Visualisation of running form changes measured by wearable sensors for conditioning management, an application of the Functional Data Analysis

Visualisation of running form changes measured by wearable sensors for conditioning management, an application of the Functional Data Analysis | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Visualisation of running form changes measured by wearable sensors for conditioning management, an application of the Functional Data Analysis Hirofumi Doi, Hidetoshi Matsui, Daisuke Nishioka, Yuri Ito, Ryuichi Saura This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-3850139/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Running is a widely-accepted activity among the general public, with runners aspiring to achieve optimal performance. However, established methods for the regular monitoring of running forms is lacking. To address this gap, we explore a versatile visualization method utilizing the widely-adopted Inertial Measurement Unit sensor. The running forms of 17-year-old male high school students were monitored during long-distance running training. Acceleration and angular velocity data were collected from a sensor attached to the lumbar region; data from the left foot contact to the next left foot contact were defined as the running cycle. Fatigue during running was assessed using the Borg Scale. The distribution of principal component scores obtained from functional principal component analysis of the running form data corresponded to changes in fatigue from one measurement session to another. However, no consistent trends or changes were observed across subjects. The running forms of participants who were measured twice exhibited a close distribution and similarity, yet unique features were also observed during each measurement. The findings suggest that changes and characteristics of runners' running forms can be readily visualized using a generic approach based on commonly-used sensors and functional data analysis. Health sciences/Health care/Disease prevention Health sciences/Health care/Patient education Biological sciences/Computational biology and bioinformatics/Statistical methods Biological sciences/Zoology/Biomechanics Biological sciences/Systems biology/Signal processing Biological sciences/Systems biology/Time series Health sciences/Biomarkers Health sciences/Health care Health sciences/Signs and symptoms Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction Running is considered as a popular activity. Runners set diverse goals, including improving personal best times, enhancing running performance through distance adjustments, refining diet and health indicators, engaging with fellow enthusiasts, and alleviating stress [ 1 , 2 ]. Aspiring to achieve a slightly better state with each run, runners constantly face the challenge of confirming whether their running performance has surpassed their previous record. Moreover, it is well-established that a majority of runners experience Running-Related Injuries (RRIs) [ 3 – 7 ]. Extensive research has been conducted on factors contributing to RRIs, encompassing variables such as time and distance of exposure, exercise load and environment, running skills, fatigue, and history of trauma [ 8 – 19 ]. In popular marathons, participants are subject to significant exposure in terms of both distance and time, leading to biomechanical changes [ 20 – 23 ]. Maas et al. highlighted that novice runners undergo substantial alterations in their running strategies driven by fatigue during long-distance running [ 24 ]. Therefore, can runners effectively assess whether biomechanical changes during races or training align with their expectations? Numerous biomechanical analyses have been conducted in running using measurement equipment in tightly-controlled environments [ 25 ]. Interpreting the analysis results, including representative values and waveforms of elements such as force and angles, requires specialized knowledge, rendering it impractical to acquire and apply such information regularly. Conversely, the information available to runners in their daily routines consists of simple metrics (such as heart rate and cadence) provided by running watches and smartphone applications, observational data from coaches, and confirmation of the running form from videos. Readily-accessible information concerning changes in the running biomechanics of runners and coaches is limited. As a potential example, standard models or running forms of elite athletes could potentially serve as evaluation criteria. However, a universally-shared running form that is a common goal applicable to all individuals and situations is highly improbable [ 26 , 27 ]. This can be attributed to the variation in runners' skeletal structure, even when factors such as race, age, and sex are consistent alongside differing background factors, including past training, experience, upbringing, medical history, footwear, and ground conditions. Furthermore, the lack of established means for acquiring and evaluating running form data over an extended period in real-world running environments makes it challenging for runners to comprehend trends in their own running form. Our objective is to propose a versatile method for daily visualization and categorization of the running form. As one device, Inertial Measurement Unit (IMU) are attracting attention. IMU sensors integrated into running watches and smartphones have found applications in activity measurement, offering biomechanical indicators through running apps and supporting research activities, rendering them suitable for daily visualization [ 28 – 31 ]. Functional Data Analysis (FDA) is one of the most advantageous methods of analysis when considering the maximum use of data collected by IMUs over time. FDA involves the transformation of time-series observations for individuals measured at discrete time points into functions (values), and employs statistical methods to analyze a set of functional data [ 32 ]. Time-series data capturing continuous running motions are recorded as discrete time point data, necessitating the capture of changes occurring over time as a coherent set of information encompassing the entire running cycle. FDA has the ability to handle datasets with different sampling rates and variations in the number of time points in a measured operating cycle. This capability is particularly advantageous when dealing with multiple time-series data comprising different observation time points and number of observation time points. We investigated the application of the widely-used Inertial Measurement Unit (IMU) sensors and Functional Data Analysis (FDA) to develop visualization methods accessible to the general running community. In this study, we explored the possibility of visualizing changes in fatigue and running form induced by long-distance pace training using FDA. Methods In the track-and-field team of the high school, we observed the running form of athletes participating in long-distance pace training, which involved running distances of 15 km or more at a consistent speed. The specifics of the training, including distance and target pace, were determined through discussions between the coaches and athletes. The observations were conducted during four practice sessions in March 2021. Regarding data related to the running form, we acquired phase and actual measurements of acceleration and angular velocity from IMU sensors affixed near the human center of mass in the lumbar region. We securely attached IMU sensors (VICON BLUETRIDENT) just above the participants' L5/S1 (5th Lumbar Sacral Joint) using a waist belt. Acceleration and angular velocity were measured at a frequency of 1125 Hz, whereas geomagnetic data were captured at 100 Hz. Fatigue levels were assessed using the Borg-RPE (15-point Borg rating of perceived exertion scale), and data were obtained at 1-km intervals [ 33 – 35 ]. The presence of fatigue was defined as Borg scores of 12 or below, indicating "no fatigue," and Borg scores of 13 or above, indicating "fatigue." Informed participants received training in advance on how to measure fatigue using the Borg RPE scale. When there was a change in participants' fatigue levels, they signed an observer, and the recall method was employed after the practice session for confirmation. Data collection included measurements of height, weight, circumference (thigh and calf), lower limb length, and joint range of motion (hip, knee, and ankle), as well as subjective condition assessments before and after practice on measurement day. These included long-lasting fatigue, discomfort, and pain after practice. The researchers evaluated the differences before and after these aspects. The subjective conditions before practice were collected after finalizing the details of the training session on that day. However, participants who experienced severe fatigue from the previous day's practice or experienced pain chose not to participate in observations on the current day. Participants The study included 17-year-old male high school students specializing in long-distance track-and-field. Information sessions were conducted with prospective participants and their guardians, where the purpose of the study was explained, and consent for participation was obtained. Participants who altered or discontinued the training session for reasons such as pain were excluded from the observation on that day. Statistical Method Statistical methods were based on the future development of applications on devices such as smartphones. Python 3.9 and R 4.1.0 were employed for statistical processing. The data obtained from the IMU sensor underwent preprocessing, including noise reduction and transformation from sensor coordinates to spatial coordinates. One complete running cycle was defined as right foot contact–jump–left foot contact–jump, and acceleration data corresponding to one running cycle was treated as an individual observation. We treat the time-series data extracted from a specific 1 km segment as acceleration data representing the running form of that segment. The number of cycles of time-series data extracted from the 1 km segment was set to 30. Specifically, linear segments were identified from the geomagnetic data, and for each 1 km interval, three linear running segments (approximately 70 seconds after the start of the 1 km interval, approximately 100 seconds after, and approximately 130 seconds after) were chosen. Time-series data for 10 running cycles were obtained for each linear segment, resulting in 30 time-series datasets per 1 km interval. The extracted dataset (distance for each individual \(\times\) 3 extraction points \(\times\) 10 cycles) were transformed into functions for each cycle and standardization of the time points using functional data analysis. One way of transforming the observed data into functions is to use basis function expansion, i.e., the function \(x\left(t\right)\) is expressed as the sum of \(m\) basis functions with coefficients \({w}_{1},\dots ,{w}_{m}\) and basis functions \({\phi }_{1}\left(t\right),\dots ,{\phi }_{m}\left(t\right)\) as follows; $$x\left(t\right)={w}_{1}{\phi }_{1}\left(t\right)+\dots +{w}_{m}{\phi }_{m}\left(t\right)+\epsilon .