Start Submission Become a Reviewer

Reading: Pattern-Matching Kinematic Analysis of Glide Phase after Start with Different Techniques in ...

Download

A- A+
Alt. Display

Research

Pattern-Matching Kinematic Analysis of Glide Phase after Start with Different Techniques in Medley Swimming: An Olympic Champion Case

Author:

Shun Wang

Beijing Sport University, Beijing, CN
X close

Abstract

Background: The analysis of kinematic characteristics in the glide phase after the start of swimming is significantly valuable for the technical optimization of competition. However, in the traditional analysis of kinematic parameters in time series are generally similar in the time axis, but their corresponding relationship is unknown.

Objective: The objective of this study is to identify the difference in kinematic parameters in underwater phases of hands-free start and hands-together start by using a case of an Olympic champion in men’s 200 m individual medley competition, through the dynamic time warp (DTW) pattern-matching approach.

Methods: An Olympic gold medal winner in the men’s 200 m individual medley competition is the subject of this study. The start techniques of hands-together and hands-free were included in the comparison. A Marker-Free Subaqueous Motion Recognition System was used to capture the motion, and nonlinear regressions between each kinematic parameter were made according to their theoretical physical relationships. And the DTW was applied to conduct the pattern matching.

Results: The majority of the resistance loads on the swimmer’s body in the subaqueous glide phase is the pressure drag of the water, and the swimmer will have a higher instantaneous and average velocity at the end of the start when the swimmer keeps his or her hands together and arms straight forward in the subaqueous glide phase due to the dynamic benefit prepared during the whole glide phase.

Conclusion: A start technique with hands-together and arms straight forwards could create mechanical and kinematical advantages for the swimmer, and the advantages might come from the optimal preload for the muscle activation at the end of the glide phase.

How to Cite: Wang, S. (2022). Pattern-Matching Kinematic Analysis of Glide Phase after Start with Different Techniques in Medley Swimming: An Olympic Champion Case. Physical Activity and Health, 6(1), 246–256. DOI: http://doi.org/10.5334/paah.201
54
Views
28
Downloads
  Published on 03 Nov 2022
 Accepted on 12 Sep 2022            Submitted on 18 Aug 2022

Introduction

From the beginning of the 1980s, the start of swimming has been intensely discussed (Counsilman, 1980). In 1985, Navarro and Arellano (Navarro and Arellano, 1985) recommended focusing on the start as one of the main factors for improving swimming performance, this suggestion was supported by Pereira in 2001 (Pereira, 2001). Since it is necessary to evaluate the starting time to understand the performance of the swimming competition, many previous studies have quantified the proportion of the start time in the overall swim times during competition to assess their relative contribution to overall race performance from various perspectives and given suggestions for training and technique optimization (Troup, 2005). For example, in 1988, Hay verified that the start represents 11% of the total time of a 50-meter in freestyle (front crawl) swimming race, and suggested an intensification in studies of the start techniques to reduce the time expense at this phase. Moreover, Maglischo and do Nascimento, as well as Cossor and Mason, all demonstrated that the start times ranged from 0.8% to 26.1% of the overall race time that depending on the race event and the proportion of start time increases when the distance of the swimming decreases (Maglischo and do Nascimento, 1999; Cossor and Mason, 2001).

In today’s swimming competitions, the FINA rules require that, in the competitions of front crawl, backstroke, and butterfly, the swimmer’s head should break the surface of the water before the 15-meter mark. Besides, there is a specific constraint to the breaststroke that the swimmer’s head must break the surface of the water before his or her hands turn inward at the widest part of the second stroke. According to previous studies, it is suggested that great velocity during the underwater phase of the start is critical to achieving high swim velocity (Seifert et al., 2007; Olstad et al., 2020; Zheng, 2021). Except for the backstroke race, swimmers will complete the process from land to underwater and then to the water surface in the start phase. Therefore, it has become a basic consensus in the research field of swimming start that a sub-division of the start phase is necessary. In 1981, Hay explored the body positions of swimmers during the race, suggesting that the swimmer had to be as horizontal as possible after start until the speed forward was less than the speed of the stroke (Hay, 1981). According to Hay’s suggestion, the swimmer’s performance at the start can be assessed by three partial times, which are the time between the start signal until the feet leaving the block (start phase), the time between the feet leaving the block until the first contact with the water (block phase), and the time from the first contact with the water until the swimmer commences stroking (glide phase).

