sensors
Article
Validity and Reliability of Inertial Measurement System for
Linear Movement Velocity in Flywheel Squat Exercise
Sergio Maroto-Izquierdo 1 , Kazunori Nosaka 2 , Jesús Alarcón-Gómez 3
Citation: Maroto-Izquierdo, S.;
Nosaka, K.; Alarcón-Gómez, J.;
Martín-Rivera, F. Validity and
Reliability of Inertial Measurement
System for Linear Movement Velocity
in Flywheel Squat Exercise. Sensors
2023, 23, 2193. https://doi.org/
10.3390/s23042193
Academic Editor: Benedicte
Vanwanseele
Received: 12 January 2023
Revised: 8 February 2023
Accepted: 14 February 2023
Published: 15 February 2023
Copyright: © 2023 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
Sensors 2023, 23, 2193. https://doi.org/10.3390/s23042193
and Fernando Martín-Rivera 3, *
1 i+HeALTH, European University Miguel de Cervantes, 47012 Valladolid, Spain
2 School of Medical and Health Sciences, Centre for Human Performance, Edith Cowan University,
Joondalup, WA 6027, Australia
3 Research Group in Prevention and Health in Exercise and Sport, University of Valencia, 46010 Valencia, Spain
* Correspondence: fernando.martin-rivera@uv.es; Tel.: +34-963-864-339
Abstract: The aim of this study was to examine the validity and reliability of an Inertial Measurement
System integrated into a secondary pulley (IMS) for determining linear velocity during flywheel
squat exercises. Thirty-one male participants who were highly experienced in a flywheel resistance
exercise training performed flywheel squat exercises with three incremental loads, and mean velocity
(MV), mean propulsive velocity (MPV) and max velocity (Vmax) of the exercises were simultaneously
recorded with a validated linear encoder and the IMS, in two different sessions. Validity was
analyzed using ordinary least products regression (OLP), Lin’s concordance correlation coefficient
(CCC), and Hedge’s g for the values from the linear encoder and the IMS. Test-retest reliability was
determined by coefficient of variation (CV), Intraclass correlation coefficient (ICC), and standard error
of measurement (SEM). Results showed a high degree of validity (OLP intercept = −0.09–0.00, OLP
slope = 0.95–1.04, CCC = 0.96–0.99, Hedge’s g < 0.192, SEM = 0.04–0.08) and reliability (CV < 0.21%,
ICC > 0.88, SEM < 0.08). These results confirm that the IMS provides valid and reliable measures of
movement velocity during flywheel squat exercises.
Keywords: isoinertial; velocity-based training; encoder; validity; reliability
1. Introduction
Velocity-based training (VBT) uses movement velocity to guide the prescription of
training load during resistance exercise [1,2]. The use of this method of load prescription
has been suggested to result in greater improvements in muscular strength and power
through a more accurate intensity prescription in comparison to traditional methods based
on the percentage of one-repetition maximum (1-RM) [3]. It has been proposed that VBT
enables the objective determination of a given relative load [4], since exercise performance
and neuromuscular fatigue can be precisely controlled during exercise, while providing
information that can maximize training outcomes and reduce the risks of overtraining [5].
The foundation of VBT centers on the quantification of concentric linear velocity of
each repetition (e.g., concentric mean velocity or mean propulsive velocity) with a linear
position transducer [1], smartphone app [6], inertial movement sensor [7] or wearable
devices [8]. In addition to the quantification of velocity, these devices can allow for the
estimation of other kinetic and kinematic variables such as force and power [4]. Generally,
the use of VBT focuses on the concentric velocity as training intensity is determined by the
individual’s capacity to lift a weight (i.e., maximal concentric muscle strength) [9]. However,
it is important to note that a higher absolute load can be used during the eccentric phase of
a given resistance exercise [10]. As such, when attempting to optimize the loading strategy
during resistance exercise, it is warranted to provide greater loads during the eccentric
phase of any given exercises [11], either by prescribing high-speed stretch–shortening cycle
exercises [12] or by providing an accentuated eccentric loading [13].
https://www.mdpi.com/journal/sensors
Sensors 2023, 23, 2193 2 of 11

Flywheel technology represents one of the most-used exercise paradigms to provide

eccentric-overload during resistance exercises [14]. Flywheel isoinertial devices use the
flywheel principle [15]. Thus, to achieve an eccentric-overload, a maximum concentric
muscle contraction is required to unwind the strap connected to the inertia wheel axis,
spinning velocity into the flywheel system and accumulating kinetic energy, which occurs
in response to increases in rotational speed. Once the concentric contraction is completed,
the flywheel strap rewinds requiring a concentrated brief maximum eccentric contraction
at the end of the eccentric phase to resist the forces that are applied in response to the
rewinding of the strap [14,16]. It has been documented that isoinertial accentuated eccentric
resistance exercise training is effective for muscle hypertrophy and increasing in muscle
strength, muscle power, jumping performance and running speed in comparison with
traditional resistance training [14,17]. It has been shown that flywheel resistance training is
effective for enhancing sports performance [18], injury prevention [19] and rehabilitation
and clinical context [20,21].
Due to the unique loading pattern associated with flywheel exercise, both concentric
and eccentric muscle contractions can inform the training process by providing velocity
and power data as well as the amount of eccentric overload [22]. Therefore, monitoring
devices that provide kinematic data (i.e., angular speed) during flywheel exercises have
been developed [23,24]. Rotary encoders are usually attached to the main axis of rotation,
providing information about wheel angular velocity that can be used to estimate force
and power when considering the moment of inertia [25]. However, the use of rotary
encoders comes with some challenges as the inertia flywheel only changes its rotational
direction between repetitions, which makes it difficult to differentiate between concentric
and eccentric muscle actions and may smooth the oscillations of high-frequency velocity
actions [7,26]. This may result in the linear encoder providing inaccurate kinetic and
kinematic data [22,24].
To overcome these issues, recent studies have proposed the use of linear velocity to
monitor concentric and eccentric velocity during flywheel exercises [7,22,26]. To this end,
some studies have used an Inertial Measurement System (IMS) sensor (IonClinics & Deionic
S.L, L’Alcudia, Spain) to control flywheel velocity [27,28]. The main difference between the
IMS sensor and the rotatory encoder is its point of attachment, with the IMS being attached
in a secondary pulley, which allows it to act like a linear encoder. Fundamentally due to
its placement the IMS allows the differentiation of the kinematic data collected during
the eccentric and concentric phases of the lift and a potential reduction of the possible
kinematic dissonances between at the level of the axis of rotation and at the trainee’s
level. The implementation of IMS sensors in flywheel devices could facilitate the correct
prescription of velocity-based training as well as the acquisition of reliable data in force
and power evaluation tests [22].
However, to the best of our knowledge, no study has analyzed the validity and
reliability of this IMS sensor to establish the reproducibility of velocity measurements at
different intensities and the accuracy in distinguishing the concentric and eccentric phases
compared to a linear encoder. Therefore, the aim of this study was to evaluate the validity
and reliability of the IMS to measure kinematic variables during squats performed on a
flywheel device.

