Analysis of a Mobile System to Register the Kinematic Parameters in

Ankle, Knee, and Hip based in Inertial Sensors

V´ıctor H. Flores-Morales, Byron G. Contreras-Bermeo, Freddy L. Bueno-Palomeque

and Luis J. Serpa-Andrade

Grupo de Investigaci´on en Ingenier´ıa Biom´edica GIIB-UPS, Universidad Polit´ecnica Salesiana, Cuenca, Ecuador

Keywords:

Kinematic Parameters, Joint Angles, Wearable System, Inertial Sensors.

Abstract:

Understanding the lower-extremity kinematic during daily and sport activities provides important information

in order to detect abnormalities in human gait or analyse the execution of different sport techniques. Following

this approach, this paper presents a kinematic data collection system of human gait in the lower extremities

using six inertial sensors MPU 6050 and a microcontroller ATMEGA328P-PU. Six tests were performed

and the angular variation was recorded during the execution. The curves obtained during the tests showed a

maximum error of ±4, ±1, and -4 degrees at the Yaw, Pitch, and Roll angles respectively. This study proposes

a mobile and inexpensive system for detecting the angular variation in reduced speed movements, ideal for

goniometric measurement or analyse the techniques in certain sports.

1 INTRODUCTION

The biomechanical analysis of the human gait us-

ing optical tracking system has been widely applied

to evaluate and diagnose different pathologies. Ad-

ditionally to determine gait patterns in populations,

monitor rehabilitation techniques, and analyse sport

techniques. However, these systems have been used

in controlled environments, limited by the study area,

with restrictions in the person’s movement, the equip-

ment’s weight, and the high cost of its implemen-

tation. As an alternative, in the last years, the ac-

celerometers and gyroscopes have been used to mea-

sure the kinematics of the human body. These sen-

sors located on the person’s body allow extending the

movement range in a larger amount of environments

(Foerster et al., 1999; Rueterbories et al., 2010; Liu

et al., 2009; Callaway, 2015; Mangin et al., 2015).

The biomechanical analysis of the human gait for

clinic diagnosis has permitted to study pathological

postures, measure or quantify the effects of a deter-

mined treatment, and compare the different proto-

cols developed (Wong et al., 2007). Several meth-

ods have been proposed in the literature to address

this subject such as the proposal to discriminate be-

tween asymptomatic patients and patients with Me-

dial Knee Osteoarthritis using accelerometers and gy-

roscopes (Turcot et al., 2008) or the system to ambu-

latory monitor in Parkinson’s patients (Hobert et al.,

2014). Focusing in older people, an airbag mobile

system for fall protection was proposed using a belt,

inertial sensors and a compressed air actuator that ac-

tivates two airbags for hips protection once the sen-

sors detect an acceleration produced by the person fall

(Shi et al., 2009).

The different tests developed to quantify the spe-

ciﬁc movements in contact sports have revealed the

generation of errors in the contact and in the fast and

abrupt change of speed during a collision (Chambers

et al., 2015); it has motivated to search for new alter-

natives to improve the acquisition of kinematic data.

Considering other sport, a wearable system was pro-

posed for monitoring track cycling, using accelerom-

eters to register speed, acceleration, and cadence dur-

ing the development of the technique (Lattes et al.,

2013).

Inertial method and an infrared camera system

were compared using sensors located on the athlete’s

forearm to analyse the accelerations during the arm-

stroke on a swimming bench (Lee et al., 2011). In

the same way, the characteristic of the main tem-

poral phases of front crawl on swimming were de-

tected, through algorithm to determine angular veloc-

ities and accelerations. The method was compared

with a video system showing similar results (Dadashi

et al., 2011).

Recent technology miniaturization has motivated

to propose smart new wearable systems for monitor-

Flores-Morales, V., Contreras-Bermeo, B., Bueno-Palomeque, F. and Serpa-Andrade, L.

Analysis of a Mobile System to Register the Kinematic Parameters in Ankle, Knee, and Hip based in Inertial Sensors.

