A Visual Analysis Approach to Static Postural Control Acquired by a

Force Plate

Thales Vinicius de Brito Uê, Danilo Medeiros Eler

and Iracimara de Anchieta Messias

Faculty of Science and Technology, São Paulo State University, Presidente Prudente, São Paulo, Brazil

Keywords: Visual Analysis, Center of Pressure, Postural Control, Force Plate, Elderly People, Sarcopenia.

Abstract: Force plates are biomechanical equipment responsible for providing data to understand the mechanics of

human movement. However, mathematical software used to process data are a barrier to researchers without

much experience and prior knowledge on areas from Exact Sciences and Information Technology. This paper

aims to implement a visual approach to analyze human static postural control obtained from a force plate as

a means of helping researchers interpret its data. By measuring ground reaction forces and their respective

torque moments, the displacements of the Center of Pressure in its medial‑lateral and anterior-posterior

directions are calculated to observe and evaluate the postural balance’s behavior. Data processing and

visualization were implemented using Python programming language. Scatter plots, heat maps, violin plots,

and box plots were used as graphic representations for data collected before and after muscular intervention

in older adults with sarcopenia. Applying the developed approach makes it possible to visually observe each

of the Center of Pressure’s oscillation values measured for data collection and how they relate. This fact

differs from statistical information, which summarizes the sample’s data in a quantified value. Therefore, data

visualization is essential to complement the statistical data and provide another view to force plate data.

1 INTRODUCTION

Biomechanical equipment is essential to obtain data

related to the mechanics of human movement and its

evaluation and understanding. In addition to the

quantitative analysis methods for this area of study,

there is the qualitative part of the evaluation, which

can be provided through data visualization for a

non

‑

numerical and visual way to display values

acquired for a variable. As an example of

biomechanical research, it is possible to qualitatively

assess the behavior of a person's postural control in a

dynamic state of movement or standing still, both

actions under different vision and surface conditions.

However, one barrier to biomechanical analysis is

the mathematical software used to process data.

Researchers must have prior knowledge or learn how

to manipulate these tools to analyze their data, which

makes this step reliant on the experience and

expertise presented by the researcher on other areas

from Exact Sciences and Information Technology

(Dunn et al., 2017). As a result, several

biomechanical analyses are quantitative, and only the

https://orcid.org/0000-0002-9493-145X

most experienced researchers use qualitative methods

to look at the data. Consequently, the use and benefits

of data visualization for biomechanical analyses

remain unknown to some researchers from

Healthcare areas.

For those with experience in analyzing data,

software such as MATLAB and Origin are employed

to create visual representations for biomechanical

variables. In the case of research based on data

collections from force plate equipment, the behavior

of a person's Center of Pressure (COP) is evaluated

by visualizing one of its displacement directions by

the other, which represents right-left and front-back

oscillations (Duarte and Freitas, 2010). Since

including programming languages, such as Python

and R, in analyses, the options and possibilities for

data visualization have become increasingly

extensive.

This paper aims to implement graphic

visualizations using data visualization libraries from

Python (Seaborn, Plotly Express and Plotly Graph

Objects) to improve the analysis of human static

postural control acquired by a force plate to help

628

Uê, T., Eler, D. and Messias, I.

A Visual Analysis Approach to Static Postural Control Acquired by a Force Plate.

DOI: 10.5220/0012599700003690

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 26th International Conference on Enterprise Information Systems (ICEIS 2024) - Volume 1, pages 628-635

ISBN: 978-989-758-692-7; ISSN: 2184-4992

evaluate, interpret, and disseminate postural analysis

results to researchers with different levels of

experience. The force plate used in this paper

measures the components of ground reaction forces

and their respective torque moments applied to its

surface. From its results, the Center of Pressure in its

medial‑lateral (ML) and anterior‑posterior (AP)

directions were calculated to assess its displacement

during data collection and evaluate the behavior of

human postural balance. In the experiments, the

analyzed data belong to static positions performed by

older adults with sarcopenia before and after being

subjected to a muscular intervention (Bertolini et al.,

2021). Both data processing and visualization were

implemented using Python programming language.