$$ It then allows the change in acceleration over a running cycle to be expressed as an acceleration function \(x\left(t\right)\) that varies over a standardized time \(t\) . Accordingly, the set of acceleration data for a running cycle can be treated as an acceleration function \(x\left(t\right)\) generated for each running cycle. B-splines were used as basis functions. Functional Principal Component Analysis (FPCA) was used to evaluate participants' running forms [ 36 ]. Cluster analysis was performed using k-means and Ward's methods on the principal component scores obtained. As the goal was to partition the data into individual measurement sessions, the number of clusters to which measurement sessions belong was determined based on the condition that at least two-thirds of the data from one measurement session should be included in a single cluster. Additionally, the distribution and trends of the scores of the first and second principal components were assessed from the results of FPCA. As fatigue is expected to cause gradual changes in running form near the limit, we categorized fatigue levels into “no-fatigue" (Borg 11 or below), "Borderline no-fatigue" (Borg 12), and “fatigue" (Borg 13 and above). We then examined the relationship between fatigue levels and principal component scores. To evaluate the relationship between fatigue and running form, we established two baseline models to determine the presence of fatigue: a group model (consisting of no-fatigue data from all participants) and an individual model (no-fatigue data from an individual participant). The relationship between running form and fatigue was modeled by Functional Logistic Regression Analysis, where univariate (in each direction) and multivariate analyses (across triaxial) are conducted [ 37 ]. In these analyses, we compared either the group or individual baseline (no-fatigue data) with the individual fatigue data obtained in a single measurement. Ethical Considerations The study was designed in compliance with the Ethical Guidelines for Medical Research Involving Human Subjects and the Personal Information Protection Law, and conducted with approval from the Ethics Committee of Osaka University of Medical and Pharmaceutical Sciences (2020 − 172). The study objective was explained to prospective participants and their guardians at a briefing session, and informed consent was obtained from both prospective participants and their guardians who agreed to participate voluntarily. Results Measurements were conducted during four practice sessions in March 2021. Consent for participation in the study was obtained from all 13 male high school students, aged 17, specialized in long-distance track-and-field. Twelve students (n = 12) participated in the measurements after excluding one who met the exclusion criteria owing to practice restrictions associated with RRIs, resulting in 24 datasets (d = 24). Data from two cases where fatigue did not reach Borg 13 (d = 2), three cases of IMU sensor malfunction (d = 3), two cases with incomplete interval time or fatigue data acquisition (d = 2), and two cases of early fatigue appearing immediately after practice initiation (d = 2) were excluded from the analysis (Fig.1). Essentially, 15 datasets (d = 15) from 10 participants (n = 10) were considered. Body measurements, including thigh and calf circumferences, lower limb length, and joint range of motion (hip, knee, ankle), could not be obtained due to COVID-19 prevention guidelines at the high school where the participants were enrolled, as they required direct physical contact. Therefore, height and weight were collected through self-reporting. Consequently, data on circumference (thigh and calf), lower limb length, and joint range of motion (hip, knee, and ankle) were not available. The long-distance pace training sessions were tailored to the participants' training goals and abilities, resulting in distances ranging from 15.0 to 21.6 km and pace timing ranging from 230 to 255 seconds per km (refer to Table1). The change in fatigue due to long-distance pace training began from Borg 9–11 and reached Borg 13–15. The extracted data included no fatigue data (270–510) and fatigue data (30–360) for each measurement. Among the five participants who were measured twice, four had different subjective conditions before and after the long-distance pace training sessions. Classification of running forms The waveform representing running form was examined based on the first and second principal components obtained from multivariate FPCA for the triaxial directions (refer to Fig.2 and Supplementary Equation S1). The cumulative contribution of the first and second principal components in the multivariate analysis was 79.24%. The first principal component contained cyclic components of the left stance phase in the vertical and sagittal directions, whereas the second principal component contained components of the right stance phase in the vertical and sagittal directions as well as the left–right direction. Clustering was performed on the multivariate data, including components in each direction (vertical, sagittal, left-right). Through exploration using K-means and Ward methods, the number of clusters between 3 and 5 was likely, and clusters containing data from only one measurement session appeared for clusters above 6 (refer to Supplementary Fig.S2). With 4 or 5 clusters, some measurement data are dispersed among multiple clusters, making it difficult to determine the appropriate cluster assignment. With 3 clusters, 14 out of 15 measurements were clearly classifiable, and even for the remaining one measurement, 69.61% of the data were attributed to one cluster, leading us to choose 3 clusters (refer to Supplementary Fig.S2). When considering cluster Cls.1 with the highest number of participants as a reference, distinct characteristics were observed: cluster Cls.2 had lower scores in the first principal component, whereas cluster Cls.3 exhibited higher scores in the second principal component. The waveforms of participants who were measured twice were closely distributed and classified into the same cluster (refer to Supplementary Fig.S3). The relationship between fatigue levels and median principal component scores varied among the participants for each measurement (see Fig.3 and Supplementary Fig.S4). For instance, in the case of Participant B, the primary change in waveform, particularly in the horizontal direction, corresponded to the fatigue level in B1, whereas in B2, the change was observed in the second principal component (vertical direction). Furthermore, some participants such as D1 and D2 exhibited differences in the magnitude of displacement, whereas others such as Participant G showed minimal changes in displacement. Moreover, the changes in principal components over time varied among participants, and participants who underwent the two measurements twice did not necessarily exhibit the same trends. Assessment of fatigue and running form Fatigue and running were assessed by evaluating the association between fatigue induced by long-distance running and acceleration waveforms representing the running form. This was performed using both univariate and multivariate functional logistic regression analyses (refer to Fig.4, Supplementary Equation S1 and Supplementary Fig.S5). The results of the univariate functional logistic regression analyses conducted for each direction, along with the multivariate functional logistic regression analysis considering all three directions, revealed associations between waveforms and fatigue in at least one of the univariate or multivariate analyses. However, even if the strength of the relationship was confirmed in the univariate analysis, it did not necessarily manifest in the multivariate analysis. Discussion In this study, we attempted to capture the variations and characteristics of running form using IMU sensors attached to the waist. We utilized both the magnitude and phase of acceleration and angular velocity near the center of mass, and applied FDA to visualize these aspects of the running form. Although previous studies have been traditionally relying on specialized motion analysis equipment and expertise for interpretation of results, attempts have been made in recent years to capture motion using IMU sensors. Studies have attempted to validate the assessment of whole-body biomechanics using partial acceleration data from various body parts, including the hips, back, wrists, and feet [ 38 , 39 ]. Human movement arises from temporal changes in posture, involving both limbs and the trunk, and manifests as motion within the spatial dimensions of the center of mass. Typically, the center of mass in an upright posture is believed to be located in the lower part of the torso. However, it is important to note that the center of mass is not fixed and can shift with changes in posture. The clustering of principal component scores derived from multivariate FPCA resulted in the classification of acceleration waveform patterns representing the running form into three distinct clusters. Although principal component scores were plotted in close proximity for each participant, their positions varied. Among the participants who underwent measurements on two occasions, the second set of principal component scores was also classified into the same cluster as the first set and their positions were also close. This suggests that within individual participants, the principal component scores may exhibit relatively similar distributions (see Supplementary Fig.S4). However, the relationship between acceleration waveform patterns representing running form and fatigue did not demonstrate consistent trends or convergence. Subsequently, we computed the average acceleration waveform representing the running form for both the entire participant group and each cluster. The results indicated a similarity in waveform patterns between the entire participant group and Cluster Cls.1, while Clusters Cls.2 and Cls.3 displayed distinct waveform shapes, each exhibiting unique characteristics. It is worth noting that these three clusters were derived from a relatively small sample size, suggesting that a larger-scale investigation in the future may unveil a greater number of distinct running form clusters. Previous studies have proposed various running-form categories, including grounding patterns, running ability, and sex, with a wide range of possible combinations [ 40 – 43 ]. This implies that variations in the running form among runners should be considered. It may be inappropriate to discuss current performance or future directions simply by comparing them to benchmarks, such as elite runners or group averages. Moreover, it is crucial to note that this study specifically focused on a group of 17-year-old male athletes engaged in the same track-and-field discipline and training environment. Consequently, subtle differences between the participants are visualized. Similar to the characteristics of conventional principal component analysis, the features of the evaluated dataset can influence the principal component functions obtained from functional principal component analysis. For example, if participant data were assessed alongside elite athletes or recreational runners, the differences in acceleration waveforms representing the running forms of the participants in this study may be perceived as relatively limited. This underscores the importance of understanding the features of the dataset being evaluated according to the intended purpose and interpreting the results appropriately. Furthermore, to visualize individual performance and condition management, it is advisable to consider readily-available running-related motion indicators, such as cadence and walking speed. It would be feasible to conduct analyses that incorporate an individual’s past and present data, compare groups and individuals, and target specific individuals. Detecting fatigue-related changes in running form is critical. Running coaches routinely assess athletes' condition and offer guidance based on their current running form. Simultaneously, many runners seek to improve their performance by referring to their own or others' running forms. Essentially, running form plays a pivotal role in comprehending conditions and enhancing performance. However, Robbie et al. asserted that even with the trained eye of a running coach, visually classifying running economy by long-distance running coaches remains a challenging task, and assessment of running form is not straightforward [ 44 – 46 ]. The findings confirm that the trend of change in principal component scores of FCPA with changes in fatigue is not constant, confirming that it is difficult to fit the trend of change into a standardized pattern (Fig. 3 ). In this study, two models were constructed: a group model based on fatigue-free data from all participants and an individual model based on fatigue-free data from each participant. The results of functional logistic regression analysis revealed a relationship between the presence or absence of fatigue and running form in various directions, either singularly or in multiples. The difference between the group and individual model baselines did not consistently yield conclusive results, emphasizing the need for careful interpretation (Fig. 4 ). The variations in running form based on the presence or absence of fatigue, as observed through comparison with the group model, may have been influenced by individual variations in running form among the participants (see Supplementary Fig.S6). Both univariate and multivariate results frequently demonstrated that the group model outperformed the individual model in terms of deviance-explained% and ROC values. This suggests that, when estimating the presence of fatigue from the running form, it is crucial to recognize the limitations of comparing oneself with others or extrapolating the results to others. The benefits and challenges of visualizing the running form are explored with the aspiration that all runners, regardless of whether they are elite athletes, students, or recreational runners, can monitor their daily condition and respond appropriately. Recognizing the signs of changes in health conditions or indicators, such as RRIs, and making decisions to adjust training or rest is a complex task [ 47 , 48 ]. With the methodology used in this study, obtaining results commonly used in motion analysis, such as joint angles or force intensity, is not feasible. However, by utilizing acceleration data from IMU sensors and applying functional data analysis, it is possible to visualize