DOI: 10.5220/0005934800290033

In Proceedings of the 4th International Congress on Sport Sciences Research and Technology Support (icSPORTS 2016), pages 29-33

ISBN: 978-989-758-205-9

ing continuously the patient’s healthcare (Chan et al.,

2012). This research presents a mobile system that

captures and describes the angular variations in six

different movements, performed by an asymptomatic

person. The system features a quantiﬁcation method

that helps specialists to complement the biomechani-

cal analysis.

2 MATERIALS AND METHODS

Adaptive physical and anatomical characteristics of

a person outﬁt was designed, to acquire and regis-

ter the angles of lower extremities during different

tests. This outﬁt consists of: a hip girdle subject

into the joint to prevent slipping, two knee supports

without pad for a better mobility in the joint, and two

open anklets to preventslipping. The sensors were lo-

cated where extended rotation occurs. The embedded

system collects the data sent by the inertial sensors

through an Inter-Integrated Circuit (I2C) communica-

tion for a posterior analyses. The kinematic and an-

gular data receipted were used to generate a ﬁle to

visualize, simulate and analyse the movements made

by the individual using OpenSim.

MPU 6050 inertial sensors were used as the basis

of the system to acquire the angles, a microcontroller

ATMEGA 328P-PU to collect and process the data, a

Max 485 to communicate, and a secure digital (SD)

card as storage device. A bidirectional I2C communi-

cation for sending and receiving data alternately was

used between the microcontroller and the inertial sen-

sors. The income data was transformed into angles

before sending it through a Serial Peripheral Interface

communication bus (SPI), to be organized and stored

in different ﬁles in a SD card.

The MPU 6050 contains a gyroscope to measure

the angular velocity on three axis independently. It al-

lows to calculate the rotate angle on each axis (Euler’s

angles:φ, θ, and ψ) using the uniform linear motion

equation. Gyroscope quaternion accumulates an error

in the course of time (drift). To complement data, a

global frame was used. A gravity vector was gener-

ated with the real values obtained in each instant of

time by the accelerometer to calculate the Pitch and

Roll angles through the equations 1 and 2. The gravity

quaternion contains information about the orientation

of the relative sensor’s axis to the ground reference

frame. The resultant quaternion is used to calculate

the angles compensating the drift error (Eq. 3):

Roll = sin





+ a





(1)

Pitch = sin

+ a

(2)















arctan

2(q

)

1−2(q

)

arcsin[2(q

− q

)]

arctan

2(q

)

1−2(q

)







(3)

A .mot ﬁle was generated for recreation of move-

ments in OpenSim with data from the SD card. Data

was processed by a low-pass Butterworth ﬁlter with

ten poles, with a sampling frequency of 100 Hz and a

cutoff frequency of 10Hz.

To determine the system’s error, a MPU 6050 ro-

tated about a ﬁxed axis. Real data was recollected and

compared against the angles given by the algorithm

loaded in the microcontroller in different positions.

Error measured is presented in table 1.

Table 1: Measurement of error in the system.

Mean error

Error (Deg) Probability

Yaw 0.55 ± 4.80

±4 59%

Pitch 0.27 ± 0.72

±1 80%

Roll -3.62 ± 2.35

-4 52%

3 RESULTS AND DISCUSSION

During the ﬁrst test, two gyroscopes in a state of in-

ertia were connected to a microcontroller. Figure 1

shows the real obtained data by the sensor and after

to use the Butterworth ﬁlter with a cutoff frequency

of 10 Hz. During the second test, one gyroscope with

rotational motion was connected to a microcontroller.

Figure 2 shows the real obtained data and after to use

the ﬁlter.

Figure 1: Inertial stability test.

Six tests were performed with the system. All of

which were taken from a 24 years-old healthy man

icSPORTS 2016 - 4th International Congress on Sport Sciences Research and Technology Support

Figure 2: Filtering test with one gyroscope.

using the proposed system Figure 3 shows the differ-

ent tests developed using the system and the motion’s

curves generated.

• Figure 3-A shows the hip ﬂexion of 48 degrees.

• Figure 3-B shows the angular variation during a

hip abduction of 22 degrees and an adduction of 0

degrees.