In the visual analysis, scatter plots, heat maps,

violin plots, and box plots corresponded to the

graphical representations used to evaluate older

adults’ postural balance through the behavior of their

COP. Thus, each oscillation in their balance can be

visually identified by displaying all COP’s values and

the relationships between them. This way of looking

at the data differs from the statistical information,

which summarizes a sample’s data in quantified

values. Applying a visual approach to look at the

results acquired from a force plate is essential to

contribute to biomechanical analyses by

complementing the statistical approach and providing

an overall view of the data.

The second section of this paper presents related

works to provide perceptions on possible applications

with a force plate and how its data has been analyzed

quantitatively and qualitatively. The third section

describes the methodology employed for visual

analysis of the data acquired from the force plate. The

fourth section provides a case study for applying the

developed visual approach to analyze one of the static

positions performed by an older adult with

sarcopenia. In the fifth and last section, conclusions

and further works are presented.

2 RELATED WORKS

Borges et al. (2016) compared static postural balance

among older adults with and without mild cognitive

impairment using a three-dimensional

electromagnetic sensor system. The calculated

variables of interest corresponded to velocity and

displacement of the Center of Pressure. Tables, line

graphs, and scatter plots were implemented using the

Origin software to analyze the data. Scatter plots

represented each direction of the COP’s displacement

by the other, and line graphs were used to combine

these directions at the time of data collection.

Pinto et al. (2019) used four force plates and eight

motion capture cameras to compare postural control

among two yoga practitioners and two

non‑practitioners. The variables for their evaluation

involved displacements of the Centers of Mass and

Pressure, components of ground reaction force, and

amplitudes of joint angles, such as the hip, lower

back, and knee. Tables, line graphs, and scatter plots

were implemented by using MATLAB software for

the analysis step. Scatter plots represented the

directions of the Center of Pressure’s displacement by

each other.

Seo et al. (2022) assessed the balance of older

adults in different vision and surface conditions: eyes

open and closed with firm and foam surfaces. A

Nintendo Wii Balance Board force plate was chosen to

capture the amplitude, velocity, area, and covariance of

the Center of Pressure’s displacements. The MATLAB

software implemented tables and scatter plots,

illustrating the medial-lateral direction of the COP’s

displacement by the anterior-posterior one.

Zychowska et al. (2022) used an AMTI force

plate to evaluate the consequences of COVID-19

infection on postural control. Within the analyzed

group of infected people, those with respiratory

abnormalities and others with olfactory abnormalities

were also observed as a consequence of COVID-19.

MATLAB software assessed the trajectory of the

Center of Pressure displacements. The analyses were

quantitatively, through tables, and qualitatively, with

scatter plots. The visualizations combined one

direction of the Center of Pressure’s displacement

with the other.

Herrera et al. (2023) developed a framework to

aid the application of Virtual Reality in the context of

upper limb rehabilitation. The software can record

and store kinematic data during the manipulation of

objects in virtual environments. Data acquired from

the developed application can be used to evaluate

each patient’s rehabilitation progress. In a case study

to test the software’s functionalities, 10 healthy

individuals aged between 15 and 39 were conducted

in three trials for the conditions of a real environment

and an immersive virtual one to measure the number

of blocks they were able to pass from one side of a

box to the other within a minute. To compare the

results, the mean value of the three attempts was

calculated and a line graph displayed a link

connecting values obtained for each environment. A

three-dimensional heat map was also used as an

example of visual analysis for the recorded kinematic

data.

A Visual Analysis Approach to Static Postural Control Acquired by a Force Plate

629

3 VISUAL ANALYSIS APPROACH

AND RESULTS

To implement a visual approach to analyze

biomechanical data acquired from a force plate, the

data used by this paper came from research on the

evaluation of static postural control during the

moments before and after a muscle training

intervention applied to older adults with sarcopenia, in

other words, a loss of muscle mass responsible for

affecting their postural balance (Bertolini et al., 2021).

During the data collection, 22 older people (12

women and 10 men aged between 59 and 93) were

given three attempts at the static positions of feet

together, feet apart, and semi-tandem, under the

condition of eyes open and closed. The first two

positions were performed for 30 seconds, and the

third was performed for 10 seconds. The last

performed position was unipodal support with the

dominant and non

‑

dominant foot for 10 seconds. Data

was collected at 100 Hz (Bertolini et al., 2021).