2. Materials and Methods

2.1. Study Design
Participants of the study participated in three sessions: one familiarization and two
testing sessions separated by 72 h. During the testing sessions, participants were evalu-
ated through an incremental loading test for three different inertial loads and performed
4 repetitions for each load. Data were collected simultaneity from two linear position
sensors: (1) a linear-position transducer (LPT) which acted as criterion measure (1000 Hz
sampling rate, T-Force, Ergotest, Murcia, Spain) and (2) the IMS (1000 Hz sampling rate,
IonClinics & Deionic S.L, L’Alcudia, Spain). For the reliability analysis, the two fastest
Sensors 2023, 23, 2193 3 of 11

repetitions of each load set were used, so as not to artificially increase the reliability, they
were determined separately for each device. For the validation analysis, the two fastest
repetitions performed with each of the three loads were also used. The variables used for
the reliability and validation analyses were mean velocity (MV), mean propulsive velocity
(MPV) and max velocity (Vmax) that were simultaneously recorded with the LPT (i.e.,
criterion validity measure) and the IMS system.

2.2. Participants
Thirty-one healthy sports science undergraduate male students volunteered for the
study (24 ± 2.9 years, height: 1.74 ± 0.92 m, body mass: 69.4 ± 15.4 kg, and body fat
percentage: 16.5 ± 6.4%). They had, at least, one year of experience with flywheel training,
and no history of neurological disorders or lower limb orthopaedical injuries. None of
them were taking performance enhancing drugs, medications or other substances that
could alter their performance during testing. Moreover, they were instructed to record
and maintain their sleeping, eating, and drinking habits in the 48 h prior to each testing
session. They were informed of the purposes and risks involved in the study before giving
voluntary informed written consent to participate. The study procedures were performed
in accordance with the principles of the Declaration of Helsinki and were approved by the
local Institutional Review Board (H1421157445503).

2.3. Testing Procedures

During the familiarization session, participants practiced performing the flywheel
squat exercise on the Epte Inertial Concept (Ionclionis and Deionic, LAlcúdia, Valencia-
Spain) using three different inertial loads (2 sets of 4 repetitions with each load, with a 2-
minute rest periods between sets), including the loads that would be used in testing sessions.
During testing sessions, participants were evaluated through a randomized incremental
loading test in which they performed 3 sets of 4 repetitions at three different intensities.
The intensities used during testing were load 1: 0.02 kg·m2 , load 2: 0.05 kg·m2 , and load 3:
0.11 kg·m2 by different flywheels provided in the Epte Inertial Concept. The first repetition
was always used to start the movement and to print speed into the flywheel system, and
the other three repetitions were recorded for further analysis. A total of 837 repetitions
were analyzed. During the next three repetitions, participants were requested to push with
maximal effort (i.e., maximum possible concentric speed) through the entire concentric
phase, which ranged from 90◦ -knee flexion to near full extension. To ensure that volunteers
employed the same squat depth at each repetition, an adjustable tripod with a telemetric
photocell (Microgate, Bolzano, Italy) was placed at the side of the flywheel device [27].
The telemetry photocell emitted a sound when participants reached 90◦ knee flexion. The
randomized incremental loading test was replicated during testing session 2.
MV, MPV and Vmax for the concentric phase of the movement were collected simulta-
neously from the LPT (1000 Hz sampling rate, T-Force, Ergotest, Murcia-Spain) and the
IMS sensor (1000 Hz sampling rate, IonClinics & Deionic S.L, L’Alcudia-Spain). The LPT
was attached to the harness belt used for the exercise (Figure 1) and interfaced with a
personal computer using a 14-bit resolution analog-to-digital data acquisition board and
custom software. During each lift the instantaneous kinematic data were collected with a
sample frequency of 1000 Hz. Reliability of the LPT used as criterion validity measure was
previously reported [29].
The IMS sensor was a linear quadrature encoder (i.e., detecting the change of each
phase when changing the input channel assigned for each phase of the movement) inte-
grated into a secondary pulley through which the device cable passes, with the purpose to
detect the linear movement of the cable and the direction of the movement, working as a
traditional LPT (Figure 1). It collected data with a resolution of 48 points per revolution
(4.8 mm) and a frequency of 1000 Hz. The two fastest repetitions (i.e., higher concentric
MV) of each set, so as not to artificially increase the reliability, were determined separately
 dynamic stretching exercises (i.e., forward leg swings, ankle dorsi- and plantar-f
side leg swings, high knees, heel flicks, squats and lunges) was performed, accord
Cuenca-Fernández et al. [30]’s procedures (each exercise was performed for 20 s, a
entire set was repeated twice). Then, two sets of five continuous unloaded squa
Sensors 2023, 23, 2193 4 of 11
non‐jumping) interspersed by 30 s were performed at a rhythm of 2/2
tric/concentric) tempo and 1/1 tempo, respectively [31].
for each device and used for further analysis. A total of 558 repetitions were analyzed for
determining the reliability.