where changes in the running form occur within the running cycle and how they compare with past data. Although this visualized information may seem insignificant, it can serve as a guideline for managing running forms and cultivating interest in one's own condition and health literacy. Furthermore, if visualization using real-world data generated by runners worldwide during their daily runs becomes possible, it could serve as a benchmark for running form management. Consequently, if there are concerns about daily changes in physical condition, taking proactive steps to seek advice from running coaches or healthcare professionals is recommended. In the future, it would be desirable to establish an environment in which all runners can recognize changes in their condition and respond accordingly. To elucidate the relationship between various factors such as conditions, performance, individual factors of the subjects, footwear, surface, weather, psychological aspects, and acceleration waveform, it is imperative to conduct repeated investigations in actual running environments. This study presents a preliminary exploration of visualization techniques using FDA. Further investigations are warranted to explore the application of this methodology to data from wrist-worn running watches or smartphones as well as devices attached to different body parts. In addition, ensuring the mutual compatibility of data from IMU devices attached to different body segments remains a crucial challenge for future research. The FPCA conducted in this study creates and classifies the principal component functions based on all available information, providing a relative visualization of existing data. Essentially, it is not an absolute evaluation metric, necessitating the collection of information from all subjects for comparison, including data from athletes or Olympians, which serve as ideal benchmarks. New methodologies that provide absolute evaluation metrics for all subjects are desired to enable a straightforward visualization of runners' conditions and performance. Limitations This study has several limitations. The relationship between fatigue induced by long-distance running and the running forms reported in this study may potentially include unmeasured confounders and errors. This study had limited participants, and factors such as running skill, sex, race, and skeletal structure were not considered. Moreover, the limited number of measurements within the participants prevented a thorough examination of the effects of the conditions, performance, and environment, as well as reproducibility. Therefore, extensive validation is necessary to generalize the results to other participants. Although geomagnetic corrections were implemented in this study, GPS corrections were not applied, which led to an accumulation of errors specific to the IMU-based methodology. This could potentially be addressed through error corrections using technologies such as GPS and higher-performance IMUs, which may yield more accurate results. Conclusions The results of this study demonstrate that utilizing general-purpose sensors and FDA techniques, rather than specialized motion analysis devices, allows the visualization of changes and characteristics in a runner's running form. The characteristics of running form and changes in running form associated with the onset of fatigue are not uniform, highlighting the importance of understanding the changes and features of running form within individuals, rather than solely comparing them with standard models or group averages, in the evaluation of general running form. The visualization of changes and characteristics in running form holds significant potential for applications such as adjusting training regimens and volumes, managing performance (physical capabilities), monitoring overall physical and mental conditions, and identifying signs of RRIs. Declarations Data Availability Statement Datasets generated and/or analyzed during this study are not publicly available due to consent conditions with the study subjects, but are available upon reasonable request from the corresponding author H.D Acknowledgements We very grateful to the various members of the Department of Rehabilitation Medicine and the Department of Medical Statistics at Osaka Medical and Pharmaceutical University for their scientific advice and insightful discussions during the conduct of this study. We would like to thank Editage (www.editage.jp) for English language editing. Author contributions H.D. conceived this study under the guidance of R.S., H.M., Y.I., N.D.. Y.I., assisted H.D. in developing the statistical analysis plan. H.D performed the statistical analysis under H.M.'s guidance. H.M. contributed to interpretation of the results. H.D. prepared the manuscript, figures and tables. R.S. supervised the conduct of the study. All authors reviewed and critically revised the manuscript for intellectual content. All the authors approved the final version of the manuscript. Competing interests The author(s) declare no competing interests. References Shipway, R. & Holloway, I. Running free: Embracing a healthy lifestyle through distance running. Perspect. Public Health. 130, 270–276 (2010). Janssen, M. et al . Understanding different types of recreational runners and how they use running-related technology. Int. J. Environ. Res. Public Health. 17, 2276; 10.3390/ijerph17072276 (2020). Kakouris, N., Yener, N. & Fong, D. T. A systematic review of running-related musculoskeletal injuries in runners. J. Sport Health Sci. 10, 513–522 (2021). Desai, P. I. A., Jungmalm, J., Börjesson, M., Karlsson, J., & Grau, S. Recreational runners with a history of injury are twice as likely to sustain a running-related injury as runners with no history of injury: a 1-year prospective cohort study. J. Orthop. Sports Phys. Ther. 51, 144–150 (2021). Toresdahl, B. et al. Factors associated with injuries in first-time marathon runners from the New York City marathon. Phys. Sportsmed. 50, 227–232 (2022). Mayne, R. S., Bleakley, C. M. & Matthews, M. Use of monitoring technology and injury incidence among recreational runners: a cross-sectional study. BMC Sports Sci. Med. Rehabil. 13, 1–7 (2021). Dempster, J., Dutheil, F. & Ugbolue, U. C. The Prevalence of Lower Extremity Injuries in Running and Associated Risk Factors: A Systematic Review. J. Phys. Act. Health Title. 5, 133–145 (2021). Hulme, A., Nielsen, R. O. Timpka, T. Verhagen, E. & Finch, C. Risk and protective factors for middle-and long-distance running-related injury. Sports Med. 47, 869–886 (2017). Saragiotto, B. T. et al . What are the main risk factors for running-related injuries? Sports Med. 44, 1153–1163 (2014). Videbæk, S., Bueno, A. M. Nielsen, R. O. & Rasmussen, S. Incidence of running-related injuries per 1000 h of running in different types of runners: a systematic review and meta-analysis. Sports Med. 45, 1017–1026 (2015). Van Poppel, D. et al. Risk factors for overuse injuries in short-and long-distance running: A systematic review. J. Sport Health Sci. 10, 14–28 (2021). Kluitenberg, B., van Middelkoop, M. Diercks, R. & van der Worp. H. What are the differences in injury proportions between different populations of runners? A systematic review and meta-analysis. Sports Med. 45, 1143–1161 (2015). Nielsen, R. O., Buist, I., Sørensen, H., Lind, M. & Rasmussen, S. Training errors and running related injuries: a systematic review. Int. J. Sports Phys. Ther. 7, 58–75 (2012). Daoud, A. I. et al . Foot strike and injury rates in endurance runners: a retrospective study. Med. Sci. Sports Exerc. 44, 1325–1334 (2012). Tenforde, A. S. et al . Overuse injuries in high school runners: lifetime prevalence and prevention strategies. PM&R . 3, 125–131 (2011). Goss, D. L. & Gross, M. T. A review of mechanics and injury trends among various running styles. US Army Med. Dep. J. 3, 62–71 (2012). Kemler, E. & Huisstede, B. Performance goals of runners are associated with the occurrence of running-related injuries. Phys. Ther. Sport. 50, 153–158 (2021). Willwacher, S. et al. Running-related biomechanical risk factors for overuse injuries in distance runners: a systematic review considering injury specificity and the potentials for future research. Sports Med . 52, 1863–1877 (2022). Desai, P. I. A., Jungmalm, J., Börjesson, M., Karlsson, J. & Grau, S. Recreational runners with a history of injury are twice as likely to sustain a running-related injury as runners with no history of injury: a 1-year prospective cohort study. J. Orthop. Sports Phys. Ther. 51, 144–150 (2021). Reenalda, J., Maartens, E. Homan, L. & Buurke, J. J. Continuous three dimensional analysis of running mechanics during a marathon by means of inertial magnetic measurement units to objectify changes in running mechanics. J. Biomech. 49, 3362–3367 (2016). Apte, S. et al . Biomechanical response of the lower extremity to running-induced acute fatigue: a systematic review. Front. Physiol. 12, 646042; 10.3389/fphys.2021.646042 (2021). Zinner, C. & Sperlich, B. Marathon Running: Physiology, Psychology, Nutrition and Training Aspects (ed. Sperlich, B.) (Springer 2016). Clermont, C. A., Benson, L. C., Edwards, W. B., Hettinga, B. A. & Ferber, R. New considerations for wearable technology data: changes in running biomechanics during a marathon. J. Appl. Biomech. 35, 401–409 (2019). Maas, E., De Bie, J., Vanfleteren, R., Hoogkamer, W. & Vanwanseele, B. Novice runners show greater changes in kinematics with fatigue compared with competitive runners. Sports Biomech. 17, 350–360 (2018). Robertson, G. E., Caldwell, G. E., Hamill, J., Kamen, G. & Whittlesey, S. N. Research methods in Biomechanics. Human Kinetics (website) (2013). Patoz, A., Lussiana, T., Breine, B., Gindre, C. & Hébert-Losier, K. There is no global running pattern more economic than another at endurance running speeds. Int. J. Sports Physiol. Perform. 17, 659–662 (2022). Hamill, J. & Gruber, A. H. Is changing foot strike pattern beneficial to runners? J. Sport Health Sci. 6, 146–153 (2017). Migueles, J. H. et al . GRANADA consensus on analytical approaches to assess associations with accelerometer-determined physical behaviors (physical activity, sedentary behavior and sleep) in epidemiological studies. Br. J. Sports Med. 56, 376–384 (2022). Benson, L. C., Clermont, C. A., Bošnjak, E. & Ferber, R. The use of wearable devices for walking and running gait analysis outside of the lab: A systematic review. Gait Posture 63, 124–138 (2018). Ahamed, N. U., Kobsar, D., Benson, L. C., Clermont, C. A., Osis, S. T. & Ferber, R. Subject-specific and group-based running pattern classification using a single wearable sensor. J. Biomech. 84, 227–233 (2019). Faber, G. S., Chang, C. C., Kingma, I. D. S. A. R. T., Dennerlein, J. T. & Van Dieën, J. H. Estimating 3D L5/S1 moments and ground reaction forces during trunk bending using a full-body ambulatory inertial motion capture system. J. Biomech. 49, 904–912 (2016). Ramsay, J. O. & Silverman, B. W. Functional Data Analysis ( 2nd ed.) (Springer-Verlag, 2005). Borg, G. Borg's Perceived Exertion and Pain Scales (Human Kinetics, 1998) Williams, N. The Borg rating of perceived exertion (RPE) scale. Occup. Med. 67, 404–405 (2017). Borg, G. A. Psychophysical bases of perceived exertion. Med. Sci. Sports Exerc. 14, 377–381 (1982). Happ, C. & Greven, S. Multivariate functional principal component analysis for data observed on different (dimensional) domains. J. Am. Stat. Assoc. 113, 649–659 (2018). Goldsmith, J., Bobb, J., Crainiceanu, C. M., Caffo, B. & Reich, D. Penalized functional regression. J. Comput. Graph. 20, 830–851 (2011). Wixted, A. J., Billing, D. C. & James, D. A. Validation of trunk mounted inertial sensors for analysing running biomechanics under field conditions, using synchronously collected foot contact data. Sports Eng . 12, 207–212 (2010). Lee, M. & Park, S. Estimation of three-dimensional lower limb kinetics data during walking using machine learning from a single IMU attached to the sacrum. Sensors 20, 6277; 10.3390/s20216277 (2020). Almeida, M. O., Davis, I. S. & Lopes, A. D. Biomechanical differences of foot-strike patterns during running: a systematic review with meta-analysis. J. Orthop. Sports Phys. Ther. 45, 738–755 (2015). Clermont, C. A., Osis, S. T., Phinyomark, A. & Ferber, R. Kinematic gait patterns in competitive and recreational runners. J. Appl. Biomech. 33, 268–276 (2017). Clermont, C. A., Phinyomark, A., Osis, S. T. & Ferber, R. Classification of higher-and lower-mileage runners based on running kinematics. J. Sport Health Sci. 8, 249–257 (2019). Eskofier, B. M., Kraus, M., Worobets, J. T., Stefanyshyn, D. J. & Nigg, B. M. Pattern classification of kinematic and kinetic running data to distinguish gender, shod/barefoot and injury groups with feature ranking. Comput. Methods Biomech. Biomed. Eng. 15, 467–474 (2012). Cochrum, R. G. et al . Visual classification of running economy by distance running coaches. Eur. J. Sport Sci. 21, 1111–1118 (2021). Benson, L. C., Ahamed, N. U., Kobsar, D. & Ferber, R. New considerations for collecting biomechanical data using wearable sensors: Number of level runs to define a stable running pattern with a single IMU. J. Biomech. 