The OR6-6 model for the force plate made by

Advanced Mechanical Technology, Inc. (AMTI) was

used to collect data from older people (Bertolini et al.,

2021). With this biomechanical equipment, the

components of the ground reaction force and their

respective torque moments are measured according to

the person's contact with the surface of the force plate.

Both forces and moments act in an orthogonal

coordinate system, in which the x and y-axis are the

horizontal components, and the z-axis is the vertical

one. From the obtained values, it is possible to

calculate, throughout the data collection's duration,

the displacement of the Center of Pressure in its

medial-lateral and anterior-posterior directions (in

centimeters), which respectively correspond to

oscillations along the x and y directions. The COP

consists of a positioning variable responsible for

indicating right-left (medial-lateral) and front-back

(anterior-posterior) displacements (Duarte and

Freitas, 2010).

For this paper, the Python programming language

was used to calculate the oscillations of the Center of

Pressure acquired by a force plate and then to

implement the graphic visualizations. The x-axis of the

graphs represents the COP's ML displacement, and its

values are increasing towards the left since older adults

were positioned on the force plate so that the x-axis has

a positive direction to their left side. The y-axis has a

positive sense towards the front of them. Therefore, the

y-axis of the graphs represents the AP displacement,

with its values increasing upwards.

The values in each calculated variable of the

Center of Pressure were translated to make the first

one collected corresponding to zero. To accomplish

this, the initial value was subtracted by itself, and all

the other values were subtracted from the initial one.

As a result, reflected in the visualizations, the person's

static postural balance is analyzed with the first

acquired value as a reference while displaying it as

the graph's central point (0,0).

3.1 Scatter Plots

Scatter plots combine the medial-lateral and

anterior‑posterior displacement variables to display

the concentration and dispersion of the Center of

Pressure’s behavior values. It is possible to notice

where the highest and lowest concentrations of data

are. In addition, the graph’s values can be colored

according to the time they occurred during the data

collection, which allows us to know which data is

from the beginning and the end of the phenomenon.

Figure 1 illustrates a scatter plot for the

semi‑tandem position with eyes closed during the first

pre-intervention execution attempt. The light colors

indicate the data collected at the beginning of the

acquisition, while the dark colors indicate the final

moments of it. Histograms can be added on the side

of the visualizations to help locate the regions with

concentration values.

Figure 1: Scatter plot colored by time (Tempo) and

combining the COP’s displacement in medial-lateral

(COP_ML) direction by its anterior-posterior (COP_AP)

direction.

Scatter plots can also be used to individually

display the COP’s displacement directions by the

time for data collection. Similarly, it is possible to

color the values according to the time they were

Figure 2: Scatter plot colored by time (Tempo) and

combining the COP’s displacement in medial-lateral

(COP_ML) direction by the time for data collection.

ICEIS 2024 - 26th International Conference on Enterprise Information Systems

630

collected. Figure 2 shows the visualization for the ML

direction by time, and Figure 3 displays this

combination for the AP direction. The represented

static position is the same as in Figure 1.

Figure 3: Scatter plot colored by time (Tempo) and

combining the COP’s displacement in anterior-posterior

(COP_AP) direction by the time for data collection.

3.2 Heat Maps

Heat maps also combine the Center of Pressure’s

directions of displacement. However, this data

visualization style displays the regions with a

concentration of values in a topographical way.

Figure 4 illustrates a heat map for the same static

position as Figure 1. Warm colors indicate the highest

concentrations of data, and cool colors show where

there are fewer values present. Line graphs on the

sides of the visualization can also be added to help

indicate the regions with the highest density of values.

Figure 4: Heat map combining the COP’s displacement in

medial-lateral (COP_ML) direction by its anterior-posterior

(COP_AP) direction.

3.3 Violin Plots and Box Plots

Box plots display the distribution behavior of the

variables’ values. With this visualization style, the

displacements of the COP in the ML and AP

directions can be analyzed individually. In addition,

they provide visual information on the maximum,

minimum, median, and quartiles of a sample’s data.

