Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2019 Jul:2019:477-483.
doi: 10.1109/compsac.2019.10252. Epub 2019 Jul 9.

Activity Segmentation Using Wearable Sensors for DVT/PE Risk Detection

Austin Gentry, William M Mongan  1 , Brent Lee  2 , Owen Montgomery  3 , Kapil R Dandekar  2
Affiliations

Activity Segmentation Using Wearable Sensors for DVT/PE Risk Detection

Austin Gentry et al. Proc COMPSAC. 2019 Jul.

Abstract

Using a wearable electromyography (EMG) and an accelerometer sensor, classification of subject activity state (i.e., walking, sitting, standing, or ankle circles) enables detection of prolonged "negative" activity states in which the calf muscles do not facilitate blood flow return via the deep veins of the leg. By employing machine learning classification on a multi-sensor wearable device, we are able to classify human subject state between "positive" and "negative" activities, and among each activity state, with greater than 95% accuracy. Some negative activity states cannot be accurately discriminated due to their similar presentation from an accelerometer (i.e., standing vs. sitting); however, it is desirable to separate these states to better inform the risk of developing a Deep Vein Thrombosis (DVT). Augmentation with a wearable EMG sensor improves separability of these activities by 30%.

Keywords: Biomedical Computing; Classification Algorithms; Wearable Sensors.

PubMed Disclaimer

Figures

Figure 1:
Figure 1:
Image used to guide placement of the EMG sensor. Anatomic variation can result in slight variations in the ideal placement of the sensor which caused differences in the magnitude of the resulting signal. Feature selection was used to reduce the effects of this variation.
Figure 2:
Figure 2:
Accelerometer data for a trial showing the effect of the data collection methodology. Excluding transition periods between activities reduced the noise level and improved classifier performance.
Figure 3:
Figure 3:
En for a trial, showing the activity labels. It is clear from the raw data the the amount of calf muscle activity varies drastically between certain activities but the EMG data alone is not enough to separate them with high granularity.
Figure 4:
Figure 4:
Confusion matrices for classifiers based on the accelerometer alone (left), the EMG sensor alone (center), and combined (right)
See this image and copyright information in PMC

References

    1. “Deep Vein Thrombosis (DVT)— Gateway to Health Communication — CDC,” https://www.cdc.gov/healthcommunication/toolstemplates/entertainmented/t....
    1. Beckman MG, Hooper WC, Critchley SE, and Ortel TL, “Venous Thromboembolism: A Public Health Concern,” American Journal of Preventive Medicine, vol. 38, no. 4, pp. S495–S501, 2010. - PubMed
    1. Tarbox AK and Swaroop M, “Pulmonary Embolism,” International Journal of Critical Illness and Injury Science, vol. 3, no. 1, pp. 69–72, January 2013. - PMC - PubMed
    1. Eichinger S, Heinze G, Jandeck LM, and Kyrle PA, “Risk Assessment of Recurrence in Patients With Unprovoked Deep Vein Thrombosis or Pulmonary Embolism,” Circulation, vol. 121, no. 14, 2010. - PubMed
    1. Hirsh J, Hull RD, and Raskob GE, “Clinical Features and Diagnosis of Venous Thrombosis,” Journal of the American College of Cardiology, vol. 8, no. 6, pp. 114B–127B, 12 1986. - PubMed

LinkOut - more resources