85, 187–192 (2019). Ahamed, N. U. et al. Subject-specific and group-based running pattern classification using a single wearable sensor. J. Biomech. 84, 227–233 (2019). Timpka, T. et al . Injury acknowledgement by the reduction of sports load in world-leading athletes (track and field) varies with their musculoskeletal health literacy and socioeconomic environment. Br. J. Sports Med. 57, 849–854 (2023) Jacobsson, J., Spreco, A., Kowalski, J., Timpka, T. & Dahlström, Ö. Assessing parents, youth athletes and coaches subjective health literacy: A cross-sectional study. J. Sci. Med. Sport 24, 627–634 (2021). Mason, R. et al. Wearables for running gait analysis: A systematic review. Sports Med. 53, 241–268 (2023). Provot, T., Chiementin, X., Oudin, E., Bolaers, F. & Murer, S. Validation of a high sampling rate inertial measurement unit for acceleration during running. Sensors 17, 1958; 10.3390/s17091958 (2017). Phinyomark, A., Hettinga, B. A., Osis, S. & Ferber, R. Do intermediate-and higher-order principal components contain useful information to detect subtle changes in lower extremity biomechanics during running? Hum. Mov. Sci. 44, 91–101 (2015). Janssen, M., Scheerder, J., Thibaut, E., Brombacher, A. & Vos, S. Who uses running apps and sports watches? Determinants and consumer profiles of event runners’ usage of running-related smartphone applications and sports watches. PloS One. 12, e0181167; 10.1371/journal.pone.0181167 (2017). Benson, L. C., Clermont, C. A. & Ferber, R. New considerations for collecting biomechanical data using wearable sensors: the effect of different running environments. Front. bioeng. biotechnol. 8, 86; 10.3389/fbioe.2020.00086 (2020). Panebianco, G. P., Bisi, M. C., Stagni, R. & Fantozzi, S. Analysis of the performance of 17 algorithms from a systematic review: Influence of sensor position, analyzed variable and computational approach in gait timing estimation from IMU measurements. Gait Posture 66, 76–82 (2018). Medina, E., Palomares, N., Page, Á. & Bazuelo-Ruiz, B. Analysis of kinematic patterns in runners. An approach based on inertial sensors and functional data analysis. In ISBS-Conference Proceedings Archive (2015). Buckley, C. et al. Binary classification of running fatigue using a single inertial measurement unit. In 2017 IEEE 14th International Conference on Wearable and Implantable Body Sensor Networks (BSN). httos:// doi: 10.1109/BSN.2017.7936040 (2017) Tomabechi, K., Ikegami, Y., Yamamoto, K. & Nakamura, Y. Learning Whole-body Effects for Biomechanics Analysis from Partial IMU Sensing. In 2022 9th IEEE RAS/EMBS International Conference for Biomedical Robotics and Biomechatronics (BioRob). https://doi:10.1109/BioRob52689.2022.9925310 . (2022). Marotta, L., Buurke, J. H., van Beijnum, B. J. F. & Reenalda, J. Towards machine learning-based detection of running-induced fatigue in real-world scenarios: Evaluation of IMU sensor configurations to reduce intrusiveness. Sensors 21, 3451; 10.3390/s21103451 (2021). da Silva Soares, J. et al. Functional data analysis reveals asymmetrical crank torque during cycling performed at different exercise intensities. J. Biomech, 122, 110478; 10.1016/j.jbiomech.2021.110478 (2021). Xu, D. et al. Explaining the differences of gait patterns between high and low-mileage runners with machine learning. Sci. Rep. 12, 2981 (2022). Patoz, A., Lussiana, T. & Breine, B. Non-South East Asians have a better running economy and different anthropometrics and biomechanics than South East Asians. Sci. Rep. 12, 6291 (2022). Table Table.1 Overview of the dataset subject to analysis ID Freq. Distance (km) Target Pace (s/km) Lap Time Mean [SD] Fatigue Median [min, max] Sampling Data (Fatigue No ≦ 12, Yes ≧ 13) Changes in conditions due to practice A 1 19 250 253[9.3] 12[11,14] 390, 180 - 2 21.6 250 252[8.3] 13[10,14] 270, 360 - B 1 19 250 250]5.5] 12[10,13] 300, 270 uncomfortable feeling 2 15 240 216[5.2] 12[9,13] 360, 90 residual fatigue C 1 19 245 252[7.4] 13[10,14] 270, 300 - 2 15 230 215[4.1] 11[9,13] 420, 30 residual fatigue D 1 19 245 243[11.1] 11[9,13] 420, 150 uncomfortable feeling 2 15 225 215[5.6] 11[9,13] 330, 120 - E 1 19 250 249[18.8] 12[11,14] 330, 240 residual fatigue 2 21.6 250 250[11.6] 12[11,13] 510, 120 - F 1 16.8 245 243[9.1] 11[10,13] 420, 90 uncomfortable feeling G 1 16.8 250 248[6.3] 11[9,13] 390, 120 uncomfortable feeling H 1 16.8 250 243[11.6] 12[11,13] 390, 120 - I 1 21.6 255 248[12.2] 12[9,15] 360, 270 uncomfortable feeling, pain J 1 19 255 250[5.5] 12[11,13] 360, 210 - Additional Declarations No competing interests reported. Supplementary Files FDoiSupplementarymaterials240110.pdf Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-3850139","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":267656475,"identity":"627b3993-c2ae-4694-baaa-19d4d34ce7f1","order_by":0,"name":"Hirofumi Doi","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABDElEQVRIie2RsUrEQBCGJywkzUo6maDkDYSBQDwx+CwJAauLXGlx6IJwNrE/uJfQxtYNC6nO/iA29hYpr1jhVtMomJhScD+mmYWPmX8WwGL5ozBNGPrMk7L9+jro8NkkCm7LrFqOV9rL7H4tI8XHLOXvP9eIhI7YpKlKNIRHq5JgOwfvuEcJVhfnSITMWWZSFQuI4pc1OWUN7ET8rFAzjTEldBmaKYWA7HEzJdgTwEgOKJKQu0ZUE90pzvsvSiAIkXNJCtxOYUNTTJb8FMgcwCvT6m6BJks9U4c19mbxD4qqAX11/aQ81W51EsbNzcPr2zzJ+y5mPuVbh59lVsKcRiqd9cFZv2KxWCz/jB0o/FfGf0bUmwAAAABJRU5ErkJggg==","orcid":"","institution":"Doctoral Course, Graduate School of Medicine, Osaka Medical and Pharmaceutical University, Osaka","correspondingAuthor":true,"prefix":"","firstName":"Hirofumi","middleName":"","lastName":"Doi","suffix":""},{"id":267656476,"identity":"0de7faf0-2bb6-4fd4-8a21-9e5f38468711","order_by":1,"name":"Hidetoshi Matsui","email":"","orcid":"","institution":"Faculty of Data Science, Shiga University, Shiga","correspondingAuthor":false,"prefix":"","firstName":"Hidetoshi","middleName":"","lastName":"Matsui","suffix":""},{"id":267656477,"identity":"e697cf96-f02e-449c-aead-d9b7323defda","order_by":2,"name":"Daisuke Nishioka","email":"","orcid":"","institution":"Department of Medical Statistics, Research and Development Center, Osaka Medical and Pharmaceutical University, Osaka","correspondingAuthor":false,"prefix":"","firstName":"Daisuke","middleName":"","lastName":"Nishioka","suffix":""},{"id":267656478,"identity":"dc4089b7-84fe-46bd-8de5-aeb10963fcf3","order_by":3,"name":"Yuri Ito","email":"","orcid":"","institution":"Department of Medical Statistics, Research and Development Center, Osaka Medical and Pharmaceutical University, Osaka","correspondingAuthor":false,"prefix":"","firstName":"Yuri","middleName":"","lastName":"Ito","suffix":""},{"id":267656479,"identity":"147d0d9f-d58f-4899-b269-19e5c689b9ea","order_by":4,"name":"Ryuichi Saura","email":"","orcid":"","institution":"Department of Physical and Rehabilitation Medicine, Division of Comprehensive Medicine, Osaka Medical and Pharmaceutical University, Osaka","correspondingAuthor":false,"prefix":"","firstName":"Ryuichi","middleName":"","lastName":"Saura","suffix":""}],"badges":[],"createdAt":"2024-01-10 11:29:07","currentVersionCode":1,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-3850139/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-3850139/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":49826281,"identity":"5ed91090-9ad3-42e2-946a-26361354ccfc","added_by":"auto","created_at":"2024-01-18 15:46:38","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":1579639,"visible":true,"origin":"","legend":"\u003cp\u003eParticipants\u003c/p\u003e","description":"","filename":"Fig1paticipants.jpg","url":"https://assets-eu.researchsquare.com/files/rs-3850139/v1/435bc907fbaf1b26af6485a4.jpg"},{"id":49826282,"identity":"6c70972c-5f7a-4838-81bc-c1e6bc60322f","added_by":"auto","created_at":"2024-01-18 15:46:38","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":2623727,"visible":true,"origin":"","legend":"\u003cp\u003ePrincipal Component Analysis and Average Running Form by Cluster\u003c/p\u003e\n\u003cp\u003ea. Distribution of principal component scores and clusters representing running form: Analysis participants were grouped into Cls.1 (A, C, D, G, H, I, J), Cls.2 (E, F), and Cls.3 (B).\u003c/p\u003e\n\u003cp\u003eb. Solid lines represent the first principal component function, while dashed lines denote the second principal component function.\u003c/p\u003e\n\u003cp\u003ec. Average functions of acceleration per cluster: The average functions of acceleration waveforms representing running form for all data (All) and for each cluster (Cls.1, Cls.2, Cls.3) are presented.\u003c/p\u003e","description":"","filename":"Fig2PCAandAv.jpg","url":"https://assets-eu.researchsquare.com/files/rs-3850139/v1/c8e29e42794307517340260a.jpg"},{"id":49826718,"identity":"f107b15d-5021-4776-8bf4-c7a354458f2d","added_by":"auto","created_at":"2024-01-18 15:54:38","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":1199301,"visible":true,"origin":"","legend":"\u003cp\u003eRelationship between fatigue levels and principal component scores\u003c/p\u003e\n\u003cp\u003eThe median principal component scores for each measurement are presented and categorized as no fatigue (f0: Borg-RPE \u0026lt; 12), borderline no fatigue (f0b: Borg-RPE = 12), and fatigue (f1: Borg-RPE \u0026gt; 12).\u003c/p\u003e","description":"","filename":"Fig3FatPCA.jpg","url":"https://assets-eu.researchsquare.com/files/rs-3850139/v1/078840337e1364031e3b95e8.jpg"},{"id":49826283,"identity":"3e5415ed-fde6-4884-bae0-abc5450f8b80","added_by":"auto","created_at":"2024-01-18 15:46:38","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":3633363,"visible":true,"origin":"","legend":"\u003cp\u003eIs the baseline for changes in running form a group model or a personal model?\u003c/p\u003e\n\u003cp\u003eResults from the multivariate functional logistic regression analysis (triaxial) and univariate logistic regression analysis (vertical, sagittal, and left–right) are presented. The baseline indicates whether “Person” (no-fatigue data of analysis participants) or “Group” (no-fatigue data of all analysis participants) was included in the function logistic regression analysis. For participants measured twice (A, B, C, D, E), “Person” includes fatigue-free data from both measurements, while for participants measured once (F, G, H, I, J), it includes fatigue-free data from a single measurement. Dev.exp. %: Deviance explained percentage is calculated as 1 - Residual Deviance/Null Deviance. A higher value indicates a better fit, where Residual Deviance approaches zero for a well-fitting model, and Null Deviance for a poor fit. Coefficients (95% CI): The estimated coefficients and 95% confidence intervals (95% CI) for the basis expansion are shown.\u003c/p\u003e","description":"","filename":"Fig4LogisticForest.jpg","url":"https://assets-eu.researchsquare.com/files/rs-3850139/v1/9bfa2802dc7f60153d9dbfce.jpg"},{"id":52896941,"identity":"54166d0e-5c22-47e9-ad3b-0e89bae97c0e","added_by":"auto","created_at":"2024-03-18 13:10:20","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":839694,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-3850139/v1/baf47117-1871-46a5-86b5-d32f0a3a28eb.pdf"},{"id":49826279,"identity":"9181545b-b542-4cd3-a783-58f30d8b31f5","added_by":"auto","created_at":"2024-01-18 15:46:38","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":2260507,"visible":true,"origin":"","legend":"","description":"","filename":"FDoiSupplementarymaterials240110.pdf","url":"https://assets-eu.researchsquare.com/files/rs-3850139/v1/620c9aeb210685f84c5459be.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Visualisation of running form changes measured by wearable sensors for conditioning management, an application of the Functional Data Analysis","fulltext":[{"header":"Introduction","content":"\u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eRunning is considered as a popular activity. Runners set diverse goals, including improving personal best times, enhancing running performance through distance adjustments, refining diet and health indicators, engaging with fellow enthusiasts, and alleviating stress [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. Aspiring to achieve a slightly better state with each run, runners constantly face the challenge of confirming whether their running performance has surpassed their previous record.