Figure 1. Squat exercise on the Epte Inertial Concept with a linear position transducer (LPT) and an
Inertial Measurement System (IMS).
Figure 1. Squat exercise on the Epte Inertial Concept with a linear position transducer (LPT)
Before
Inertial each session,
Measurement a warm-up
System (IMS).consisting of 5 min of cycling followed by 5 min of
dynamic stretching exercises (i.e., forward leg swings, ankle dorsi- and plantar-flexion, side
leg swings, high knees, heel flicks, squats and lunges) was performed, according to Cuenca-
2.4. Statistical Analyses
Fernández et al. [30]’s procedures (each exercise was performed for 20 s, and the entire set
All study
was repeated variables
twice). Then, two(MV, MPV
sets of and Vmax)
five continuous were squats
unloaded normally distributed accord
(i.e., non-jumping)
interspersed
the by 30 s were performed
Kolmogorov–Smirnoff test. at
The a rhythm of 2/2 (eccentric/concentric)
concurrent validation of the IMS tempo and the cr
versus
1/1 tempo, respectively [31].
measure LPT (gold standard) was assessed using ordinary least products (OLP)
sion [32]. Fixed
2.4. Statistical bias (a significant systematic difference) was deemed present if th
Analyses
confidence
All studyinterval (CI)
variables (MV,ofMPV
the intercept
and Vmax)did werenot include
normally zero, while
distributed proportional
according to
significant proportional difference) was deemed present if the 95% CI of the slope d
the Kolmogorov–Smirnoff test. The concurrent validation of the IMS versus the criterion
measure one
include LPT (gold
[33]. standard) was assessedcorrelation
Lin's concordance using ordinary least products
coefficient (CCC) (OLP)wasregres-
also used
sion [32]. Fixed bias (a significant systematic difference) was deemed present if the 95%
sessing concurrent validity between the devices [34,35]. The criteria
confidence interval (CI) of the intercept did not include zero, while proportional bias (a
used to det
the strength
significant of CCCdifference)
proportional were ‘almost perfect’
was deemed >0.99,
present if the 'substantial'
95% CI of the slope between
did not0.95 an
'moderate' between
include one [33]. 0.90 and 0.95
Lin’s concordance or ‘poor’
correlation <0.90(CCC)
coefficient [36]. was
Hedges’
also used g effect sizes we
to assess-
ing concurrent validity between the devices [34,35]. The criteria used
calculated to estimate the magnitude of differences between devices [37] and inter to determine the
strength of CCC were ‘almost perfect’ >0.99, ‘substantial’ between 0.95 and 0.99, ‘moderate’
as trivial (<0.2), small (0.2–0.5), moderate (>0.5–0.8) or large (>0.8) [38]. Reliabili
between 0.90 and 0.95 or ‘poor’ <0.90 [36]. Hedges’ g effect sizes were also calculated to
determined by the standard
estimate the magnitude error
of differences of measurement
between devices [37] and(SEM), the coefficient
interpreted of variatio
as trivial (<0.2),
and the intraclass correlation coefficient (ICC). The criteria used
small (0.2–0.5), moderate (>0.5–0.8) or large (>0.8) [38]. Reliability was determined by the to determine rel
standard
were a SEMerror of
< measurement
0.2, a CV < (SEM),10% andthe coefficient
an ICC of < variation
0.70 [39,40].(CV) and Thethevalidity
intraclassand rel
analyses were performed for each load. A statistical significance was<set
correlation coefficient (ICC). The criteria used to determine reliability were a SEM 0.2,at
a p < 0.0
CV < 10% and an ICC < 0.70 [39,40]. The validity and reliability analyses were performed
for each load. A statistical significance was set at p < 0.05 level and a confidence limit
was set at 95%. All statistical analyses, and accompanying plots, were performed using R
software (The R Project for Statistical Computing, version 4.2.0).
Sensors 2023, 23, 2193 5 of 11

3. Results
3.1. Validity
Descriptive statistics from both LPT and IMS sensor for each variable and load together
with Hedges’ g effect size are included in Table 1. Hedges’ g effect size analysis showed
no difference (<0.19) in any of the variables measured between both sensors. In addition,
Table 2 includes the validity test results, showing a very high association between the
measurements of both devices. The ordinary least products (OLP, intercept and slope data)
regression showed that no fixed or proportional bias for any variable at any load. Similarly,
Lin’s CCC showed values ranged from 0.96 to 0.99 for all variables studied, indicating that
the degree of agreement between both measurement devices was almost perfect (Table 2,
Figure 2).

Table 1. Mean concentric velocity (MV), maximum velocity (Vmax) and mean propulsive velocity
(MPV) at three different flywheel loads (0.02 kg·m2 , 0.05 kg·m2 , 0.11 kg·m2 ) measured by the T-Force
and IMS, and Hedge’s g effect size between T-Force and IMS. All data are presented as mean (SD) of
31 participants.

