Conference PaperPDF Available

Sprinter Physiological State Measurement a Wearable Heart-Rate Sensor

Authors:

Abstract and Figures

Quantifying athletes' training helps to improve the efficiency of the training. Though training quantification using multi-modal sensing has been widely applied recently, few cases dealt with anaerobic sports such as short-distance running. This study attempts to provide a quantitative evaluation of the track and field short-distance running using a wearable heart-rate sensor, by gauging the relation between the autonomic nervous system activity and various types of training.
Content may be subject to copyright.
Sprinter Physiological State Measurement
a Wearable Heart-Rate Sensor
Chihiro Yoshida
Aoyama Gakuin University
5-10-1, Fuchinobe, Chuo-ku,
Kanagawa, Japan
tyoshida@wil-aoyama.jp
Junji Takahashi
Aoyama Gakuin University
5-10-1, Fuchinobe, Chuo-ku,
Kanagawa, Japan
takahashi@it.aoyama.jp
Guillaume Lopez
Aoyama Gakuin University
5-10-1, Fuchinobe, Chuo-ku,
Kanagawa, Japan
guillaume@it.aoyama.ac.jp
Naoya Isoyama
Aoyama Gakuin University
5-10-1, Fuchinobe, Chuo-ku,
Kanagawa, Japan
isoyama@it.aoyama.ac.jp
ABSTRACT
Quantifying athletes’ training helps to improve the efficiency of
the training. Though training quantification using multimodal
sensing has been widely applied recently, few cases dealt with
anaerobic sports such as short-distance running. This study
attempts to provide a quantitative evaluation of the track and field
short-distance running using a wearable heart-rate sensor, by
gauging the relation between the autonomic nervous system
activity and various types of training.
CCS Concepts
Mathematics of computing~Time series
analysis Information systems~Mobile information processing
systemsApplied computing~Health informatics
Keywords
Wearable sensor, sprint, heart rate variability analysis, autonomic
nervous system activity
1. INTRODUCTION
In recent years, lifestyle related diseases and associated health
problems have increased. To deal with such problems, a balanced
lifestyle through physical exercise is highly important. Therefore,
much focus has been placed on wearable activity measuring
devices such as pedometers to monitor people’s physical activities.
Such devices digitize and continuously monitor various daily
physical activities. Wearable devices can also be used for
performance improvement in sports. People practice sports in
order to become better; thus, quantitatively understanding the
change in performance is paramount for sports training as it could
help discover the evolution of one’s physiological and physical
performance. Presently, however, in many situations, only
physical performances of the athletes are measured and
physiological performances are entirely ignored.
In this study, the athlete’s heartbeat and energy consumption
are explored to study both his physiological and physical
performance. A wearable heart-rate sensor is used to measure the
heart-rate during short-track athletics (sprint) practice and during
rest to understand their influence to the performance of an athlete.
To achieve this, we use indices extracted from the spectral
analysis of heart rate variability (HRV), to estimate the athlete’s
physiological state. Indeed, HRV frequency power spectrum
reflects the activity of the autonomic nervous system (ANS).
Applying fast Fourier transform (FFT) to data of HRV, the power
spectrum is calculated. Thereafter, the low frequencies (LF) and
high frequencies (HF) of the electrocardiogram are calculated. In
this study, LF is the sum of PS for frequencies between 0.05 and
0.15 Hz, while HF is the sum of PS for frequencies between 0.15
and 0.35 Hz. LF and HF have a connection with respectively the
sympathetic nervous system and the parasympathetic nervous
system. When the sympathetic nervous system (SNS) activity is
dominant, the value of LF/HF is high. Likewise, when the
parasympathetic nervous system activity is dominant, the value of
LF/HF is small. Usually, SNS activity is regarded as dominant
during physical exercises. We report hereafter how ANS activity
changes in relation to various sprint training activities, using the
LF/HF index
2. PREVIOUS RESEARCHES
Hayashi et. al have studied the influence of the sleep time and
the oxygen pressure drop on the autonomic nervous activity of
track and field’s athletes [1].Theeffect on exercise capacity
improvement has been expected for high altitude training as an
important training method that can be applied to a variety of
sports. Though there is a variety in altitude training, "living in
high altitude area and training in low altitude area" is common in
many types of research. Researchers have created the same high
altitude environment and measured the autonomic nervous system
activity with a heart rate monitor. It was concluded that sleeping
in a high altitude increases parasympathetic nervous system
activity.