The “glide phase” is also called the “underwater phase”. Most previous kinematical analyses of the start phases have focused on the reaction time on the starting blocks, the time of flight phase, and the time of entry phase, comparing the grab with the track start (Ruschel et al., 2007). Although some differences have been identified in start techniques, researchers emphasized that, in all cases, great consideration should be accorded to the underwater phase. For example, in 1985, Guimarães and Hay found that the most important phase in swimming is the underwater phase (Guimaraes and Hay, 1985), and Cossor and Mason found a negative correlation between the underwater velocity and the overall start time in the 100-m breaststroke (Mason and Cossor, 2000). In 2001, Sanders and Byatt-Smith established a mathematical model based on an exponential function that fit the velocity change during the glide phase for start technique optimization (Sanders and Byatt-Smith, 2001). In 2005, Troup found that 95% of the variance in start times can be explained by the glide phase (Troup, 2005). And in 2021, a study conducted by Sakai’s team clarified the magnitudes and functions of each joint torque acting on the extremities during the track start, claiming that the force of the hands was mainly influenced by extension torque at the shoulder joint, while the hip joint extension torque on the front side lower limb was mainly used for supporting the body weight until hands off (Sakai et al., 2021). Therefore, the analysis of kinematic characteristics in the underwater phase is significantly valuable for the start technique optimization of swimming competition.

Kinematics parameters related to the underwater phase such as coordinates of body paths, velocities, and accelerations, depend on swimmers and the application of different start techniques (Hermosilla et al., 2022). Although these kinematic parameters can be obtained in three-dimensional underwater images, and the body path can be calculated with the orientation of the swimmer’s mass center, it is still difficult to determine where is the similar or different phase between the two start processes and quantify how much each one differs from the other. In the field of biomechanics, motions of single cycles such as walking, running, and swimming have been standardized into 100% normalized time duration. However, there is no argument about the validity of this time normalization. The kinematic parameters in time series are generally similar in the time axis, but their corresponding relationship is unknown. For instance, when a swimmer strokes with two different techniques, although the two techniques are visually the same type of movement, the corresponding relationship of a parameter such as velocity and acceleration at the same time may not be the same because of the different motion patterns in the two kinds of stroke techniques.

The dynamic time warping algorithm (DTW) is a traditional method used in time series analysis for measuring similarity between two temporal sequences that vary in timing-dependent parameters (Olsen et al., 2018). Any data that can be turned into a linear sequence can be analyzed with DTW such as temporal sequences of video, audio, graphics data, and human motions. At present, DTW has been wildly applied in the field of speech recognition to determine different utterances, whose duration varies and depends on the speaker or their voice. Moreover, similarities in walking and running gaits have been detected by using DTW, even if one person was walking or running faster than the other, or if there were accelerations and decelerations during an observation (Müller, 2007; Rakthanmanon et al., 2013). Considering that the wave parameters of sound and the kinematic parameters of human motions both can be expressed as time series functions, DTW could be used to compare the kinematic characteristics of the underwater phase in different start techniques.

The objective of this study is to identify the difference between instantaneous velocity, average velocity, distance per kick, time per kick, instantaneous acceleration, and average acceleration in underwater phases of different start techniques by using a case of an Olympic champion in men’s 200 m individual medley competition. The novelty of this study is the application of pattern-matching analysis based on the DTW algorithm.

Methods

Study Design

An Olympic gold medal winner in men’s 200 m individual medley competition is invited to be the subject of this study (age: 27 years old, height: 191 cm, body weight: 85 kg). The swimmer was asked to perform grab start, which was also his usual way of starting, by using 2 different techniques. The one was entering the water with his hands separated and stretched forward naturally, while the other was entering the water with his hands together and intentionally kept them as extended forward as possible.

In order to eliminate the interaction effect between the two similar motions, the tests of the two starts with different techniques were scheduled in random order on two different days. Since the swimmer’s training plan consisted of several one-week cycles, the second test day was scheduled one week after the first test, which was the same training day of the same content. Thus, there was a 7-day wash-out between the two tests. The flow diagram of the study design is presented in Figure 1.