Load T-Force IMS Hedge’s g

0.02 kg·m2
MV (m·s−1 ) 0.998 (0.152) 0.977 (0.148) 0.141
Vmax (m·s−1 ) 1.560 (0.274) 1.596 (0.265) 0.133
MPV (m·s−1 ) 1.072 (0.168) 1.050 (0.169) 0.091
0.05 kg·m2
MV (m·s−1 ) 0.905 (0.140) 0.895 (0.134) 0.071
Vmax (m·s−1 ) 1.464 (0.208) 1.491 (0.210) 0.135
MPV (m·s−1 ) 0.971 (0.160) 0.962 (0.155) 0.061
0.11 kg·m2
MV (m·s−1 ) 0.689 ( 0.119) 0.713 (0.119) 0.191
Vmax (m·s−1 ) 1.301 (0.191) 1.363 (0.201) 0.136
MPV (m·s−1 ) 0.764 (0.129) 0.764 (0.127) 0.003

Table 2. Validity test results by ordinary least products (OLP) regression (Intercept, Slope) and
Lin’s concordance correlation coefficient (CCC) with 95% confidence interval (min, max) for mean
concentric velocity (MV), maximum velocity (Vmax) and mean propulsive velocity (MPV) at three
different flywheel loads (0.02 kg·m2 , 0.05 kg·m2 , 0.11 kg·m2 ) measured by the T-Force and IMS.

Variable Intercept Slope Lin’s CCC

0.02kg·m2
MV −0.00 (−0.06, 0.07) 1.02 (0.95, 1.08) 0.96 (0.94, 0.98)
Vmax −0.09 (−0.18, 0.01) 1.04 (0.97, 1.1) 0.97 (0.96, 0.99)
MPV 0.02 (−0.05, 0.19) 0.99 (0.92, 1.06) 0.96 (0.93, 0.98)
0.05 kg·m2
MV −0.03 (−0.07, 0.02) 1.04 (0.99, 1.09) 0.98 (0.98, 0.99)
Vmax −0.01 (−0.083, 0.042) 0.99 (0.954, 1.041) 0.99 (0.98, 0.99)
MPV −0.02 (−0.06, 0.03) 1.03 (0.98, 1.07) 0.99 (0.98, 0.99)
0.11 kg·m2
MV −0.03 (−0.06, 0.001) 1.01 (0.97, 1.05) 0.98 (0.97, 0.99)
Vmax 0.00 (−0.04, 0.05) 0.95 (0.92, 1.00) 0.96 (0.95, 0.98)
MPV −0.01 (−0.052, 0.018) 1.02 (0.98, 1,07) 0.98 (0.97, 0.99)
Sensors 2023, 23, x FOR PEER REVIEW 6 of 11
Sensors 2023, 23, 2193 6 of 11

Figure 2. Lin’s concordance correlation coefficient (CCC) between the T-Force device (X-axis) and
the
the IMS
IMS sensor
sensor (Y-axis)
(Y-axis) for
for mean
mean velocity
velocity (MV: a, d,
(MV: a, d, g),
g), maximum
maximum velocity
velocity (Vmax:
(Vmax: b,
b, e,
e, h)
h) and
and mean
mean
propulsive velocity (MPV: c, f, i). MV—0.02 kg·m 2 (a), MPV—0.02 kg·m2 (b), Vmax—0.02 kg·m2 (c),
propulsive velocity (MPV: c, f, i). MV—0.02 kg·m (a), MPV—0.02 kg·m (b), Vmax—0.02 kg·m2
2 2
MV—0.05 kg·m2 (d), MPV—0.05 kg·m2 (e),2 Vmax—0.05 kg·m2 (f),2MV—0.11 kg·m2 (g),2 MPV—0.11
(c), MV—0.05 kg·m2 (d), MPV—0.05 kg·m (e), Vmax—0.05 kg·m (f), MV—0.11 kg·m (g), MPV—
kg·m (h),2Vmax—0.11 kg·m2 (i). Data
2
2 are shown in m/s, CCC concordance correlation coefficient =
0.11 kg·m (h), Vmax—0.11 kg·m (i). Data are shown in m/s, CCC concordance correlation coeffi-
statistic value (Confidence Interval).
cient = statistic value (Confidence Interval).
3.2.
3.2. Reliability
Reliability
As
As shown
shown inin Table 3, all
Table 3, all variables
variables demonstrated
demonstrated good
good reliability
reliability between
between testing
testing day
day
1 and 2, regardless of exercise intensity and the measurement device
1 and 2, regardless of exercise intensity and the measurement device used (CV used (CV < 10% and
< 10%
ICC > 0.70).
and ICC BothBoth
> 0.70). devices exhibited
devices excellent
exhibited relative
excellent reliability
relative reliability(ICCs
(ICCs>>0.9)
0.9) and
and good
good
absolute reliability (CV < 5%) for each variable studied regardless of the load
absolute reliability (CV < 5%) for each variable studied regardless of the load tested (Tabletested (Ta-
3).
ble 3). In addition,
In addition, the standard
the standard error of error of measurement
measurement (SEM)(SEM) was lower
was lower thanfor
than 0.09 0.09each
for each
load
load
and forand for LTP
both bothand
LTP and
IMS IMS devices,
devices, indicating
indicating the highthe high precision
precision of the measure-
of the measurements.
ments.
Sensors 2023, 23, 2193 7 of 11

Table 3. Coefficient of variation (CV), Intraclass correlation (ICC) with 95% confidence interval (min,
max), and standard error of measurement (SEM) values for three loads (0.02 kg·m2 , 0.05 kg·m2 ,
0.11 kg·m2 ) for mean concentric velocity (MV), maximum velocity (Vmax) and mean propulsive
velocity measured by the T-Force and IMS.