\u003c/p\u003e \u003cp\u003eMoreover, it is well-established that a majority of runners experience Running-Related Injuries (RRIs) [\u003cspan additionalcitationids=\"CR4 CR5 CR6\" citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. Extensive research has been conducted on factors contributing to RRIs, encompassing variables such as time and distance of exposure, exercise load and environment, running skills, fatigue, and history of trauma [\u003cspan additionalcitationids=\"CR9 CR10 CR11 CR12 CR13 CR14 CR15 CR16 CR17 CR18\" citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. In popular marathons, participants are subject to significant exposure in terms of both distance and time, leading to biomechanical changes [\u003cspan additionalcitationids=\"CR21 CR22\" citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e]. Maas et al. highlighted that novice runners undergo substantial alterations in their running strategies driven by fatigue during long-distance running [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]. Therefore, can runners effectively assess whether biomechanical changes during races or training align with their expectations?\u003c/p\u003e \u003cp\u003eNumerous biomechanical analyses have been conducted in running using measurement equipment in tightly-controlled environments [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. Interpreting the analysis results, including representative values and waveforms of elements such as force and angles, requires specialized knowledge, rendering it impractical to acquire and apply such information regularly. Conversely, the information available to runners in their daily routines consists of simple metrics (such as heart rate and cadence) provided by running watches and smartphone applications, observational data from coaches, and confirmation of the running form from videos. Readily-accessible information concerning changes in the running biomechanics of runners and coaches is limited. As a potential example, standard models or running forms of elite athletes could potentially serve as evaluation criteria. However, a universally-shared running form that is a common goal applicable to all individuals and situations is highly improbable [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]. This can be attributed to the variation in runners' skeletal structure, even when factors such as race, age, and sex are consistent alongside differing background factors, including past training, experience, upbringing, medical history, footwear, and ground conditions. Furthermore, the lack of established means for acquiring and evaluating running form data over an extended period in real-world running environments makes it challenging for runners to comprehend trends in their own running form.\u003c/p\u003e \u003cp\u003eOur objective is to propose a versatile method for daily visualization and categorization of the running form. As one device, Inertial Measurement Unit (IMU) are attracting attention. IMU sensors integrated into running watches and smartphones have found applications in activity measurement, offering biomechanical indicators through running apps and supporting research activities, rendering them suitable for daily visualization [\u003cspan additionalcitationids=\"CR29 CR30\" citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]. Functional Data Analysis (FDA) is one of the most advantageous methods of analysis when considering the maximum use of data collected by IMUs over time. FDA involves the transformation of time-series observations for individuals measured at discrete time points into functions (values), and employs statistical methods to analyze a set of functional data [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e]. Time-series data capturing continuous running motions are recorded as discrete time point data, necessitating the capture of changes occurring over time as a coherent set of information encompassing the entire running cycle. FDA has the ability to handle datasets with different sampling rates and variations in the number of time points in a measured operating cycle. This capability is particularly advantageous when dealing with multiple time-series data comprising different observation time points and number of observation time points. We investigated the application of the widely-used Inertial Measurement Unit (IMU) sensors and Functional Data Analysis (FDA) to develop visualization methods accessible to the general running community.\u003c/p\u003e \u003cp\u003eIn this study, we explored the possibility of visualizing changes in fatigue and running form induced by long-distance pace training using FDA.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv class=\"BlockQuote\"\u003e\n \u003cp\u003eIn the track-and-field team of the high school, we observed the running form of athletes participating in long-distance pace training, which involved running distances of 15 km or more at a consistent speed. The specifics of the training, including distance and target pace, were determined through discussions between the coaches and athletes. The observations were conducted during four practice sessions in March 2021. Regarding data related to the running form, we acquired phase and actual measurements of acceleration and angular velocity from IMU sensors affixed near the human center of mass in the lumbar region.\u003c/p\u003e\n\u003c/div\u003e\n\u003cp\u003eWe securely attached IMU sensors (VICON BLUETRIDENT) just above the participants\u0026apos; L5/S1 (5th Lumbar Sacral Joint) using a waist belt. Acceleration and angular velocity were measured at a frequency of 1125 Hz, whereas geomagnetic data were captured at 100 Hz. Fatigue levels were assessed using the Borg-RPE (15-point Borg rating of perceived exertion scale), and data were obtained at 1-km intervals [\u003cspan class=\"CitationRef\"\u003e33\u003c/span\u003e\u0026ndash;\u003cspan class=\"CitationRef\"\u003e35\u003c/span\u003e]. The presence of fatigue was defined as Borg scores of 12 or below, indicating \u0026quot;no fatigue,\u0026quot; and Borg scores of 13 or above, indicating \u0026quot;fatigue.\u0026quot; Informed participants received training in advance on how to measure fatigue using the Borg RPE scale. When there was a change in participants\u0026apos; fatigue levels, they signed an observer, and the recall method was employed after the practice session for confirmation. Data collection included measurements of height, weight, circumference (thigh and calf), lower limb length, and joint range of motion (hip, knee, and ankle), as well as subjective condition assessments before and after practice on measurement day. These included long-lasting fatigue, discomfort, and pain after practice. The researchers evaluated the differences before and after these aspects. The subjective conditions before practice were collected after finalizing the details of the training session on that day. However, participants who experienced severe fatigue from the previous day\u0026apos;s practice or experienced pain chose not to participate in observations on the current day.\u003c/p\u003e\n\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\n \u003ch2\u003eParticipants\u003c/h2\u003e\n \u003cdiv class=\"BlockQuote\"\u003e\n \u003cp\u003eThe study included 17-year-old male high school students specializing in long-distance track-and-field. Information sessions were conducted with prospective participants and their guardians, where the purpose of the study was explained, and consent for participation was obtained. Participants who altered or discontinued the training session for reasons such as pain were excluded from the observation on that day.\u003c/p\u003e\n \u003c/div\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\n \u003ch2\u003eStatistical Method\u003c/h2\u003e\n \u003cdiv class=\"BlockQuote\"\u003e\n \u003cp\u003eStatistical methods were based on the future development of applications on devices such as smartphones. Python 3.9 and R 4.1.0 were employed for statistical processing.\u003c/p\u003e\n \u003cp\u003eThe data obtained from the IMU sensor underwent preprocessing, including noise reduction and transformation from sensor coordinates to spatial coordinates. One complete running cycle was defined as right foot contact\u0026ndash;jump\u0026ndash;left foot contact\u0026ndash;jump, and acceleration data corresponding to one running cycle was treated as an individual observation. We treat the time-series data extracted from a specific 1 km segment as acceleration data representing the running form of that segment. The number of cycles of time-series data extracted from the 1 km segment was set to 30. Specifically, linear segments were identified from the geomagnetic data, and for each 1 km interval, three linear running segments (approximately 70 seconds after the start of the 1 km interval, approximately 100 seconds after, and approximately 130 seconds after) were chosen. Time-series data for 10 running cycles were obtained for each linear segment, resulting in 30 time-series datasets per 1 km interval. The extracted dataset (distance for each individual \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\times\\)\u003c/span\u003e\u003c/span\u003e 3 extraction points \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\times\\)\u003c/span\u003e\u003c/span\u003e 10 cycles) were transformed into functions for each cycle and standardization of the time points using functional data analysis. One way of transforming the observed data into functions is to use basis function expansion, i.e., the function \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(x\\left(t\\right)\\)\u003c/span\u003e\u003c/span\u003eis expressed as the sum of \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(m\\)\u003c/span\u003e\u003c/span\u003e basis functions with coefficients \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({w}_{1},\\dots ,{w}_{m}\\)\u003c/span\u003e\u003c/span\u003e and basis functions\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({\\phi }_{1}\\left(t\\right),\\dots ,{\\phi }_{m}\\left(t\\right)\\)\u003c/span\u003e\u003c/span\u003e as follows;\u003c/p\u003e\n \u003c/div\u003e\n \u003cdiv id=\"Equa\" class=\"Equation\"\u003e\n \u003cdiv class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e$$x\\left(t\\right)={w}_{1}{\\phi }_{1}\\left(t\\right)+\\dots +{w}_{m}{\\phi }_{m}\\left(t\\right)+\\epsilon .$$\u003c/div\u003e\n \u003c/div\u003e\n \u003cdiv class=\"BlockQuote\"\u003e\n \u003cp\u003eIt then allows the change in acceleration over a running cycle to be expressed as an acceleration function \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(x\\left(t\\right)\\)\u003c/span\u003e\u003c/span\u003e that varies over a standardized time \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(t\\)\u003c/span\u003e\u003c/span\u003e. Accordingly, the set of acceleration data for a running cycle can be treated as an acceleration function \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(x\\left(t\\right)\\)\u003c/span\u003e\u003c/span\u003e generated for each running cycle. B-splines were used as basis functions.\u003c/p\u003e\n \u003cp\u003eFunctional Principal Component Analysis (FPCA) was used to evaluate participants\u0026apos; running forms [\u003cspan class=\"CitationRef\"\u003e36\u003c/span\u003e]. Cluster analysis was performed using k-means and Ward\u0026apos;s methods on the principal component scores obtained. As the goal was to partition the data into individual measurement sessions, the number of clusters to which measurement sessions belong was determined based on the condition that at least two-thirds of the data from one measurement session should be included in a single cluster. Additionally, the distribution and trends of the scores of the first and second principal components were assessed from the results of FPCA. As fatigue is expected to cause gradual changes in running form near the limit, we categorized fatigue levels into \u0026ldquo;no-fatigue\u0026quot; (Borg 11 or below), \u0026quot;Borderline no-fatigue\u0026quot; (Borg 12), and \u0026ldquo;fatigue\u0026quot; (Borg 13 and above). We then examined the relationship between fatigue levels and principal component scores. To evaluate the relationship between fatigue and running form, we established two baseline models to determine the presence of fatigue: a group model (consisting of no-fatigue data from all participants) and an individual model (no-fatigue data from an individual participant).\u003c/p\u003e\n \u003cp\u003eThe relationship between running form and fatigue was modeled by Functional Logistic Regression Analysis, where univariate (in each direction) and multivariate analyses (across triaxial) are conducted [\u003cspan class=\"CitationRef\"\u003e37\u003c/span\u003e]. In these analyses, we compared either the group or individual baseline (no-fatigue data) with the individual fatigue data obtained in a single measurement.\u003c/p\u003e\n \u003c/div\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec5\" class=\"Section2\"\u003e\n \u003ch2\u003eEthical Considerations\u003c/h2\u003e\n \u003cdiv class=\"BlockQuote\"\u003e\n \u003cp\u003eThe study was designed in compliance with the Ethical Guidelines for Medical Research Involving Human Subjects and the Personal Information Protection Law, and conducted with approval from the Ethics Committee of Osaka University of Medical and Pharmaceutical Sciences (2020\u0026thinsp;\u0026minus;\u0026thinsp;172). The study objective was explained to prospective participants and their guardians at a briefing session, and informed consent was obtained from both prospective participants and their guardians who agreed to participate voluntarily.