The flow diagram of the study
Figure 1 

The flow diagram of the study.

Outcome Measures

The Marker-Free Subaqueous Motion Recognition System (Version 1.0, Ningbo, China) developed by Ningbo University was used to capture the swimmer’s motion in the glide phase after the start.

According to the requirement of the system, 16 subaqueous cameras were positioned parallel to each other on both sides of the lane. There was a 2-meter distance between two cameras on each side to ensure the underwater phase of the swimmer from the start block to the 15-meter marker could be completely captured. At the same time, 16 super-aqueous cameras were positioned parallel to each other on both sides of the lane and right on top of each subaqueous camera. The function of these 16 super-aqueous cameras was to provide images above the surface of the water for the calibration and correction of the subaqueous 3D coordinates in the system and to capture the swimmer’s motion above the surface of the water in the start and block phase. Moreover, there were 3 super-aqueous cameras positioned behind the start block to capture the swimmer’s motion on the start block on the sagittal plane. All cameras captured the whole start of the swimmer synchronously and simultaneously with genlocked views and a 50 Hz frame rate. The schematic diagram of the testing site is presented in Figure 2, while the set positions of cameras are provided in Figure 3.

The schematic diagram of the testing site. (a) glide phase; (b) start phase
Figure 2 

The schematic diagram of the testing site. (a) glide phase; (b) start phase.

Set positions of cameras. (a) super-aqueous cameras; (b) subaqueous cameras
Figure 3 

Set positions of cameras. (a) super-aqueous cameras; (b) subaqueous cameras.

The kinematical analysis included 6 parameters, which were instantaneous velocity, average velocity, distance per kick, time per kick, instantaneous acceleration, and average acceleration in underwater phases. All units were presented according to the international standard units.

Statistical Analysis

The basic descriptive statistical analysis for all original data of the kinematic parameters was conducted to show their variation trends, while the linear and nonlinear regression curves between each kinematic parameter and time or position parameters were fitted according to their theoretical physical relationships.

Since the subaqueous pressure drag of a certain-shaped object had a linear relationship with the square of its velocity, whereas the velocity and acceleration were the first and second derivatives of the distance versus the time, a quartic polynomial function was used to fit the regression equation of distance versus the time in the underwater phase, a quadratic function was used to fit the regression equation of acceleration versus distance in the underwater phase, a power series function was used to fit the regression equation of velocity versus distance in the underwater phase, while segmental linear functions were used to fit the regression equations of the time or distance in each kick versus distance in the underwater phase.

The GraphPad software (Version Prism 9.1, Dotmatics Co., Ltd. San Diego, USA) was applied to conduct the statistical analysis and draw the relevant figures.

Dynamic Time Warping (DTW)

The Microsoft Visual Code Studio (Version 1.64.2, Microsoft, IL, USA) was used to conduct the DTW algorithm. The kinematic parameters collected on the first test day were set as the template characteristic sequences, while those collected on the second test day were set as the inquiry response sequences, while the Euclidean distance was applied to calculate the smallest cumulative distance of neighboring element under the boundary condition, the continuity condition, and the monotonicity condition.

Result

Kinematic Parameters

The kinematic parameters collected in the two test days was presented in Table 1. The regression curve between the distance and time was presented in Figure 4, while the regression curves between other kinematic parameters and distance were presented in Figure 5.

Table 1

The kinematic parameters in the tests.


START TECHNIQUE TIME (S) POSITION (M) VELOCITY (M/S) AVE.VELOCITY (M/S) ACCELERATION (M/S2) AVE.ACCELERATION (M/S2) DISTANCE/KICK (M) TIME/KICK (S)

