine and epidemiological research studies, for example, managing and identifying triggers of asthma exacerbations. A growing number of reportedly accurate machine learning algorithms for human activity recognition (HAR) have been developed using data from wearable devices (eg, smartwatch and smartphone). However, many HAR algorithms depend on fixed-size sampling windows that may poorly adapt to real-world conditions in which activity bouts are of unequal duration. A small sliding window can produce noisy predictions under stable conditions, whereas a large sliding window may miss brief bursts of intense activity.

Objective

We aimed to create an HAR framework adapted to variable duration activity bouts by (1) detecting the change points of activity bouts in a multivariate time series and (2) predicting activity for each homogeneous window defined by these change points.

Methods

We applied standard fixed-width sliding windows (4-6 different sizes) or greedy Gaussian segmentation (GGS) to identify break points in filtered triaxial accelerometer and gyroscope data. After standard feature engineering, we applied an Xgboost model to predict physical activity within each window and then converted windowed predictions to instantaneous predictions to facilitate comparison across segmentation methods. We applied these methods in 2 datasets: the human activity recognition using smartphones (HARuS) dataset where a total of 30 adults performed activities of approximately equal duration (approximately 20 seconds each) while wearing a waist-worn smartphone, and the Biomedical REAl-Time Health Evaluation for Pediatric Asthma (BREATHE) dataset where a total of 14 children performed 6 activities for approximately 10 min each while wearing a smartwatch. To mimic a real-world scenario, we generated artificial unequal activity bout durations in the BREATHE data by randomly subdividing each activity bout into 10 segments and randomly concatenating the 60 activity bouts. Each dataset was divided into ~90% training and ~10% holdout testing.

Results

In the HARuS data, GGS produced the least noisy predictions of 6 physical activities and had the second highest accuracy rate of 91.06% (the highest accuracy rate was 91.79% for the sliding window of size 0.8 second). In the BREATHE data, GGS again produced the least noisy predictions and had the highest accuracy rate of 79.4% of predictions for 6 physical activities.

Conclusions

In a scenario with variable duration activity bouts, GGS multivariate segmentation produced smart-sized windows with more stable predictions and a higher accuracy rate than traditional fixed-size sliding window approaches. Overall, accuracy was good in both datasets but, as expected, it was slightly lower in the more real-world study using wrist-worn smartwatches in children (BREATHE) than in the more tightly controlled study using waist-worn smartphones in adults (HARuS). We implemented GGS in an offline setting, but it could be adapted for real-time prediction with streaming data.

CHILDHOOD RESPIRATORY MORBIDITY AND COOKING PRACTICES AMONG HOUSEHOLDS IN A PREDOMINANTLY RURAL AREA OF GHANA

Background:

Household air pollution is a leading risk factor for respiratory morbidity and mortality in developing countries where biomass fuel is mainly used for cooking.

Materials and Method:

A household cross-sectional survey was conducted in a predominantly rural area of Ghana in 2007 to determine the prevalence of respiratory symptoms and their associated risk factors. Household cooking practices were also assessed as part of the survey.

Results:

Household heads of twelve thousand, three hundred and thirty-three households were interviewed. Fifty-seven percent (7006/12333) of these households had at least one child less than five years of age. The prevalence of symptoms of acute lower respiratory infections (ALRI) was 13.7% (n= 957, 95% CI 12.8 – 15.5%). A majority (77.8%, 95% CI, 77.7 - 78.5%) of households used wood as their primary fuel. Majority of respondents who used wood as their primary fuel obtained them by gathering wood from their neighborhood (95.6%, 9177/9595) and used a 3-stone local stove for cooking (94.9%, 9101/9595). In a randomly selected subset of respondents, females were the persons who mostly gathered firewood from the fields (90.8%, 296/326) and did the cooking (94.8%, 384/406) for the household.

Conclusion:

Symptoms of ALRI reported by caregivers is high in the Kintampo area of Ghana where biomass fuel use is also high. There is the need to initiate interventions that use improved cook stoves and to test the health benefits of such interventions.