Thus, it is possible to be aware of the amplitude and

symmetry present in the data. A box plot can be

included within a violin plot, whose purpose is to

display the density of the data as well, but in a

mirrored way. In this visualization, the stretch of the

violin’s extremities along the graph’s axis indicates

how far apart the values are, and its peaks indicate

how concentrated they are.

For the same static position in Figure 1, Figure 5

provides a violin plot with a box plot to observe the

values acquired in the ML direction. Based on the

same visual style, Figure 6 shows the performance of

the Center of Pressure’s AP direction. Both figures

have each of the data displayed next to the graph as

support for the visualization.

Figure 5: Violin plot with a box plot within it for the

behavior of the COP’s displacement in the medial-lateral

(COP_ML) direction.

Figure 6: Violin plot with a box plot within it for the

behavior of the COP’s displacement in the

anterior‑posterior (COP_AP) direction.

4 DISCUSSIONS

As a demonstration for applying the visual approach

developed in this paper, the results for the static

semi

‑

tandem position with eyes closed during pre-

and post-intervention are evaluated and compared

based on the graphic visualizations implemented with

Python language.

From the scatter plots and heat maps combining

one direction of the Center of Pressure’s displacement

by the other, it is possible to notice large dispersions

in the values from the first and third attempts of

pre‑intervention and the first one of post-intervention.

These results indicate the older adult had more

difficulty in finding their balance point control

throughout the data collection, which caused their

COP to oscillate at practically every moment. On the

A Visual Analysis Approach to Static Postural Control Acquired by a Force Plate

631

second attempt for each moment of intervention,

dispersions also occurred, but on a smaller scale. On

the second one, for pre-intervention, older adults

started the data collection with more minor variations

in their balance. On the second attempt in

post‑intervention, they finished it with the values of

the COP’s displacements, concentrating on a single

region.

Additionally, an improvement can be noticed in

the post-intervention results, culminating in its third

attempt to present values much closer to and around

Figure 7: Scatter plots for the pre-intervention moment with

the first, second and third (from top to bottom respectively)

attempts at the semi-tandem position with eyes closed.

Figure 8: Scatter plots for the post-intervention moment

with the first, second and third (from top to bottom

respectively) attempts at the semi-tandem position with

eyes closed.

the graph's central point (0,0). Such behavior from the

results indicates an improvement in the older person's

ability to maintain their balance throughout the data

collection by displacing their Center of Pressure

much less than in previous collections. Figures 7 and

8 illustrate the pre-and post-intervention results with

scatter plots. Likewise, Figures 9 and 10 display the

data with heat maps.

Figure 9: Heat maps for the pre-intervention moment with

the first, second and third (from top to bottom respectively)

attempts at the semi-tandem position with eyes closed.

Figure 10: Heat maps for the post-intervention moment

with the first, second and third (from top to bottom

respectively) attempts at the semi-tandem position with

eyes closed.

From the scatter plots combining the

medial‑lateral direction of the COP’s displacement by

the duration of data collection, a more prominent

oscillation in the values was noticed during most

pre‑intervention attempts. In contrast, most

post‑intervention data was closer to zero, with the

second attempt showing the tiniest variations.

Applying the same visualization style for the

anterior‑posterior direction over time makes it

possible to notice a similarity in most of the values’

behavior for both moments of intervention. The

second attempt in pre-intervention and the third one

in post-intervention show the closest data to zero.

Figure 11 displays the results for the ML direction in

pre- and post-intervention. Similarly, Figure 12

illustrates the results for the AP direction.

Instead of displaying the temporal information

alongside the Center of Pressure’s displacement

values, violin graphs with box plots within them

enable the density of a single sample’s data to be

observed in a more detailed way. For the ML

direction visualization, it is possible to notice most of

the pre‑intervention attempts, and the first one in

post‑intervention displays the most prominent

ICEIS 2024 - 26th International Conference on Enterprise Information Systems

632

Figure 11: Scatter plots for the medial-lateral direction

during pre- (left) and post-intervention (right) with the first,

second and third (from top to bottom respectively) attempts

at the semi‑tandem position with eyes closed.