Variable CV (%) ICC (95%CI) SEM

T-Force
MV—0.02 kg·m2 0.150 0.90 (0.80,0.95) 0.048
MV—0.05 kg·m2 0.187 0.94 (0.89,0.97) 0.041
MV—0.11 kg·m2 0.206 0.92 (0.84,0.96) 0.038
Vmax—0.02 kg·m2 0.173 0.95 (0.90,0.98) 0.061
Vmax—0.05 kg·m2 0.201 0.94 (0.88,0.97) 0.071
Vmax—0.11 kg·m2 0.186 0.95 (0.89,0.97) 0.054
MPV—0.02 kg·m2 0.155 0.87 (0.76,0.94) 0.061
MPV—0.05 kg·m2 0.196 0.95 (0.89,0.97) 0.042
MPV—0.11 kg·m2 0.190 0.93 (0.85,0.96) 0.039
IMS
MV—0.02 kg·m2 0.150 0.90 (0.80,0.95) 0.044
MV—0.05 kg·m2 0.183 0.89 (0.78,0.94) 0.054
MV—0.11 kg·m2 0.201 0.93 (0.85,0.96) 0.038
Vmax—0.02 kg·m2 0.163 0.95 (0.89,0.97) 0.059
Vmax—0.05 kg·m2 0.200 0.91 (0.82,0.95) 0.088
Vmax—0.11 kg·m2 0.187 0.95 (0.89,0.97) 0.057
MPV—0.02 kg·m2 0.156 0.88 (0.77,0.94) 0.059
MPV—0.05 kg·m2 0.194 0.92 (0.84,0.96) 0.053
MPV—0.11 kg·m2 0.187 0.90 (0.81,0.95) 0.045

4. Discussion
The results of this study showed that the IMS could be used as a valid and reliable
system to control execution velocity during flywheel training. It was observed that the
IMS sensor differentiated between concentric and eccentric phase; therefore, it is a valid
and reliable tool for the measurements of MV, Vmax, MPV and ROM. The three loads
tested showed Lin’s CCC values between 0.96 and 0.98 for MV, 0.96 and 0.99 for Vmax
and 0.96 and 0.99 for MPV between the LPT criterion measures and the IMS (Table 2,
Figure 2). Similarly, the IMS sensor showed significant ICC values (ranging from 0.88 to
0.95) for each load and output (Table 3). In addition, the standard error of measurement
and CV were lower than 0.09 and 0.2%, respectively, for each tested load (see Table 3).
Thus, this technology provides a better way to prescribe flywheel velocity-based exercise,
hence providing information that can maximize training outcomes and reduce the risks
of overtraining. In addition, the results of the present study permit the use of flywheel
devices for testing when a IMS sensor is used.
Resistance training intensity has been traditionally prescribed relative to the maximum
strength (based on the percentage of one-repetition maximum (1-RM)) [9], or based on an
individual’s ability to lift the load at a given velocity [4]. However, in the case of flywheel
technology, it is not possible to determine the 1-RM since there is not a maximum load
that can be lifted [22]. Instead, during a flywheel exercise, the user needs to accelerate the
inertia of the flywheel through a maximum concentric muscle contraction, and then apply
a braking contraction to return the stored energy and invert the movement of the flywheel
during the following eccentric phase of the exercise [22]. Therefore, considering the maxi-
mum nature of each concentric contraction together with the need to know the magnitude
of the load during the eccentric contraction (i.e., eccentric:concentric ratio), concentric linear
velocity is monitored during flywheel exercises [3,7,22]. Although concentric and eccentric
velocity can be used to effectively prescribe exercise intensity [3], their accurate monitoring
is quite complicated and requires the use of advanced technologies (e.g., 3D motion capture
or quadrature encoders) that are not commonly available to practitioners.
Sensors 2023, 23, 2193 8 of 11

This has prompted practitioners to use rotary encoders (integrated into the axis of
the system), which provide information about the angular velocity of the flywheel [23,24].
However, a dissonance between what happens on the axis (i.e., angular velocity in a conical
cylinder flywheel device) [41,42] and what happens at kinematic level while the participant
is performing the exercise should be considered, which limits the use of velocity for such
purposes. Although angular velocity can be converted to linear velocity mathematically,
this relationship relies on a known radius (which can vary within some devices) and the
lack of slack in cable/rope. As preliminary research has shown, rotary encoders function
as a low-pass filter (i.e., smoothing the oscillations of high-frequency velocity actions [26],
which is even accentuated when higher inertial loads are used, smoothing the instantaneous
velocity record [43]. This could affect the calculations used to determine the mechanical
power executed during the different phases of the movement and other estimated kinetic
outcomes, altering not only the validity of these data, but also being responsible for the low
reproducibility demonstrated when compared with LPT or force plates [23,24]. Moreover,
rotary encoders do not allow the precise measurement of the concentric acceleration (or
MPV in terms of velocity, impulse in kinetic terms) that the subject experiences during the
accelerating phase of the concentric muscle contraction which determines the nature of
the flywheel stimulus, and hence, the possibility of kinetically overloading the eccentric
action [3]. It should also be considered that a rotary encoder does not physically differentiate
between the concentric and eccentric phase of the movement [7]. This differentiation is
usually performed by software prediction, which may lead to estimation errors (normally
underestimating the concentric velocity). Moreover, in flywheel exercises performed with
a load lower than 78% 1RM [5], rotary encoders will likely assign part of the concentric
phase to the eccentric one, since participants usually stop applying force before completing
the concentric contraction to avoid jumping, which may lead to erroneous estimates when
controlling training load [2]. For all these reasons, it is necessary to validate a measurement
instrument that can adequately differentiate between the concentric and eccentric phases
during flywheel exercise, and thus can accurately quantify concentric linear velocity.
The results of the present study showed that the measurement of linear concentric
velocity by means of an IMS sensor was a valid strategy for the control of flywheel exer-
cise. Compared with a previously validated LPT device, which acted as a gold standard,
Hedge’s effect sizes between 0.07–0.14 for MV, 0.13–0.14 for Vmax and 0.00–0.09 for MPV
were observed at the three inertial loads tested. The validity of the IMS device has also
been contrasted through the correlations of Lin’s CCC (ranging from 0.96 to 0.99 for all
variables and loads) and OLP intercept (ranging from −0.09 to 0.02 for all variables and
loads) and slope (ranging from 0.95 to 1.04 for all variables and loads) analysis, demon-
strating an excellent correlation between the IMS and the LPT devices on MV, Vmax and
MPV data (Table 2, Figure 2). Furthermore, it should be noted that these two statistical
analyses (Lin’s CCC and OLP), despite having been used separately in previous validation
studies [31,44,45], were for the first time applied simultaneously in the present study. Fur-
thermore, this was the first study to graphically represent Lin’s CCC (Figure 2). Despite
this, we cannot affirm that the IMS device has greater validity than other encoders, since
the studies in which the validity of rotatory encoders [24,28] was studied have not applied
a similar statistical treatment. Similarly, the IMS device showed excellent reproducibility
between the two measurements analyzed for the measurement of MV (CV: 0.15–0.20%;
ICC: 0.89–0.93; SEM: 0.04–0.05), Vmax (CV: 0.16–0.20%; ICC: 0.91–0.95; SEM: 0.06–0.09) and
MPV (CV: 0.16–0.19%; ICC: 0.88–0.92; SEM: 0.04–0.06), showing a high similarity with the
results obtained by the criterion measure LPT (Table 3). The present study showed similar
values to those obtained by other articles about the validation and reliability of different
devices for measuring exercise velocity in the variables studied (MV, MPV and Vmax), and
similar results for SEM, CV, ICC and CCC, which coincide with the values considered as
indicators of high similarity between results and high reliability [31,44,45].
However, some potential delimitations are worthy of mention. Despite the sample
being homogeneous with respect to age, sex, strength level, and training experience,
Sensors 2023, 23, 2193 9 of 11