Hand Free 0.00 0.00

1.86 6.62 2.30 2.62 1.24 1.41 0.68 0.26

2.38 7.84 1.51 2.35 –1.52 –0.52 1.22 0.52

2.90 8.93 1.35 2.10 –0.31 –0.48 1.09 0.52

3.48 10.03 1.07 1.90 –0.48 –0.34 1.10 0.58

4.08 11.11 1.07 1.80 0.00 –0.17 1.08 0.60

4.74 12.16 1.15 1.59 0.12 –0.32 1.05 0.66

5.94 14.26 1.19 1.75 0.03 0.13 2.10 1.20

Hands-together 0.00 0.00

1.78 6.63 2.47 3.09 1.39 1.74 0.99 0.32

2.30 7.91 1.71 2.60 –1.46 –0.94 1.28 0.52

2.82 9.02 1.38 2.13 –0.63 –0.90 1.11 0.52

3.36 10.09 1.22 1.98 –0.30 –0.28 1.07 0.54

3.92 11.13 1.25 1.86 0.05 –0.21 1.04 0.56

4.48 12.14 1.25 1.8 0.00 –0.11 1.01 0.56

5.12 13.22 1.13 1.69 –0.19 –0.17 1.08 0.64

6.32 15.35 1.26 1.78 0.11 0.08 2.13 1.20

The regression curve between the distance and time
Figure 4 

The regression curve between the distance and time.

The regression curves between kinematic parameters and distance. (a) instantaneous velocity; (b) average velocity; (c) instantaneous acceleration; (d) average acceleration; (e) distance of per kick; (f) time of per kick
Figure 5 

The regression curves between kinematic parameters and distance. (a) instantaneous velocity; (b) average velocity; (c) instantaneous acceleration; (d) average acceleration; (e) distance of per kick; (f) time of per kick.

According to the results, in the underwater phase, the regression curves, which were fitted based on each pair of kinematic parameters, were well consistent with the functional equations, which were previously assumed according to the physical principles between the kinematic parameters.

Pattern Matching Analysis

The results of pattern matching analysis based on DTW were provided in Figure 6. The numbers on the X and Y axis represented the lengths of the reshaped parameter arrays, the white line represented the shortest path, while the elements in each figure with different grayscales represent the accumulated cost (Euclidean distance) matrix of the two series. In these figures, the vertical segments in the path represented the different parts within the two series, indicating the difference between the two motions. While the diagonal segments in the path represented the similar parts within the two series, indicating the similar parts within the two motions.

The results of DTW pattern matching analysis of each kinematic parameter. (a) instantaneous velocity; (b) average velocity; (c) instantaneous acceleration; (d) average acceleration; (e) distance of per kick; (f) time of per kick
Figure 6 

The results of DTW pattern matching analysis of each kinematic parameter. (a) instantaneous velocity; (b) average velocity; (c) instantaneous acceleration; (d) average acceleration; (e) distance of per kick; (f) time of per kick.

Discussion

The objective of this study is to identify the difference of kinematic parameters in underwater phases of hands-free and together start techniques by using a case of an Olympic champion in men’s 200 m individual medley competition, through the DTW pattern-matching approach. There are two main findings of the study. On one hand, in the underwater phase in the start, the kinematic characteristics of the swimmer’s body in the underwater phase are following basic hydromechanical principles, which is that the subaqueous pressure drag of the swimmer has a linear relationship with the square of his or her velocity, as well as the Newtonian kinematic principles, which is that the velocity and acceleration are the first and second derivatives of the distance versus the time. On the other hand, the results of DTW pattern-matching analysis indicate that the swimmer will have a higher instantaneous and average velocity at the end of the start (the white line that represents the shortest path is above the diagonal line of the elements matrix) when the swimmer keeps hand together and arms straight forward in the subaqueous glide phase, this advantage comes from the dynamic benefit prepared during the whole glide phase. By applying this start technique, the swimmer can gain more propulsion at the end of the glide phase, during which the swimmer makes a full stroke and then lets his or her head break the surface of the water.

The first finding of this study once again confirms one of the consensuses of the academic community regarding the resistance mechanism in swimming, that is, the majority of the resistance loads on the swimmer’s body in the subaqueous glide phase is the pressure drag of the water, which is proportional to the square of the swimmers’ underwater velocity (Gardano and Dabnichki, 2006; Psycharakis, 2006; Schnitzler et al., 2006a; b).

During swimming, swimmers usually streamline their bodies as much as possible to reduce the cross-sectional area of their bodies relative to the water current for the reduction of pressure drag. However, since the human body is composed of many segments with different shapes, it is so difficult for a swimmer to maintain a perfectly streamlined posture all the time that the local pressure drags seem inevitable (Clarys-Robion, 1979). A study conducted by Clarys-Robin’s team identified that the pressure drag could be affected by the technique of swimming (Clarys-Robion, 1979), this finding was also confirmed by Huijing et al. in 1988 (Huijing et al., 1988). Moreover, a literature review by Zamparo et al. identified the difference in pressure drags of swimmers of various ages, gender, swimming strokes, and competition levels (Zamparo et al., 2020).