• Figure 3-C shows the angular variation during a

hip rotation of -25.

• Figure 3-D shows a knee ﬂexion of -73 degrees

from rest.

• Figure 3-E shows an angle variation of 32 degrees

during the dorsal ﬂexion and -10 degrees during

the plantar ﬂexion.

Finally, the ﬁgure 4 shows curves corresponding

to the hip, knee and ankle during a complete gait

cycle in the right leg. The measurements were all

made in the sagittal plane using our proposed system

and the motion tracking software Kinovea 0.8.15 for

Windows, considering that this method is reliable and

valid (Balsalobre-Fern´andez et al., 2014).

This study presents a system based on inertial sen-

Figure 3: Curve describing: A) ﬂexion in the hip, B) abduction and adduction in the hip, C) rotation in the hip, D) ﬂexion in

the knee, and E) dorsal ﬂexion and plantar ﬂexion in the ankle.

Analysis of a Mobile System to Register the Kinematic Parameters in Ankle, Knee, and Hip based in Inertial Sensors

Figure 4: Angular variation of the right leg during a gait cy-

cle on the sagittal plane. IC = initial contact, OT = opposite

toe-off, HR = heel rise, OI = opposite initial contact, TO =

toe-off, FA = feet adjacent, and TV = tibia vertical.

sors to collect data from the kinematics of a per-

son and represent them graphically in an environment

where it is possible to make a detailed biomechanical

analysis. The wearable system allowed located the

sensors on the hip, knee, and ankle joints in order to

develop the different tests on an asymptomatic per-

son. The system has limitations in certain conditions

where movement and speed are high and complex.

Different studies have focused in decrease and to

eliminate the drift error occasioned by the use of gy-

roscopes (Alonge et al., 2014). In this ﬁeld our study

used a gravity vector to decrease the accumulate error

as shown in table 1. This error is negligible consid-

ering the gyroscopes range used. The margin of error

as the mean were increased due to false information

caused by noise. The results obtained during the static

tests (Figure 2) show that the angular variation on the

joints have a small variation comparing with the ﬁnal

angle reached during the six tests. The use of the ﬁlter

caused a shift in the curve of movement on the time

axis which is interpreted as a delay episode. As the

ﬁlter is the same for all the signals, the time delay is

the same for all of them. This shows that the delay is

irrelevant in this application since they are not being

displayed in real time (Figure 2).

Figure 3 shows the angle variation of the volun-

teer’s right leg registered with our system during the

execution of six different tests. To validate the value

obtained, we used a graduated device to compare

the measurements registered. Additionally, ﬁgure 4

shows the average angle obtained during a gait cycle

test, considering it as a normal speed activity. Our

system shows a delayed measurement, comparing it

with the results provided by Kinovea system. Partic-

ularly, in the hip joint, the maximum angular varia-

tion registered is 7.0

◦

at the initial contact. The maxi-

mum variation on the knee is 9.1

◦

on the feet-adjacent

phase and ﬁnally, on the ankle, the maximum varia-

tion is 5.0

◦

on the opposite toe-off phase. The gait cy-

cle time of 1.44 s affects directly on the measurements

of our system, although the curves have the same pat-

tern, our system registered a lower and delayed an-

gle than the Kinovea system. The microcontroller re-

sponsible for data storage and transmission coordina-

tion, optimizes the largest amount of data collected

considering the complexity in handling serial commu-

nication, i2c, SPI, and data processing in the memory

space available. To achieve higher speed data capture,

it is important to improve the data management; ad-

ditionally, the calculation should run in a processor of

higher capacity.

4 CONCLUSIONS

An embedded system was developed that can be used

as orthopaedic tool for recording and analysing kine-

matic data of a person. The static tests demonstrated

that the system can be used for goniometric measure-

ments, to perform outdoor tests, and analysis in cer-

tain sporting techniques. The results presented in this

study could contribute to the development of a low

cost wearable system for the gait analysis using ac-

celerometers. With the respective corrections in the

time of data collection and improving the ﬁltering sys-

icSPORTS 2016 - 4th International Congress on Sport Sciences Research and Technology Support

tem could be implemented to complement the sports

studies at high speed disciplines.