it is not known whether similar results would be found in women or individuals less
exercise training experience. Furthermore, the inclusion of other parameters such as the
measurement of the ground reactive force with a force plate system or the measurement of
other kinematic parameters such as the eccentric velocity or the eccentric:concentric ratio is
warranted. Future research is warranted to contribute to the knowledge and development
of load-velocity profiles and their particularities in flywheel training by measuring linear
velocity instead of angular velocity, as well as to develop measuring instruments that
facilitate flywheel velocity-based training in order to individually monitor and adjust
training volume and intensity in a better way to fit the needs of each participant.

5. Conclusions
In conclusion, the IMS was a valid and reliable tool for measuring concentric linear
velocity during flywheel training. In comparison with a previous validated LPT, the
IMS sensor provided almost the same values for MV, Vmax, MPV and ROM, as well as
trivial Hedge’s g values for each outcome. In addition, the IMS presented good test–retest
reliability for all loads tested between two different testing sessions. For the first time, this
study provided evidence that the IMS accurately differentiated between concentric and
eccentric phases of the movement during flywheel squat exercises. This in turn, enables
practitioners to precisely estimate other kinetic values, such as eccentric:concentric ratio.
In addition, it provides the feasible assessment of mean propulsive velocity, since this
parameter is decisive in characterizing the nature of flywheel exercise. Therefore, the use of
an IMS has the potential to be used as an avenue of intensity prescription, as well as provide
valid and reliable data to monitor flywheel training and to control the training process.

Author Contributions: Conceptualization: S.M.-I., J.A.-G. and F.M.-R.; data collection: S.M.-I. and
F.M.-R.; software: S.M.-I., J.A.-G., F.M.-R.; validation: S.M.-I., J.A.-G., K.N. and F.M.-R.; data analyses,
S.M.-I., J.A.-G., K.N. and F.M.-R.; manuscript: S.M.-I., K.N. and F.M.-R. All authors have read and
agreed to the published version of the manuscript.
Funding: This research received no external funding.
Institutional Review Board Statement: The study was conducted in accordance with the Decla-
ration of Helsinki and approved by the Institutional Ethics Committee of University of Valencia
(H1421157445503) for studies involving humans.
Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.
Data Availability Statement: The data presented in this study are available on request from the
corresponding author. The data are not publicly available due to participants’ privacy.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Jonathon, W.; Bryan, M.; Harry, B.; Shaun, M.; Tannath, S.; Amador, G.-R. Velocity-Based Training: From Theory to Application.
Strength Cond. J. 2021, 43, 31–49. [CrossRef]
2. González-Badillo, J.J.; Sánchez-Medina, L.; Ribas-Serna, J.; Rodríguez-Rosell, D. Toward a New Paradigm in Resistance Training
by Means of Velocity Monitoring: A Critical and Challenging Narrative. Sport Med.—Open 2022, 8, 118. [CrossRef] [PubMed]
3. Carroll, K.M.; Wagle, J.P.; Sato, K.; Taber, C.B.; Yoshida, N.; Bingham, G.E.; Stone, M.H. Characterising Overload in Inertial
Flywheel Devices for Use in Exercise Training. Sports Biomech. 2019, 18, 390–401. [CrossRef] [PubMed]
4. Banyard, H.G.; Tufano, J.J.; Delgado, J.; Thompson, S.W.; Nosaka, K. Comparison of the Effects of Velocity-Based Training
Methods and Traditional 1RM-Percent-Based Training Prescription on Acute Kinetic and Kinematic Variables. Int. J. Sports Physiol.
Perform. 2019, 4, 246–255. [CrossRef] [PubMed]
5. Sanchez-Medina, L.; Perez, C.E.; Gonzalez-Badillo, J.J. Importance of the propulsive phase in strength assessment. Int. J. Sports
Med. 2010, 31, 123–129. [CrossRef] [PubMed]
6. Balsalobre-Fernández, C.; Marchante, D.; Muñoz-López, M.; Jiménez, S.L. Validity and reliability of a novel iPhone app for the
measurement of barbell velocity and 1RM on the bench-press exercise. J. Sports Sci. 2018, 36, 64–70. [CrossRef] [PubMed]
7. Martín-Rivera, F.; Beato, M.; Alepuz-Moner, V.; Maroto-Izquierdo, S. Use of concentric linear velocity to monitor flywheel exercise
load. Front. Physiol. 2022, 13, 961572. [CrossRef] [PubMed]
Sensors 2023, 23, 2193 10 of 11