Figure 12: Scatter plots for the anterior-posterior direction

during pre- (left) and post-intervention (right) with the first,

second and third (from top to bottom respectively) attempts

at the semi‑tandem position with eyes closed.

oscillations in the older adult’s postural balance. On

these attempts, the violin stretches throughout the

graph, indicating a dispersion in the values. However,

post-intervention results indicate an improvement in

minimizing the COP’s displacement in its ML

direction and maintaining a more concentrated

posture control at each sample taken. Values in these

attempts are closer to zero, especially in the second

one, and the graph’s violin is not stretched as much.

Finally, for AP direction, attempts to present the

smallest dispersions correspond to the second data

collection in pre-intervention and the last two in

post‑intervention. These results indicate a more

considerable concentration of data approaching the

zero value. In the post-intervention moment, it is

possible to observe the values becoming increasingly

concentrated with each attempt taken. Figure 13

illustrates the ML direction obtained in pre- and

post

‑

intervention. Likewise, Figure 14 displays the

results for AP direction.

Figure 13: Violin plot with a box plot within it for the

medial-lateral direction in pre- (left) and post-intervention

(right) with first, second and third (from top to bottom

respectively) attempts at semi‑tandem position with eyes

closed.

Figure 14: Violin plot with a box plot within it for the

anterior‑posterior direction in pre- (left) and

post‑intervention (right) with first, second and third (from

top to bottom respectively) attempts at semi‑tandem

position with eyes closed.

5 CONCLUSIONS AND FURTHER

WORKS

The mechanics of human movement are measured

and evaluated through a variety of biomechanical

equipment. However, its data analyses consist

primarily of quantitative steps, and not all researchers

are aware of the qualitative methods for looking at

their data. To develop a visual approach to analyze

data acquired from a force plate, this paper aims to

apply data visualization to the evaluation of static

postural control in older adults with sarcopenia

submitted to a muscular intervention to demonstrate

visualizations' contributions to biomechanical

research.

A Visual Analysis Approach to Static Postural Control Acquired by a Force Plate

633

The Center of Pressure was calculated in its

medial-lateral and anterior-posterior directions to

observe the postural balance’s behavior. Data

processing and visualization were implemented using

Python programming language. Scatter plots, heat

maps, violin plots, and box plot visualizations were

modeled to provide qualitative information about the

values acquired for each of the COP’s displacement

directions, individually combined by the data

collections’ duration and the combinations of one of

the oscillation directions by the other.

When the graphic visualizations were applied to

compare the static semi-tandem position with eyes

closed performed for an older adult during the data

collections at pre- and post-intervention, it was

possible to identify the behavior of their postural

balance throughout every attempt of positioning. The

concentrations, dispersions, and variations in the

COP’s oscillation values were displayed in the graphs

to be interpreted. After visually comparing the results,

it was possible to determine an improvement after the

muscular intervention in the older adult’s postural

control for the analyzed static position.

Applying a visual approach to biomechanical

analyses with data acquired from a force plate enables

each value present in a sample to be visualized

graphically. This way, it is possible to know all the

collected data's behavior and how they relate. This

overall view of the results differs from statistical

information by not presenting a numerical value to

characterize the phenomenon analyzed. For static

postural control analyses, there is a visual perception

of how the person's Center of Pressure behaves

throughout the data collection, not a summarization

of such action into a numerical quantification. The

addition of temporal information in the visualizations

is also relevant to the exact awareness of postural

balance's performance at the beginning, middle, and

end of a static positioning execution.

In addition, the variety of graphic visualization

styles opens up a wide range of possibilities for

representing a sample's data. Each visual style has

characteristics that can enhance the others, and more

than one visualization can complement one another in

the same graph.

Therefore, graphic visualizations are essential to

support and contribute to force plate users during the

analyses of their results and provide different ways of

looking at their data. Visual information is also

necessary to increase and complement the

quantitative side of biomechanical research.

For this paper, no dynamic postural control data

was used, and no other variables related to the Center

of Pressure were calculated. With a force plate, it is

possible to capture performances regarding dynamic

actions (e.g., gait, jump), and to calculate additional

COP-related variables (e.g., velocity, frequency,

area).