\u003c/p\u003e\n \u003c/div\u003e\n\u003c/div\u003e"},{"header":"Results","content":"\u003cp\u003eMeasurements were conducted during four practice sessions in March 2021. Consent for participation in the study was obtained from all 13 male high school students, aged 17, specialized in long-distance track-and-field. Twelve students (n = 12) participated in the measurements after excluding one who met the exclusion criteria owing to practice restrictions associated with RRIs, resulting in 24 datasets (d = 24). Data from two cases where fatigue did not reach Borg 13 (d = 2), three cases of IMU sensor malfunction (d = 3), two cases with incomplete interval time or fatigue data acquisition (d = 2), and two cases of early fatigue appearing immediately after practice initiation (d = 2) were excluded from the analysis (Fig.1). Essentially, 15 datasets (d = 15) from 10 participants (n = 10) were considered.\u003c/p\u003e\n\u003cp\u003eBody measurements, including thigh and calf circumferences, lower limb length, and joint range of motion (hip, knee, ankle), could not be obtained due to COVID-19 prevention guidelines at the high school where the participants were enrolled, as they required direct physical contact. Therefore, height and weight were collected through self-reporting. Consequently, data on circumference (thigh and calf), lower limb length, and joint range of motion (hip, knee, and ankle) were not available. The long-distance pace training sessions were tailored to the participants' training goals and abilities, resulting in distances ranging from 15.0 to 21.6 km and pace timing ranging from 230 to 255 seconds per km (refer to Table1). The change in fatigue due to long-distance pace training began from Borg 9–11 and reached Borg 13–15. The extracted data included no fatigue data (270–510) and fatigue data (30–360) for each measurement. Among the five participants who were measured twice, four had different subjective conditions before and after the long-distance pace training sessions.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eClassification of running forms\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe waveform representing running form was examined based on the first and second principal components obtained from multivariate FPCA for the triaxial directions (refer to Fig.2 and Supplementary Equation S1). The cumulative contribution of the first and second principal components in the multivariate analysis was 79.24%. The first principal component contained cyclic components of the left stance phase in the vertical and sagittal directions, whereas the second principal component contained components of the right stance phase in the vertical and sagittal directions as well as the left–right direction.\u003c/p\u003e\n\u003cp\u003eClustering was performed on the multivariate data, including components in each direction (vertical, sagittal, left-right). Through exploration using K-means and Ward methods, the number of clusters between 3 and 5 was likely, and clusters containing data from only one measurement session appeared for clusters above 6 (refer to Supplementary Fig.S2). With 4 or 5 clusters, some measurement data are dispersed among multiple clusters, making it difficult to determine the appropriate cluster assignment. With 3 clusters, 14 out of 15 measurements were clearly classifiable, and even for the remaining one measurement, 69.61% of the data were attributed to one cluster, leading us to choose 3 clusters (refer to Supplementary Fig.S2). When considering cluster Cls.1 with the highest number of participants as a reference, distinct characteristics were observed: cluster Cls.2 had lower scores in the first principal component, whereas cluster Cls.3 exhibited higher scores in the second principal component.\u003c/p\u003e\n\u003cp\u003eThe waveforms of participants who were measured twice were closely distributed and classified into the same cluster (refer to Supplementary Fig.S3). The relationship between fatigue levels and median principal component scores varied among the participants for each measurement (see Fig.3 and Supplementary Fig.S4). For instance, in the case of Participant B, the primary change in waveform, particularly in the horizontal direction, corresponded to the fatigue level in B1, whereas in B2, the change was observed in the second principal component (vertical direction). Furthermore, some participants such as D1 and D2 exhibited differences in the magnitude of displacement, whereas others such as Participant G showed minimal changes in displacement. Moreover, the changes in principal components over time varied among participants, and participants who underwent the two measurements twice did not necessarily exhibit the same trends.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eAssessment of fatigue and running form\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFatigue and running were assessed by evaluating the association between fatigue induced by long-distance running and acceleration waveforms representing the running form. This was performed using both univariate and multivariate functional logistic regression analyses (refer to Fig.4, Supplementary Equation S1 and Supplementary Fig.S5). The results of the univariate functional logistic regression analyses conducted for each direction, along with the multivariate functional logistic regression analysis considering all three directions, revealed associations between waveforms and fatigue in at least one of the univariate or multivariate analyses. However, even if the strength of the relationship was confirmed in the univariate analysis, it did not necessarily manifest in the multivariate analysis.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eIn this study, we attempted to capture the variations and characteristics of running form using IMU sensors attached to the waist. We utilized both the magnitude and phase of acceleration and angular velocity near the center of mass, and applied FDA to visualize these aspects of the running form. Although previous studies have been traditionally relying on specialized motion analysis equipment and expertise for interpretation of results, attempts have been made in recent years to capture motion using IMU sensors. Studies have attempted to validate the assessment of whole-body biomechanics using partial acceleration data from various body parts, including the hips, back, wrists, and feet [\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e]. Human movement arises from temporal changes in posture, involving both limbs and the trunk, and manifests as motion within the spatial dimensions of the center of mass. Typically, the center of mass in an upright posture is believed to be located in the lower part of the torso. However, it is important to note that the center of mass is not fixed and can shift with changes in posture.\u003c/p\u003e\u003cp\u003eThe clustering of principal component scores derived from multivariate FPCA resulted in the classification of acceleration waveform patterns representing the running form into three distinct clusters. Although principal component scores were plotted in close proximity for each participant, their positions varied. Among the participants who underwent measurements on two occasions, the second set of principal component scores was also classified into the same cluster as the first set and their positions were also close. This suggests that within individual participants, the principal component scores may exhibit relatively similar distributions (see Supplementary Fig.S4). However, the relationship between acceleration waveform patterns representing running form and fatigue did not demonstrate consistent trends or convergence. Subsequently, we computed the average acceleration waveform representing the running form for both the entire participant group and each cluster. The results indicated a similarity in waveform patterns between the entire participant group and Cluster Cls.1, while Clusters Cls.2 and Cls.3 displayed distinct waveform shapes, each exhibiting unique characteristics. It is worth noting that these three clusters were derived from a relatively small sample size, suggesting that a larger-scale investigation in the future may unveil a greater number of distinct running form clusters. Previous studies have proposed various running-form categories, including grounding patterns, running ability, and sex, with a wide range of possible combinations [\u003cspan additionalcitationids=\"CR41 CR42\" citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e]. This implies that variations in the running form among runners should be considered. It may be inappropriate to discuss current performance or future directions simply by comparing them to benchmarks, such as elite runners or group averages.\u003c/p\u003e\u003cp\u003eMoreover, it is crucial to note that this study specifically focused on a group of 17-year-old male athletes engaged in the same track-and-field discipline and training environment. Consequently, subtle differences between the participants are visualized. Similar to the characteristics of conventional principal component analysis, the features of the evaluated dataset can influence the principal component functions obtained from functional principal component analysis. For example, if participant data were assessed alongside elite athletes or recreational runners, the differences in acceleration waveforms representing the running forms of the participants in this study may be perceived as relatively limited. This underscores the importance of understanding the features of the dataset being evaluated according to the intended purpose and interpreting the results appropriately. Furthermore, to visualize individual performance and condition management, it is advisable to consider readily-available running-related motion indicators, such as cadence and walking speed. It would be feasible to conduct analyses that incorporate an individual\u0026rsquo;s past and present data, compare groups and individuals, and target specific individuals.\u003c/p\u003e\u003cp\u003eDetecting fatigue-related changes in running form is critical. Running coaches routinely assess athletes' condition and offer guidance based on their current running form. Simultaneously, many runners seek to improve their performance by referring to their own or others' running forms. Essentially, running form plays a pivotal role in comprehending conditions and enhancing performance. However, Robbie et al. asserted that even with the trained eye of a running coach, visually classifying running economy by long-distance running coaches remains a challenging task, and assessment of running form is not straightforward [\u003cspan additionalcitationids=\"CR45\" citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e]. The findings confirm that the trend of change in principal component scores of FCPA with changes in fatigue is not constant, confirming that it is difficult to fit the trend of change into a standardized pattern (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). In this study, two models were constructed: a group model based on fatigue-free data from all participants and an individual model based on fatigue-free data from each participant. The results of functional logistic regression analysis revealed a relationship between the presence or absence of fatigue and running form in various directions, either singularly or in multiples.\u003c/p\u003e\u003cp\u003eThe difference between the group and individual model baselines did not consistently yield conclusive results, emphasizing the need for careful interpretation (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). The variations in running form based on the presence or absence of fatigue, as observed through comparison with the group model, may have been influenced by individual variations in running form among the participants (see Supplementary Fig.S6). Both univariate and multivariate results frequently demonstrated that the group model outperformed the individual model in terms of deviance-explained% and ROC values. This suggests that, when estimating the presence of fatigue from the running form, it is crucial to recognize the limitations of comparing oneself with others or extrapolating the results to others.