8. Balsalobre-Fernández, C.; Marchante, D.; Baz-Valle, E.; Alonso-Molero, I.; Jiménez, S.L.; Muñóz-López, M. Analysis of Wearable
and Smartphone-Based Technologies for the Measurement of Barbell Velocity in Different Resistance Training Exercises. Front.
Physiol. 2017, 8, 649. [CrossRef]
9. Suchomel, T.J.; Nimphius, S.; Bellon, C.R.; Stone, M.H. The Importance of Muscular Strength: Training Considerations. Sports
Med. 2018, 48, 765–785. [CrossRef]
10. Douglas, J.; Pearson, S.; Ross, A.; McGuigan, M. Chronic Adaptations to Eccentric Training: A Systematic Review. Sports Med.
2017, 47, 917–941. [CrossRef]
11. Suchomel, T.J.; Wagle, J.P.; Douglas, J.; Taber, C.B.; Harden, M.; Haff, G.G.; Stone, M.H. Implementing Eccentric Resistance
Training—Part 1: A Brief Review of Existing Methods. J. Funct. Morphol. Kinesiol. 2019, 4, 38. [CrossRef] [PubMed]
12. Hernández-Davó, J.L.; Sabido, R.; Blazevich, A.J. High-speed stretch-shortening cycle exercises as a strategy to provide eccentric
overload during resistance training. Scand. J. Med. Sci. Sports 2021, 31, 2211–2220. [CrossRef] [PubMed]
13. Wagle, J.P.; Taber, C.B.; Cunanan, A.J.; Bingham, G.E.; Carroll, K.M.; Deweese, B.H.; Sato, K.; Stone, M.H. Accentuated eccentric
loading for training and performance: A review. Sports Med. 2017, 47, 2473–2495. [CrossRef] [PubMed]
14. Maroto-Izquierdo, S.; García-López, D.; Fernandez-Gonzalo, R.; Moreira, O.C.; González-Gallego, J.; de Paz, J.A. Skeletal muscle
functional and structural adaptations after eccentric overload flywheel resistance training: A systematic review and meta-analysis.
J. Sci. Med. Sport 2017, 20, 943–951. [CrossRef] [PubMed]
15. Marco, B.; Antonio, D.I. Implementing Flywheel (Isoinertial) Exercise in Strength Training: Current Evidence, Practical Recom-
mendations, and Future Directions. Front. Physiol. 2020, 11, 569. [CrossRef]
16. Fernandez-Gonzalo, R.; Lundberg, T.R.; Alvarez-Alvarez, L.; de Paz, J.A. Muscle damage responses and adaptations to eccentric-
overload resistance exercise in men and women. Eur. J. Appl. Physiol. 2014, 114, 1075–1084. [CrossRef]
17. de Keijzer, K.L.; McErlain-Naylor, S.A.; Brownlee, T.E.; Raya-González, J.; Beato, M. Perception and application of flywheel
training by professional soccer practitioners. Biol. Sport 2022, 39, 809–817. [CrossRef]
18. Beato, M.; Bigby, A.E.J.; De Keijzer, K.L.; Nakamura, F.Y.; Coratella, G.; McErlain-Naylor, S.A. Post-activation potentiation effect
of eccentric overload and traditional weightlifting exercise on jumping and sprinting performance in male athletes. PLoS ONE
2019, 14, 0222466. [CrossRef]
19. Beato, M.; Maroto-Izquierdo, S.; Turner, A.N.; Bishop, C. Implementing strength training strategies for injury prevention in soccer:
Scientific rationale and methodological recommendations. Int. J. Sports Physiol. Perform. 2021, 16, 456–461. [CrossRef]
20. Tesch, P.A.; Fernandez-Gonzalo, R.; Lundberg, T.R. Clinical Applications of Iso-Inertial, Eccentric-Overload (YoYo™) Resistance
Exercise. Front. Physiol. 2017, 27, 241. [CrossRef]
21. Maroto-Izquierdo, S.; McBride, J.; Gonzalez-Diez, N.; García-López, D.; González-Gallego, J.; de Paz, J.A. Comparison of Flywheel
and Pneumatic Training on Hypertrophy, Strength, and Power in Professional Handball Players. Res. Q. Exerc. Sport 2022, 93,
1–15. [CrossRef] [PubMed]
22. Maroto-Izquierdo, S.; Raya-González, J.; Hernández-Davó, J.L.; Beato, M. Load Quantification and Testing Using Flywheel
Devices in Sports. Front. Physiol. 2021, 12, 739399. [CrossRef]
23. Weakley, J.; Till, K.; Sampson, J.; Banyard, H.; Leduc, C.; Wilson, K.; Roe, G.; Jones, B. The Effects of Augmented Feedback on
Sprint, Jump, and Strength Adaptations in Rugby Union Players After a 4-Week Training Program. Int. J. Sports Physiol. Perform.
2019, 29, 1205–1211. [CrossRef]
24. Bollinger, L.M.; Brantley, J.T.; Tarlton, J.K.; Baker, P.A.; Seay, R.F.; Abel, M.G. Construct validity, test-retest reliability, and
repeatability of performance variables using a flywheel resistance training device. J. Strength Cond. Res. 2020, 34, 3149–3156.
[CrossRef] [PubMed]
25. Beato, M.; Maroto-Izquierdo, S.; Hernández-Davó, J.L.; Raya-González, J. Flywheel Training Periodization in Team Sports. Front.
Physiol. 2021, 12, 732802. [CrossRef] [PubMed]
26. Yan, Z.; Zhang, X.-P. General Energy Filters for Power Smoothing, Tracking and Processing Using Energy Storage. IEEE Access
2017, 5, 19373–19382. [CrossRef]
27. Maroto-Izquierdo, S.; Bautista, I.; Rivera, F. Post-activation performance enhancement (PAPE) after a single-bout of high-intensity
flywheel resistance training. Biol. Sport 2020, 37, 343–350. [CrossRef]
28. Martín-San Agustín, R.; Sánchez-Barbadora, M.; García-Vidal, J.A. Validity of an inertial system for measuring velocity, force, and
power during hamstring exercises performed on a flywheel resistance training device. PeerJ 2020, 8, e10169. [CrossRef]
29. Courel-Ibáñez, J.; Martínez-Cava, A.; Morán-Navarro, R.; Escribano-Peñas, P.; Chavarren-Cabrero, J.; González-Badillo, J.J.;
Pallarés, J.G. Reproducibility and Repeatability of Five Different Technologies for Bar Velocity Measurement in Resistance
Training. Ann. Biomed. Eng. 2019, 47, 1523–1538. [CrossRef]
30. Cuenca-Fernández, F.; López-Contreras, G.; Arellano, R. Effect on Swimming Start Performance of Two Types of Activation
Protocols: Lunge and YoYo Squat. J. Strength Cond. Res. 2015, 29, 647–655. [CrossRef]
31. Mina, M.A.; Blazevich, A.J.; Tsatalas, T.; Giakas, G.; Seitz, L.B.; Kay, A.D. Variable, But not Free-Weight, Resistance back Squat
Exercise Potentiates Jump Performance Following a Comprehensive Task-Specific Warm-Up. Scand. J. Med. Sci. Sports 2018, 29,
380–392. [CrossRef] [PubMed]
32. Ludbrook, J. Statistical techniques for comparing measurers and methods of measurement: A critical review. Clin. Exp. Pharmacol.
Physiol. 2002, 29, 527–536. [CrossRef] [PubMed]
Sensors 2023, 23, 2193 11 of 11