Moreover, since the pressure drag of the water, which is proportional to the square of the swimmers’ underwater velocity there will be an “optimal velocity” in the subaqueous glide phase after the start (Guimaraes and Hay, 1985; Sanders and Byatt-Smith, 2001; Arellano et al., 2002; Ruschel et al., 2007). In 2004, Schleihalf conducted a study in which the Measurement of Active Drag System (MAD-System) was used to measure the propulsive force of swimmers during swimming. This study found that the swimming velocity could not increase indefinitely with the increase in their stroke rates (Schleihauf, 2004). A study published in 2008 that was conducted by Sanders et al. confirmed this finding (Sanders et al., 2008).

Another important finding of the study is that the swimmer will have a higher instantaneous and average velocity at the end of the start when the swimmer keeps hands together and arms straight forward in the subaqueous glide phase and this advantage comes from the dynamic benefit prepared during the whole glide phase. And the swimmer can gain more propulsion at the end of the glide phase, during which the swimmer makes a full stroke and then lets his or her head break the surface of the water.

The mechanism might come from the optimal preload for the muscle activation at the end of the glide phase, in other words, it is probable that the motor control process of hands-together and arms straight forwards induces the isometric contraction of the swimmer’s upper body muscle groups, through which the swimmer can get a better timing of stroke in the final stage of the underwater glide phase, where the swimmer is asked to conduct a full stroke with arms according to the FINA rules (Matsuuchi et al., 2007; Mullen, 2018; Masud et al., 2022). This mechanism assumption comes from two perspectives.

On one hand, in 2003, Vilas-Boas et al. observed that any advantage in horizontal velocity established at the moment of entry would be quickly lost during entry and gliding. Therefore, the swimmer’s ability to reduce drag during the glide phase would have a significant impact on overall performance (Cossor and Mason, 2001; Vilas-Boas et al., 2003). Pereira et al. conducted a study in 2006 of biomechanical analysis on the glide phase after start by using subaqueous photography technology and finally found that the maximum depth in the glide phase had an impact on overall start performance, and the average velocity had the greatest impact on overall start time. Therefore, Pereira et al. recommended paying attention to starting technique optimization for the glide phase in swimming training (Pereira et al., 2006). Swimmers who are good at gliding underwater can glide further and gain an advantage before surfacing in swimming competitions. Considering that the subject of this study is an international elite swimmer and the overall distances of his starts almost reached the 15-meter marker, these results indicates that the application of different start techniques might have no effect on drag reduction for the world’s top swimmers.

On the other hand, the final stage of the start is a stroke. Previous studies have emphasized the importance of the timing for the first stroke in the final stage of the glide phase after the start and turn. Counsilman et al. in 1980 emphasized the importance of the timing for the first stroke after the start (Counsilman, 1980). Hay et al. in 1981 found that the best timing for the first stroke after start and turn was the moment in which the swimmer’s velocity had dropped to equal the velocity that his or her stroke could maintain (Hay, 1981). Later, a study conducted by Lyttle and Blanksby evaluated the difference in velocity before and after the first stroke in the final glide phase and found that when the swimmer started to stroke, the velocity decreased, and the magnitude of the decrease was negatively correlated with the swimmer’s overall start performance, which was represented by the time of the first 15 meters. Therefore, Lyttle and Blanksby recommended that swimmers should keep as much velocity as possible during the glide phase after the turn and start. At the same time, this study considered that the optimal stroke timing is individual, thus the angle and depth of entry should be individually assessed for different swimmers (Lyttle and Blanksby, 2000).

There are also some limitations of this study. For example, there might be a huge specialty in the Olympic champion case so the universality of the results is unclear. Moreover, the first stroke style of medley swimming competition is the butterfly. Therefore, the strategies of starts in other swimming strokes might be different from that used in this study. Last but not the least, in this study, there was no other start technique included in the comparison, the possibility of another start technique optimization protocol with more mechanical advantage is non-ignorable.