Another research was conducted on cardio-respiratory function
maintenance and improvement to deal with health problems,
healthy lifestyle promotion, and obesity reduction [2]. This study
uses aerobic exercise, which is a low-intensity workout. The
research investigated and compared the level of oxygen intake
during training exercises. The research could not, however,
observe any significant difference in oxygen intake for the
different exercises.
Nevertheless, the aforementioned researches have been carried
out under aerobic exercise conditions. Moreover, currently
available wearable activity measurement devices have also been
used to measure calories consumption and heart rate in regular
exercise amount such long distance running that is considering
aerobic exercise conditions. In addition, at the moment no study
has measured short-time high power anaerobic exercises such as
short-track running (sprint). This is mainly due to its short running
duration and distance.
3. EXPERIMENT
3.1 Sensor Selection
In this study, wearable heart rate sensor “myBeat” (Union Tool
co.) was used as shown in figure 1 [3]. This sensor is attached to
the person’s chest using adhesive electrodes pads that are clipped
to the back of the sensor. The sensor can measure three types of
physiological information: body surface temperature, 3 axis
acceleration, and heart beats. From the heartbeats, continuous
heart rate and heartbeats interval are automatically calculated. The
sensor can transmit the acquired data via an embedded 2.4GHz
wireless transmitter for a real time monitoring, or, using the
16MB built-in flash memory, the data can be recorded. In the
latter scenario, the recorded data can be offline viewed and
analyzed using the supplied software that runs on a personal
computer (PC).
The experiment uses four male university athletes in their
twenties from Aoyama Gakuin University Track and Field Club,
specialized in short-track running. Figure 2 shows a typical
attachment of the wearable heart rate sensor (myBeat) to the left
side of the chest. Sensor and electrodes pads were sturdily
attached to the body using surgical tape, to reduce motion artifacts
as much as possible, and to avoid electrodes detachment due to
sweating during training.
Figure 1: Union tool co. sensor “myBeat”
Figure 2: Mounting the “myBeat”
3.2 Sensor Comparison
Since in short-track running high power and short time training
is carried out, it is necessary to consider how much motion artifact
would affect the heart rate variations. Before starting the main
experiments, we compared the performances of myBeat to another
heart rate sensor. An ideal device must be able to measure the
heart rate continuously (beat by beat), and serviceable in sprint
conditions. Micoach smart run” (Adidas co.), a smartwatch with a
built-in heart rate monitor shown in figure 3 meets these
conditions [4]. Figure 4 shows a comparison between myBeat and
micoach smart run. As one can observe, the two sensors perform
almost equally in terms of heart rate monitoring. For further
comparison, Table 1 provides average heart rates and correlation
coefficients measured by the two sensors.
Figure 3: Adidas co. “micoach smart run”
Figure 4: Performance Comparision of tmyBeat and micoach
smart run
Table 1. Correlation coefficient between the measurement
data and average heart rate of both sensors evaluated three
times
Correlation
Coefficient myBeat Micoach
Smart Run
0.81 115 115
0.77 106 105
0.65 113 113
3.3 Experimental methods
Each subject wore the wearable heart rate sensor at the
beginning to the training. At the end of the training, the recording
is stopped and the sensor detached. Afterward, the recorded data,
which is saved on the built-in memory, is extracted using a PC
dedicated software that enables to log it in CSV file format. Then,
the physiological data (heart rate and LF/HF) was segmented and
analyzed. This process is summed-up in the flow chart presented
in figure 5.
Figure 5: Experimental methods flowchart
After gathering the data, PC dedicated viewer software is used
to check the heart rate waveform. Figure 6 show a sample plot of
the data (top: heart rate, bottom: LF/HF ratio) with superposed
segmentation marking corresponding to each training activity that
will be analyzed.
