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
. 2025 Jul 9;25(14):4262.
doi: 10.3390/s25144262.

Investigating Performance of an Embedded Machine Learning Solution for Classifying Postural Behaviors

Bruno Andò  1 , Salvatore Baglio  1 , Mattia Manenti  1 , Valeria Finocchiaro  1 , Vincenzo Marletta  2 , Sreeraman Rajan  3 , Ebrahim Ali Nehary  3 , Valeria Dibilio  4 , Mario Zappia  4 , Giovanni Mostile  4   5
Affiliations

Investigating Performance of an Embedded Machine Learning Solution for Classifying Postural Behaviors

Bruno Andò et al. Sensors (Basel). .

Abstract

Postural instability is one of the main critical aspects to be monitored in the case of degenerative diseases, and is also a predictor of potential falls. This paper presents a multi-layer perceptron approach for the classification of four different classes of postural behaviors that is implemented by an embedded sensing architecture. The robustness of the methodology against noisy data and the effects of using different sets of classification features have been investigated. In the case of noisy input data, a reliability index of almost 100% has been obtained, with a negligible drop (less than 5%) being shown for the whole range of noise levels that was investigated. Such an achievement substantiates the better robustness of this approach with respect to threshold-based algorithms, which have been also considered for the sake of comparison.

Keywords: experimental assessment; inertial sensor; multi-layer perceptron; noise robustness; postural sway classification.

PubMed Disclaimer

Conflict of interest statement

Author Vincenzo Marletta was employed by the company STMicroelectronics S.r.l. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Figures

Figure 2
Figure 2
(a) Data flow for features generation. Examples of features are reported in the stabilogram; (b) the architecture adopted to reproduce postural sway behaviors.
Figure 1
Figure 1
Main steps of the methodology and results presented in this work.
Figure 3
Figure 3
The dataset and its distribution among different cases and training/test subsets. The overall number of patterns for each class is: 331 ST, 340 ML, 334 AP, and 333 UNST.
Figure 4
Figure 4
The proposed MLP architecture, in the case where all four features are used as the model input.
Figure 5
Figure 5
Performance of threshold and MLP algorithms as a function of noise levels in the input data, for both training and test subsets using different features.
Figure 5
Figure 5
Performance of threshold and MLP algorithms as a function of noise levels in the input data, for both training and test subsets using different features.
Figure 6
Figure 6
Results obtained by the experimental test to assess features estimated by the sensor node (+) against features estimated in MATLAB® (o).
Figure 7
Figure 7
Results of a real-time test consisting of a sequence of postural dynamics. (a) Classes estimated by the sensor node (+) and predicted by the same algorithm running in MATLAB® (o); (b) RI estimated by the sensor node (+) and by the MATLAB® routine (o).
See this image and copyright information in PMC

References

    1. Baig M.M., Afifi S., GholamHosseini H., Mirza F. A Systematic Review of Wearable Sensors and IoT-Based Monitoring Applications for Older Adults—A Focus on Ageing Population and Independent Living. J. Med. Syst. 2019;43:233. doi: 10.1007/s10916-019-1365-7. - DOI - PubMed
    1. Ando B. Instrumentation Notes—Sensors That Provide Security for People with Depressed Receptors. IEEE Instrum. Meas. Mag. 2006;9:56–61. doi: 10.1109/mim.2006.1634992. - DOI
    1. Chaccour K., Darazi R., El Hassani A.H., Andres E. From Fall Detection to Fall Prevention: A Generic Classification of Fall-Related Systems. IEEE Sens. J. 2017;17:812–822. doi: 10.1109/jsen.2016.2628099. - DOI
    1. Nicoletti A., Mostile G., Stocchi F., Abbruzzese G., Ceravolo R., Cortelli P., D’Amelio M., De Pandis M.F., Fabbrini G., Pacchetti C., et al. Factors Influencing Psychological Well-Being in Patients with Parkinson’s Disease. PLoS ONE. 2017;12:e0189682. doi: 10.1371/journal.pone.0189682. - DOI - PMC - PubMed
    1. Contrafatto D., Mostile G., Nicoletti A., Raciti L., Luca A., Dibilio V., Lanzafame S., Distefano A., Drago F., Zappia M. Single Photon Emission Computed Tomography Striatal Asymmetry Index May Predict Dopaminergic Responsiveness in Parkinson Disease. Clin. Neuropharmacol. 2011;34:71–73. doi: 10.1097/wnf.0b013e318211f945. - DOI - PubMed

LinkOut - more resources