Conclusion

To sum up, in elite men’s 200 m individual medley competition, a start technique with hands-together and arms straight forwards could create mechanical and kinematical advantages for the swimmer, and the advantages might come from the optimal preload for the muscle activation at the end of glide phase.

Competing Interests

The author has no competing interests to declare.

References

  1. Arellano, R., Pardillo, S., & Gavilán, A. (2002). “Underwater undulatory swimming: Kinematic characteristics, vortex generation and application during the start, turn and swimming strokes”. Universidad de Extremadura, Caceras (pp. 29–41). 

  2. Clarys-Robion, J. P. (1979). “Human morphology and hydrodynamics.” In Biomechanics of aquatics. Haya-Cnopt, 39–71. 

  3. Cossor, J., & Mason, B. (2001). “Swim start performances at the Sydney 2000 Olympic Games”. In ISBS-Conference Proceedings Archive. 

  4. Counsilman, J. E. (1980). “Swimming III”. LWW). DOI: https://doi.org/10.1249/00005768-198024000-00014 

  5. Gardano, P., & Dabnichki, P. (2006). On hydrodynamics of drag and lift of the human arm. Journal of biomechanics, 39(15), 2767–2773. DOI: https://doi.org/10.1016/j.jbiomech.2005.10.005 

  6. Guimaraes, A. C. S., & Hay, J. G. (1985). A Mechanical Analysis of the Grab Starting Technique in Swimming. J International Journal of Sport Biomechanics, 1(1), 25–35. DOI: https://doi.org/10.1123/ijsb.1.1.25 

  7. Hay, J. G. (1981). Biomecânica das técnicas desportivas. Interamericana. 

  8. Hermosilla, F., Yustres, I., Psycharakis, S., Santos del Cerro, J., González-Mohíno, F., & González-Rave, J. M. (2022). Which variables may affect underwater glide performance after a swimming start? J European journal of sport science, 22(8), 1141–1148. DOI: https://doi.org/10.1080/17461391.2021.1944322 

  9. Huijing, P. A., Toussaint, H. M., Mackay, R., Vervoon, K., Clarys, J., & Hollander, A. P. (1988). Active drag related to body dimensions. Swimming science V, 18, 31–37. 

  10. Lyttle, A., & Blanksby, B. (2000). “A look at gliding and underwater kicking in the swim turn”. In ISBS-Conference Proceedings Archive. 

  11. Maglischo, E. W., & do Nascimento, F. (1999). Nadando ainda mais rápido Manole. J São Paulo. 

  12. Mason, B., & Cossor, J. (2000). “What can we learn from competition analysis at the 1999 Pan Pacific Swimming Championships?”. In ISBS-Conference Proceedings Archive. 

  13. Masud, M. H., La Mantia, M., & Dabnichki, P. (2022). Estimate of Strouhal and Reynolds numbers for swimming penguins. Journal of Avian Biology, 2022(2). DOI: https://doi.org/10.1111/jav.02886 

  14. Matsuuchi, K., Yamada, K., Nomura, T., Sakakibara, J., Shintani, H., & Miwa, T. (2007). Motion analysis of front crawl swimmer’s hand and visualization of flow fields using PIV. Nihon Kikai Gakkai Ronbunshu, B Hen/Transactions of the Japan Society of Mechanical Engineers, Part B, 73(10), 2027–2032. DOI: https://doi.org/10.1299/kikaib.73.2027 

  15. Mullen, G. J. (2018). Swimming science: Optimizing training and performance. University of Chicago Press. DOI: https://doi.org/10.7208/chicago/9780226287980.001.0001 

  16. Müller, M. (2007). “Dynamic Time Warping.” In M. Müller (Ed.), Information Retrieval for Music and Motion (pp. 69–84). Berlin, Heidelberg: Springer Berlin Heidelberg. DOI: https://doi.org/10.1007/978-3-540-74048-3_4 

  17. Navarro, F., & Arellano, R. (1985). Natación. Espanha: Comité Olímpico Español. Madrid/EI Corte Inglés. 

  18. Olsen, N. L., Markussen, B., & Raket, L. L. (2018). Simultaneous inference for misaligned multivariate functional data. Journal of the Royal Statistical Society, 67(5), 1147–1176. DOI: https://doi.org/10.1111/rssc.12276 

  19. Olstad, B. H., Wathne, H., & Gonjo, T. (2020). Key Factors Related to Short Course 100 m Breaststroke Performance. Int. J. Environ. Res. Public Health, 17(17), 6257. DOI: https://doi.org/10.3390/ijerph17176257 

  20. Pereira, S. (2001). Análise da Performance da Saída de Nadadores velocistas em Diferentes Alturas e Inclinações do Bloco de Partida. Dissertação (Mestrado em Ciências do Movimento Humano), Universidade do Estado de Santa Catarina. 