\u003c/p\u003e\u003cp\u003eThe benefits and challenges of visualizing the running form are explored with the aspiration that all runners, regardless of whether they are elite athletes, students, or recreational runners, can monitor their daily condition and respond appropriately. Recognizing the signs of changes in health conditions or indicators, such as RRIs, and making decisions to adjust training or rest is a complex task [\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e, \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e]. With the methodology used in this study, obtaining results commonly used in motion analysis, such as joint angles or force intensity, is not feasible. However, by utilizing acceleration data from IMU sensors and applying functional data analysis, it is possible to visualize where changes in the running form occur within the running cycle and how they compare with past data. Although this visualized information may seem insignificant, it can serve as a guideline for managing running forms and cultivating interest in one's own condition and health literacy. Furthermore, if visualization using real-world data generated by runners worldwide during their daily runs becomes possible, it could serve as a benchmark for running form management. Consequently, if there are concerns about daily changes in physical condition, taking proactive steps to seek advice from running coaches or healthcare professionals is recommended. In the future, it would be desirable to establish an environment in which all runners can recognize changes in their condition and respond accordingly.\u003c/p\u003e\u003cp\u003eTo elucidate the relationship between various factors such as conditions, performance, individual factors of the subjects, footwear, surface, weather, psychological aspects, and acceleration waveform, it is imperative to conduct repeated investigations in actual running environments. This study presents a preliminary exploration of visualization techniques using FDA. Further investigations are warranted to explore the application of this methodology to data from wrist-worn running watches or smartphones as well as devices attached to different body parts. In addition, ensuring the mutual compatibility of data from IMU devices attached to different body segments remains a crucial challenge for future research. The FPCA conducted in this study creates and classifies the principal component functions based on all available information, providing a relative visualization of existing data. Essentially, it is not an absolute evaluation metric, necessitating the collection of information from all subjects for comparison, including data from athletes or Olympians, which serve as ideal benchmarks. New methodologies that provide absolute evaluation metrics for all subjects are desired to enable a straightforward visualization of runners' conditions and performance.\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e"},{"header":"Limitations","content":"\u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eThis study has several limitations. The relationship between fatigue induced by long-distance running and the running forms reported in this study may potentially include unmeasured confounders and errors. This study had limited participants, and factors such as running skill, sex, race, and skeletal structure were not considered. Moreover, the limited number of measurements within the participants prevented a thorough examination of the effects of the conditions, performance, and environment, as well as reproducibility. Therefore, extensive validation is necessary to generalize the results to other participants. Although geomagnetic corrections were implemented in this study, GPS corrections were not applied, which led to an accumulation of errors specific to the IMU-based methodology. This could potentially be addressed through error corrections using technologies such as GPS and higher-performance IMUs, which may yield more accurate results.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e"},{"header":"Conclusions","content":"\u003cp\u003eThe results of this study demonstrate that utilizing general-purpose sensors and FDA techniques, rather than specialized motion analysis devices, allows the visualization of changes and characteristics in a runner's running form.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe characteristics of running form and changes in running form associated with the onset of fatigue are not uniform, highlighting the importance of understanding the changes and features of running form within individuals, rather than solely comparing them with standard models or group averages, in the evaluation of general running form.\u003c/p\u003e\n\u003cp\u003eThe visualization of changes and characteristics in running form holds significant potential for applications such as adjusting training regimens and volumes, managing performance (physical capabilities), monitoring overall physical and mental conditions, and identifying signs of RRIs.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eData Availability Statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eDatasets generated and/or analyzed during this study are not publicly available due to consent conditions with the study subjects, but are available upon reasonable request from the corresponding author H.D\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe very grateful to the various members of the Department of Rehabilitation Medicine and the Department of Medical Statistics at Osaka Medical and Pharmaceutical University for their scientific advice and insightful discussions during the conduct of this study. We would like to thank Editage (www.editage.jp) for English language editing.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eH.D. conceived this study under the guidance of R.S., H.M., Y.I., N.D.. Y.I., assisted H.D. in developing the statistical analysis plan. H.D performed the statistical analysis under H.M.'s guidance. H.M. contributed to interpretation of the results. H.D. prepared the manuscript, figures and tables. R.S. supervised the conduct of the study. All authors reviewed and critically revised the manuscript for intellectual content. All the authors approved the final version of the manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe author(s) declare no competing interests.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eShipway, R. \u0026amp; Holloway, I. Running free: Embracing a healthy lifestyle through distance running. Perspect. Public Health. 130, 270\u0026ndash;276 (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJanssen, M. \u003cem\u003eet al\u003c/em\u003e. Understanding different types of recreational runners and how they use running-related technology. Int. J. Environ. Res. Public Health. 17, 2276; \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/ijerph17072276\u003c/span\u003e\u003cspan address=\"10.3390/ijerph17072276\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKakouris, N., Yener, N. \u0026amp; Fong, D. T. A systematic review of running-related musculoskeletal injuries in runners. J. Sport Health Sci. 10, 513\u0026ndash;522 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDesai, P. I. A., Jungmalm, J., B\u0026ouml;rjesson, M., Karlsson, J., \u0026amp; Grau, S. Recreational runners with a history of injury are twice as likely to sustain a running-related injury as runners with no history of injury: a 1-year prospective cohort study. J. Orthop. Sports Phys. Ther. 51, 144\u0026ndash;150 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eToresdahl, B. \u003cem\u003eet al.\u003c/em\u003e Factors associated with injuries in first-time marathon runners from the New York City marathon. Phys. Sportsmed. 50, 227\u0026ndash;232 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMayne, R. S., Bleakley, C. M. \u0026amp; Matthews, M. Use of monitoring technology and injury incidence among recreational runners: a cross-sectional study. BMC Sports Sci. Med. Rehabil. 13, 1\u0026ndash;7 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDempster, J., Dutheil, F. \u0026amp; Ugbolue, U. C. The Prevalence of Lower Extremity Injuries in Running and Associated Risk Factors: A Systematic Review. J. Phys. Act. Health Title. 5, 133\u0026ndash;145 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHulme, A., Nielsen, R. O. Timpka, T. Verhagen, E. \u0026amp; Finch, C. Risk and protective factors for middle-and long-distance running-related injury. Sports Med. 47, 869\u0026ndash;886 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSaragiotto, B. T. \u003cem\u003eet al\u003c/em\u003e. What are the main risk factors for running-related injuries? Sports Med. 44, 1153\u0026ndash;1163 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVideb\u0026aelig;k, S., Bueno, A. M. Nielsen, R. O. \u0026amp; Rasmussen, S. Incidence of running-related injuries per 1000 h of running in different types of runners: a systematic review and meta-analysis. Sports Med. 45, 1017\u0026ndash;1026 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVan Poppel, D. \u003cem\u003eet al.\u003c/em\u003e Risk factors for overuse injuries in short-and long-distance running: A systematic review. J. Sport Health Sci. 10, 14\u0026ndash;28 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKluitenberg, B., van Middelkoop, M. Diercks, R. \u0026amp; van der Worp. H. What are the differences in injury proportions between different populations of runners? A systematic review and meta-analysis. Sports Med. 45, 1143\u0026ndash;1161 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNielsen, R. O., Buist, I., S\u0026oslash;rensen, H., Lind, M. \u0026amp; Rasmussen, S. Training errors and running related injuries: a systematic review. Int. J. Sports Phys. Ther. 7, 58\u0026ndash;75 (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDaoud, A. I. \u003cem\u003eet al\u003c/em\u003e. Foot strike and injury rates in endurance runners: a retrospective study. Med. Sci. Sports Exerc. 44, 1325\u0026ndash;1334 (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTenforde, A. S. \u003cem\u003eet al\u003c/em\u003e. Overuse injuries in high school runners: lifetime prevalence and prevention strategies. \u003cem\u003ePM\u0026amp;R\u003c/em\u003e. 3, 125\u0026ndash;131 (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGoss, D. L. \u0026amp; Gross, M. T. A review of mechanics and injury trends among various running styles. US Army Med. Dep. J. 3, 62\u0026ndash;71 (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKemler, E. \u0026amp; Huisstede, B. Performance goals of runners are associated with the occurrence of running-related injuries. Phys. Ther. Sport. 50, 153\u0026ndash;158 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWillwacher, S. \u003cem\u003eet al.\u003c/em\u003e Running-related biomechanical risk factors for overuse injuries in distance runners: a systematic review considering injury specificity and the potentials for future research. \u003cem\u003eSports Med\u003c/em\u003e. 52, 1863\u0026ndash;1877 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDesai, P. I. A., Jungmalm, J., B\u0026ouml;rjesson, M., Karlsson, J. \u0026amp; Grau, S. Recreational runners with a history of injury are twice as likely to sustain a running-related injury as runners with no history of injury: a 1-year prospective cohort study. J. Orthop. Sports Phys. Ther. 51, 144\u0026ndash;150 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eReenalda, J., Maartens, E. Homan, L. \u0026amp; Buurke, J. J. Continuous three dimensional analysis of running mechanics during a marathon by means of inertial magnetic measurement units to objectify changes in running mechanics. J. Biomech. 49, 3362\u0026ndash;3367 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eApte, S. \u003cem\u003eet al\u003c/em\u003e. Biomechanical response of the lower extremity to running-induced acute fatigue: a systematic review. Front. Physiol. 12, 646042; \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fphys.2021.646042\u003c/span\u003e\u003cspan address=\"10.3389/fphys.2021.646042\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZinner, C. \u0026amp; Sperlich, B. \u003cem\u003eMarathon Running: Physiology, Psychology, Nutrition and Training Aspects\u003c/em\u003e (ed. Sperlich, B.) (Springer 2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eClermont, C. A., Benson, L. C., Edwards, W. B., Hettinga, B. A. \u0026amp; Ferber, R. New considerations for wearable technology data: changes in running biomechanics during a marathon. J. Appl. Biomech. 35, 401\u0026ndash;409 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMaas, E., De Bie, J., Vanfleteren, R., Hoogkamer, W. \u0026amp; Vanwanseele, B. Novice runners show greater changes in kinematics with fatigue compared with competitive runners. Sports Biomech. 17, 350\u0026ndash;360 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRobertson, G. E., Caldwell, G. E., Hamill, J., Kamen, G. \u0026amp; Whittlesey, S. N. \u003cem\u003eResearch methods in Biomechanics.