As further works, developing a non-paid,

intuitive, and user-friendly computational interface

for enabling researchers with different experiences in

data analysis to create graphic visualizations and use

them to contribute to their biomechanical evaluation

of postural control acquired from a force plate is

proposed, disseminating the use of visualization in

analyses from force plate data. Furthermore, the tool

can provide the calculation and visualization for other

COP-related variables and add statistical information

provided in tables to complement the qualitative data.

Additional functionalities can correspond to enable

users to change the variable chosen to color values in

scatter plots. Parameters that differ according to the

research with the force plate (e.g., frequency of data

acquisition) can also be altered by users. The type of

force plate from which the data will be collected must

be previously decided, as force plates can generate

different data and file formats, saving them. Finally,

the recommendations and suggestions provided by

the participations of force plate users are essential to

validate the tool’s usability and functionalities, as

well as the visualizations used for postural control

analyses.

ACKNOWLEDGEMENTS

This study was financed in part by the Coordenação

de Aperfeiçoamento de Pessoal de Nível Superior -

Brasil (CAPES).

REFERENCES

Bertolini, G. N., Silva, B. S. A., Santos, V. R., Messias, I.

A., Ribeiro, J. P. J., Marini, E., Gobbo, L. A. (2021).

Are bioelectrical parameters and functionality

associated with postural control in the elderly? Clinical

Biomechanics. Bristol, v. 82.

Borges, A. P. O., Carneiro, J. A. O., Zaia, J. E., Carneiro,

A. A. O., Takayanagui, O. M. (2016). Evaluation of

postural balance in mild cognitive impairment through

a three-dimensional electromagnetic system. Brazilian

Journal of Otorhinolaryngol. v. 82, n. 4.

Duarte, M., Freitas, S. M. S. F. (2010). Revisão sobre

posturografia baseada em plataforma de força para

avaliação do equilíbrio. Rev. Bras. Fisioter. São Carlos,

v. 14, n. 3.

Dul, J., Weerdmeester, B. (2012). Ergonomia prática,

Blucher. São Paulo, 3

edition.

ICEIS 2024 - 26th International Conference on Enterprise Information Systems

634

Dunn, W., Burgun, A., Krebs, M., Rance, B. (2017).

Exploring and visualizing multidimensional data in

translational research platforms. Brief Bioinform. v. 18,

n. 6.

Falzon, P. (2018). Ergonomia, Blucher. São Paulo, 2

edition.

Falzon, P. (2007). Ergonomia Construtiva, Blucher. São

Paulo, 1

edition.

Hall, S. J. (2016). Biomecânica Básica, Guanabara Koogan.

Rio de Janeiro, 7

edition.

Hamill, J., Knutzen, K. M., Derrick, T. R. (2016). Bases

biomecânicas do movimento humano, Manole. Barueri,

São Paulo, 4

edition.

Hatze, H. (1974). The meaning of the term “biomechanics”.

Journal of Biomechanics. v. 7, n. 2.

Pheasant, S. (2003). Bodyspace: Anthropometry,

Ergonomics and the Design of Work, CRC Press.

London, 2

edition.

Pinto, D. P., Sarmet, P., Pino, A. V., Menegaldo, L. L.

(2019). Biomechanics of Postural Control in Yoga: A

Pilot Study. In: XXVI Brazilian Congress on

Biomedical Engineering. Armação dos Búzios, Rio de

Janeiro, Brazil.

Seo, J. -W., Kim, J. I., Kim, T., Jang, K. -M.; Jeong, Y., Do,

J. -H. (2022). Sensory Interaction and Balancing

Ability Evaluation of the Elderly Using a Simplified

Force Plate System. Sensors. Basel, v. 22.

Ward, M., Grinstein, G. G., Keim, D. (2010). Interactive

Data Visualization: Foundations, Techniques, and

Applications, A K Peters, Ltd. Natick, Massachusetts.

Zychowska, M., Jaworecka, K., Mazur, E., Slomka, K.,

Marszalek, W., Rzepko, M., Czarny, W., Reich, A.

(2022). COVID-19 and Postural Control - A

Stabilographic Study Using Rambling-Trembling

Decomposition Method. Medicina. v. 58, n. 2.

A Visual Analysis Approach to Static Postural Control Acquired by a Force Plate

635