  21. Pereira, S., Araujo, L., Freitas, E., Gatti, R., Silveira, G., Roesler, H. J. B., et al. (2006). Biomechanical analysis of the turn in front crawl swimming. International Symposium on Biomechanics and Medicine in Swimming, 6(Suppl 2), 77–79. 

  22. Psycharakis, S. (2006). A 3D analysis of intra-cycle kinematics during 200 m freestyle swimming. Unpublished PhD, University of Edinburgh, Edinburgh. 

  23. Rakthanmanon, T., Campana, B., Mueen, A., Batista, G., Westover, B., Zhu, Q., et al. (2013). Addressing Big Data Time Series: Mining Trillions of Time Series Subsequences Under Dynamic Time Warping. ACM Trans Knowl Discov Data, 7(3). DOI: https://doi.org/10.1145/2500489 

  24. Ruschel, C., Araujo, L. G., Pereira, S. M., & Roesler, H. (2007). “Kinematical analysis of the swimming start: block, flight and underwater phases”. In ISBS-Conference Proceedings Archive. 

  25. Sakai, S., Koike, S., Takeda, T., Sengoku, Y., Homma, M., & Takagi, H. (2021). Kinetics of four limb joints during kick-start motion in competitive swimming. Sports Biomech (pp. 1–19). DOI: https://doi.org/10.1080/14763141.2021.1963465 

  26. Sanders, R., & Byatt-Smith, J. (2001). “Improving feedback on swimming turns and starts exponentially”. In ISBS-Conference Proceedings Archive. 

  27. Sanders, R., Psycharakis, S., Naemi, R., & McCabe, C. (2008). “Advancing swimming science with CARE”. In 5th International Scientific Conference on Kinesiology (pp. 81–89). University of Zagreb. 

  28. Schleihauf, R. E. (2004). Biomechanics of human movement. Bloomington, Ind: Authorhouse. 

  29. Schnitzler, C., Ernwein, V., Seifert, L., & Chollet, D. (2006a). Intracyclic velocity signal as a tool to evaluate propulsive phase duration. Biomechanics and Medicine in Swimming X, 88–90. 

  30. Schnitzler, C., Ernwein, V., Seifert, L., & Chollet, D. (2006b). The stability of IDC during maximal and submaximal swim trials questioned. Biomechanics and Medicine in Swimming X, 6(Suppl. 2), 255–257. 

  31. Seifert, L., Vantorre, J., & Chollet, D. (2007). Biomechanical analysis of the breaststroke start. Int J Sports Med, 28(11), 970–976. DOI: https://doi.org/10.1055/s-2007-965005 

  32. Troup, J. (2005). “A device for quantitative measurement of starting time in swimming.” In R. Arellano, F. Moreno, M. Martínez & A. Qña (Eds.), Biomechanics and Medicine in Swimming VII (pp. 219–224) Routledge. 

  33. Vilas-Boas, J., Cruz, J., Sousa, F., Conceicao, F., Fernandes, R., & Carvalho, J. (2003). “Biomechanical analysis of ventral swimming starts: comparison of the grab start with two track-start techniques”. In IXth World Symposium on Biomechanics and Medicine in Swimming. Saint Etienne: University of Saint Etienne (pp. 249–253). 

  34. Zamparo, P., Cortesi, M., & Gatta, G. (2020). The energy cost of swimming and its determinants. European journal of applied physiology, 120(1), 41–66. DOI: https://doi.org/10.1007/s00421-019-04270-y 

  35. Zheng, K. (2021). Research on the Physiological Monitoring and Evaluation of Pre-Competition Altitude Training for Zhejiang Elite Swimmers. Physical Activity and Health, 5(1), 64–70. DOI: https://doi.org/10.5334/paah.91 

comments powered by Disqus