Figure 6. Measurement plot viewer software output (top:
heart rate, bottom: LF/HF ratio) with training contents
segmentation
3.4 Results
Figure 7 shows a comparative mean of 4 subjects’s(A,B,C,D)
LF/LH for different workouts, namely stretch (up), dash, three
times stretch (down) that all the subjects undertook. It is
noticeable that the activity of the sympathetic nerve is more
pronounced during exercise (dash). It decreases otherwise
(stretch). In addition, it is obvious that there is a meaningful
difference between the average LF/LH during the stretch (up) and
the dash workout (student’s t-test significance level 0.1).
Figure 7: Comparison of the Autonomous Nervous System
Activity Evaluation Index for the three Workout Exercises
A more detailed analysis is shown in figure 8, which shows a
detailed analysis of subject A from whom more data have been
collected. The investigated training contents were respectively:
warming-up (up), stretch (up), dash, dash (spike), start dash 30
meters (SD30), cooling down (down), and stretch (down). Seven
training contents were sorted from high to low values of LF/HF of
the heart rate variations during each training segment. The
analysis of the stretch (up) and stretch (down) are considered as
resting state, while up, down, dash, dash (spike), and SD30 are
considered as exercising state.
Though down showed higher LF/HF than stretch (down), as
showed in figure 7. In figure 8, also we can observe that
sympathetic nervous system activity was significantly higher at
rest than during training. As shown in figure 8, the experiment
proved statistically significant differences among six types of
training activities: 0.05 significance level for stretch (up) and dash
(spike), dash (spike) and down, stretch (down) and dash (spike),
and 0.1 significance level for stretch (up) and stretch (down),
stretch (up) and dash (spike).
Figure 8: Comparison of the ANS Activity Index for various
workout activities for Subject A.
4. CONCLUSIONS
The initial assumption was that the activity of sympathetic
nervous system would become active during sprint exercises
because it was expected that intense physical activity and
concentration during exercises would boost the autonomous
nervous system activity. However, it was found that, at rest, the
autonomic nervous activity index LF/HF were higher than during
spring exercise time. This may be due to a higher mental
concentration at rest that increases LF/HF. Also, it is not clear if
this behavior applies equally to both male and female alike since
this study involved only male. Thus, further research is needed
examine if there are any discrepancies. Finally, we plan to connect
our findings with knowledge about methods to control autonomic
nervous system activity and develop a smartphone tool that will
enable automatic and on site quick feedback to the athlete during
training, to support performance level improvement.
5. REFERENCES
[1] M. Hayashi, “The Effects of Nocturnal Sleep during the
Normobaric Hypoxic Environment on the Autonomic
Nervous Activity and the Condition in the Hakone-Ekiden
Athletes” Tokai J.Sports Med. Sci. No. 27, 43-49, 2015
[2] I. Han, “Effects of intermittent bouts of aerobic exercise on
oxygen consumption during and after exercise” NSSU
Journal of Sport Sciences, Vol. 1, 1-7, 2012
[3] UNION TOOL CO. “Wearable Heart rate Sensor myBeat”
http://www.uniontool.co.jp/en/product/sensor/
[4] SOLE COLLECTOR ”adidas miCoach Smart Run Launches
Today”
http://solecollector.com/news/adidas-micoarch-smart-run-
launches-today/
ResearchGate has not been able to resolve any citations for this publication.
The Effects of Nocturnal Sleep during the Normobaric Hypoxic Environment on the Autonomic Nervous Activity and the Condition in the Hakone-Ekiden Athletes
  • M Hayashi
M. Hayashi, "The Effects of Nocturnal Sleep during the Normobaric Hypoxic Environment on the Autonomic Nervous Activity and the Condition in the Hakone-Ekiden Athletes" Tokai J.Sports Med. Sci. No. 27, 43-49, 2015
Wearable Heart rate Sensor myBeat
  • Union
  • Co
UNION TOOL CO. "Wearable Heart rate Sensor myBeat" http://www.uniontool.co.jp/en/product/sensor/
Effects of intermittent bouts of aerobic exercise on oxygen consumption during and after exercise
  • I Han
I. Han, "Effects of intermittent bouts of aerobic exercise on oxygen consumption during and after exercise" NSSU Journal of Sport Sciences, Vol. 1, 1-7, 2012