33. Ludbrook, J. A primer for biomedical scientists on how to execute model ii linear regression analysis. Clin. Exp. Pharmacol. Physiol.
2012, 39, 329–335. [CrossRef] [PubMed]
34. Lin, L.I.-K. A concordance correlation coefficient to evaluate reproducibility. Biometrics 1989, 45, 255–268. [CrossRef]
35. Lin, L.I.-K. A note on the concordance correlation coefficient. Biometrics 2000, 56, 324–325.
36. Cortés-Reyes, É.; Rubio-Romero, J.A.; Gaitán-Duarte, H. Statistical methods of assessing the concordance and reproductibility of
diagnostic tests. Rev. Colomb. De Obstet. Y Ginecol. 2010, 61, 247–255. [CrossRef]
37. Hedges, L.V.; Olkin, I. Estimation of a single effect size: Parametric and nonparametric methods. In Statistical Methods for
Meta-Analysis; Academic Press: San Diego, CA, USA, 1985; pp. 75–106. [CrossRef]
38. Cohen, J. The t test for means. In Statistical Power Analysis for the Behavioral Sciences; Lawrence Earlbaum Associates: Hilldale, NJ,
USA, 1988; pp. 19–74.
39. Duthie, G.; Pyne, D.; Hooper, S. The reliability of video-based time motion analysis. J. Hum. Mov. Stud. 2003, 44, 259–272.
40. Cormack, S.J.; Newton, R.U.; McGuigan, M.R. Reliability of measures obtained during single and repeated countermovement
jumps. Int. J. Sports Physiol. Perform. 2008, 3, 131–144. [CrossRef]
41. Sabido, R.; Hernández-Davó, J.L.; Pereyra-Gerber, G.T. Influence of Different Inertial Loads on Basic Training Variables During
the Flywheel Squat Exercise. Int. J. Sports Physiol. Perform. 2018, 13, 482–489. [CrossRef]
42. Muñoz-López, A.; Floría, P.; Sañudo, B.; Pecci, J.; Carmona Pérez, J.; Pozzo, M. The Maximum Flywheel Load: A Novel Index to
Monitor Loading Intensity of Flywheel Devices. Sensors 2021, 21, 8124. [CrossRef]
43. Rojas-Delgado, B.; Alonso, M.; Amaris, H.; de Santiago, J. Wavepower output smoothing through the use of a high-speed kinetic
buffer. Energies 2019, 12, 2196. [CrossRef]
44. Peña García-Orea, G.; Belando-Pedreño, N.; Merino-Barrero, J.A.; Jiménez-Ruiz, A.; Heredia-Elvar, J.R. Validation of an opto-
electronic instrument for the measurement of weighted countermovement jump execution velocity. Sports Biomech. 2021, 20,
150–164. [CrossRef] [PubMed]
45. Peña García-Orea, G.; Belando-Pedreño, N.; Merino-Barrero, J.A.; Heredia-Elvar, J.R. Validation of an opto-electronic instrument
for the measurement of execution velocity in squat exercise. Sports Biomech. 2021, 20, 706–719. [CrossRef] [PubMed]

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.