REFERENCES

Alonge, F., Cucco, E., D’Ippolito, F., and Pulizzotto, A.

(2014). The use of accelerometers and gyroscopes to

estimate hip and knee angles on gait analysis. Sensors,

14(5):8430–8446.

Balsalobre-Fern´andez, C., Tejero-Gonz´alez, C. M., del

Campo-Vecino, J., and Bavaresco, N. (2014). The

concurrent validity and reliability of a low-cost, high-

speed camera-based method for measuring the ﬂight

time of vertical jumps. The Journal of Strength &

Conditioning Research, 28(2):528–533.

Callaway, A. J. (2015). Measuring kinematic variables in

front crawl swimming using accelerometers: A vali-

dation study. Sensors, 15(5):11363–11386.

Chambers, R., Gabbett, T. J., Cole, M. H., and Beard, A.

(2015). The use of wearable microsensors to quantify

sport-speciﬁc movements. Sports Medicine, pages 1–

17.

Chan, M., Est`eve, D., Fourniols, J.-Y., Escriba, C., and

Campo, E. (2012). Smart wearable systems: Current

status and future challenges. Artiﬁcial intelligence in

medicine, 56(3):137–156.

Dadashi, F., Crettenand, F., Millet, G., Seifert, L., Komar,

J., and Aminian, K. (2011). Front crawl propulsive

phase detection using inertial sensors. Port. J. Sport

Sci, 11:855–858.

Foerster, F., Smeja, M., and Fahrenberg, J. (1999). Detec-

tion of posture and motion by accelerometry: a vali-

dation study in ambulatory monitoring. Computers in

Human Behavior, 15(5):571–583.

Hobert, M. A., Maetzler, W., Aminian, K., and Chiari, L.

(2014). Technical and clinical view on ambulatory

assessment in parkinson’s disease. Acta Neurologica

Scandinavica, 130(3):139–147.

Lattes, M., Gatti, R., Roa, Y., Fruett, F., and Cunha, S.

(2013). Assessment of micro sensor (sm-mae) in mon-

itoring of cycling. In XXIV Congress of the Interna-

tional Society of Biomechanics.

Lee, J. B., Burkett, B. J., Thiel, D. V., and James, D. A.

(2011). Inertial sensor, 3d and 2d assessment of stroke

phases in freestyle swimming. Procedia Engineering,

13:148–153.

Liu, K., Liu, T., Shibata, K., Inoue, Y., and Zheng, R.

(2009). Novel approach to ambulatory assessment

of human segmental orientation on a wearable sensor

system. Journal of biomechanics, 42(16):2747–2752.

Mangin, M., Valade, A., Costes, A., Bouillod, A., Acco, P.,

and Soto-Romero, G. (2015). An instrumented glove

for swimming performance monitoring. In Proceed-

ings of the 3rd International Congress on Sport Sci-

ences Research and Technology Support, pages 53–

58.

Rueterbories, J., Spaich, E. G., Larsen, B., and Andersen,

O. K. (2010). Methods for gait event detection and

analysis in ambulatory systems. Medical engineering

& physics, 32(6):545–552.

Shi, G., Chan, C. S., Li, W. J., Leung, K.-S., Zou, Y., and

Jin, Y. (2009). Mobile human airbag system for fall

protection using mems sensors and embedded svm

classiﬁer. Sensors Journal, IEEE, 9(5):495–503.

Turcot, K., Aissaoui, R., Boivin, K., Pelletier, M., Hage-

meister, N., De Guise, J., et al. (2008). New accelero-

metric method to discriminate between asymptomatic

subjects and patients with medial knee osteoarthritis

during 3-d gait. Biomedical Engineering, IEEE Trans-

actions on, 55(4):1415–1422.

Wong, W. Y., Wong, M. S., and Lo, K. H. (2007). Clinical

applications of sensors for human posture and move-

ment analysis: a review. Prosthetics and orthotics in-

ternational, 31(1):62–75.

Analysis of a Mobile System to Register the Kinematic Parameters in Ankle, Knee, and Hip based in Inertial Sensors