\u003c/em\u003e Human Kinetics (website) (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePatoz, A., Lussiana, T., Breine, B., Gindre, C. \u0026amp; H\u0026eacute;bert-Losier, K. There is no global running pattern more economic than another at endurance running speeds. Int. J. Sports Physiol. Perform. 17, 659\u0026ndash;662 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHamill, J. \u0026amp; Gruber, A. H. Is changing foot strike pattern beneficial to runners? J. Sport Health Sci. 6, 146\u0026ndash;153 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMigueles, J. H. \u003cem\u003eet al\u003c/em\u003e. GRANADA consensus on analytical approaches to assess associations with accelerometer-determined physical behaviors (physical activity, sedentary behavior and sleep) in epidemiological studies. Br. J. Sports Med. 56, 376\u0026ndash;384 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBenson, L. C., Clermont, C. A., Bošnjak, E. \u0026amp; Ferber, R. The use of wearable devices for walking and running gait analysis outside of the lab: A systematic review. Gait Posture 63, 124\u0026ndash;138 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAhamed, N. U., Kobsar, D., Benson, L. C., Clermont, C. A., Osis, S. T. \u0026amp; Ferber, R. Subject-specific and group-based running pattern classification using a single wearable sensor. J. Biomech. 84, 227\u0026ndash;233 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFaber, G. S., Chang, C. C., Kingma, I. D. S. A. R. T., Dennerlein, J. T. \u0026amp; Van Die\u0026euml;n, J. H. Estimating 3D L5/S1 moments and ground reaction forces during trunk bending using a full-body ambulatory inertial motion capture system. J. Biomech. 49, 904\u0026ndash;912 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRamsay, J. O. \u0026amp; Silverman, B. W. \u003cem\u003eFunctional Data Analysis\u003c/em\u003e (\u003cem\u003e2nd ed.)\u003c/em\u003e (Springer-Verlag, 2005).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBorg, G. \u003cem\u003eBorg's Perceived Exertion and Pain Scales\u003c/em\u003e (Human Kinetics, 1998)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWilliams, N. The Borg rating of perceived exertion (RPE) scale. \u003cem\u003eOccup. Med.\u003c/em\u003e 67, 404\u0026ndash;405 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBorg, G. A. Psychophysical bases of perceived exertion. Med. Sci. Sports Exerc. 14, 377\u0026ndash;381 (1982).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHapp, C. \u0026amp; Greven, S. Multivariate functional principal component analysis for data observed on different (dimensional) domains.\u0026lt;ivertical-align:sub;\u0026gt; \u0026lt;/ivertical-align:sub;\u0026gt;J. Am. Stat. Assoc. 113, 649\u0026ndash;659 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGoldsmith, J., Bobb, J., Crainiceanu, C. M., Caffo, B. \u0026amp; Reich, D. Penalized functional regression. J. Comput. Graph. 20, 830\u0026ndash;851 (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWixted, A. J., Billing, D. C. \u0026amp; James, D. A. Validation of trunk mounted inertial sensors for analysing running biomechanics under field conditions, using synchronously collected foot contact data. \u003cem\u003eSports Eng\u003c/em\u003e. 12, 207\u0026ndash;212 (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLee, M. \u0026amp; Park, S. Estimation of three-dimensional lower limb kinetics data during walking using machine learning from a single IMU attached to the sacrum. Sensors 20, 6277; \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/s20216277\u003c/span\u003e\u003cspan address=\"10.3390/s20216277\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAlmeida, M. O., Davis, I. S. \u0026amp; Lopes, A. D. Biomechanical differences of foot-strike patterns during running: a systematic review with meta-analysis. J. Orthop. Sports Phys. Ther. 45, 738\u0026ndash;755 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eClermont, C. A., Osis, S. T., Phinyomark, A. \u0026amp; Ferber, R. Kinematic gait patterns in competitive and recreational runners. J. Appl. Biomech. 33, 268\u0026ndash;276 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eClermont, C. A., Phinyomark, A., Osis, S. T. \u0026amp; Ferber, R. Classification of higher-and lower-mileage runners based on running kinematics. J. Sport Health Sci. 8, 249\u0026ndash;257 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEskofier, B. M., Kraus, M., Worobets, J. T., Stefanyshyn, D. J. \u0026amp; Nigg, B. M. Pattern classification of kinematic and kinetic running data to distinguish gender, shod/barefoot and injury groups with feature ranking. Comput. Methods Biomech. Biomed. Eng. 15, 467\u0026ndash;474 (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCochrum, R. G. \u003cem\u003eet al\u003c/em\u003e. Visual classification of running economy by distance running coaches. Eur. J. Sport Sci. 21, 1111\u0026ndash;1118 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBenson, L. C., Ahamed, N. U., Kobsar, D. \u0026amp; Ferber, R. New considerations for collecting biomechanical data using wearable sensors: Number of level runs to define a stable running pattern with a single IMU. J. Biomech. 85, 187\u0026ndash;192 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAhamed, N. U. \u003cem\u003eet al.\u003c/em\u003e Subject-specific and group-based running pattern classification using a single wearable sensor. J. Biomech. 84, 227\u0026ndash;233 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTimpka, T. \u003cem\u003eet al\u003c/em\u003e. Injury acknowledgement by the reduction of sports load in world-leading athletes (track and field) varies with their musculoskeletal health literacy and socioeconomic environment. Br. J. Sports Med. 57, 849\u0026ndash;854 (2023)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJacobsson, J., Spreco, A., Kowalski, J., Timpka, T. \u0026amp; Dahlstr\u0026ouml;m, \u0026Ouml;. Assessing parents, youth athletes and coaches subjective health literacy: A cross-sectional study. J. Sci. Med. Sport 24, 627\u0026ndash;634 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMason, R. \u003cem\u003eet al.\u003c/em\u003e Wearables for running gait analysis: A systematic review. Sports Med. 53, 241\u0026ndash;268 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eProvot, T., Chiementin, X., Oudin, E., Bolaers, F. \u0026amp; Murer, S. Validation of a high sampling rate inertial measurement unit for acceleration during running. \u003cem\u003eSensors\u003c/em\u003e 17, 1958; \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/s17091958\u003c/span\u003e\u003cspan address=\"10.3390/s17091958\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePhinyomark, A., Hettinga, B. A., Osis, S. \u0026amp; Ferber, R. Do intermediate-and higher-order principal components contain useful information to detect subtle changes in lower extremity biomechanics during running? Hum. Mov. Sci. 44, 91\u0026ndash;101 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJanssen, M., Scheerder, J., Thibaut, E., Brombacher, A. \u0026amp; Vos, S. Who uses running apps and sports watches? Determinants and consumer profiles of event runners\u0026rsquo; usage of running-related smartphone applications and sports watches. PloS One. 12, e0181167; \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1371/journal.pone.0181167\u003c/span\u003e\u003cspan address=\"10.1371/journal.pone.0181167\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBenson, L. C., Clermont, C. A. \u0026amp; Ferber, R. New considerations for collecting biomechanical data using wearable sensors: the effect of different running environments. Front. bioeng. biotechnol. 8, 86; \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fbioe.2020.00086\u003c/span\u003e\u003cspan address=\"10.3389/fbioe.2020.00086\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePanebianco, G. P., Bisi, M. C., Stagni, R. \u0026amp; Fantozzi, S. Analysis of the performance of 17 algorithms from a systematic review: Influence of sensor position, analyzed variable and computational approach in gait timing estimation from IMU measurements. Gait Posture 66, 76\u0026ndash;82 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMedina, E., Palomares, N., Page, \u0026Aacute;. \u0026amp; Bazuelo-Ruiz, B. Analysis of kinematic patterns in runners. An approach based on inertial sensors and functional data analysis. In \u003cem\u003eISBS-Conference Proceedings Archive\u003c/em\u003e (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBuckley, C. \u003cem\u003eet al.\u003c/em\u003e Binary classification of running fatigue using a single inertial measurement unit. In 2017 \u003cem\u003eIEEE 14th International Conference on Wearable and Implantable Body Sensor Networks (BSN).\u003c/em\u003e \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttos://\u003c/span\u003e\u003cspan address=\"http://httos://\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003edoi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1109/BSN.2017.7936040\u003c/span\u003e\u003cspan address=\"10.1109/BSN.2017.7936040\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2017)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTomabechi, K., Ikegami, Y., Yamamoto, K. \u0026amp; Nakamura, Y. Learning Whole-body Effects for Biomechanics Analysis from Partial IMU Sensing. In \u003cem\u003e2022 9th IEEE RAS/EMBS International Conference for Biomedical Robotics and Biomechatronics (BioRob).\u003c/em\u003e \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi:10.1109/BioRob52689.2022.9925310\u003c/span\u003e\u003cspan address=\"https://doi:10.1109/BioRob52689.2022.9925310\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMarotta, L., Buurke, J. H., van Beijnum, B. J. F. \u0026amp; Reenalda, J. Towards machine learning-based detection of running-induced fatigue in real-world scenarios: Evaluation of IMU sensor configurations to reduce intrusiveness. Sensors 21, 3451; \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/s21103451\u003c/span\u003e\u003cspan address=\"10.3390/s21103451\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eda Silva Soares, J. \u003cem\u003eet al.\u003c/em\u003e Functional data analysis reveals asymmetrical crank torque during cycling performed at different exercise intensities. J. Biomech, 122, 110478; \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.jbiomech.2021.110478\u003c/span\u003e\u003cspan address=\"10.1016/j.jbiomech.2021.110478\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXu, D. \u003cem\u003eet al.\u003c/em\u003e Explaining the differences of gait patterns between high and low-mileage runners with machine learning. Sci. Rep. 12, 2981 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePatoz, A., Lussiana, T. \u0026amp; Breine, B. Non-South East Asians have a better running economy and different anthropometrics and biomechanics than South East Asians. Sci. Rep. 12, 6291 (2022).\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"},{"header":"Table","content":"\u003cp\u003eTable.1 Overview of the dataset subject to analysis\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"690\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd width=\"4.196816208393632%\"\u003e\n \u003cp\u003e\u003cstrong\u003eID\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"6.9464544138929085%\"\u003e\n \u003cp\u003e\u003cstrong\u003eFreq.\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.104196816208393%\"\u003e\n \u003cp\u003e\u003cstrong\u003eDistance\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e(km)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.709117221418234%\"\u003e\n \u003cp\u003e\u003cstrong\u003eTarget Pace\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e(s/km)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.590448625180898%\"\u003e\n \u003cp\u003e\u003cstrong\u003eLap Time\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003eMean [SD]\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.050651230101302%\"\u003e\n \u003cp\u003e\u003cstrong\u003eFatigue\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003eMedian [min, max]\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"21.852387843704776%\"\u003e\n \u003cp\u003e\u003cstrong\u003eSampling Data\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e(Fatigue No\u003c/strong\u003e\u003cstrong\u003e≦\u003c/strong\u003e\u003cstrong\u003e12, Yes\u003c/strong\u003e\u003cstrong\u003e≧\u003c/strong\u003e\u003cstrong\u003e13)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"20.549927641099856%\"\u003e\n \u003cp\u003e\u003cstrong\u003eChanges in conditions due to practice\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"4.196816208393632%\" rowspan=\"2\" valign=\"top\"\u003e\n \u003cp\u003eA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"6.9464544138929085%\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.104196816208393%\"\u003e\n \u003cp\u003e19\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.709117221418234%\"\u003e\n \u003cp\u003e250\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.590448625180898%\"\u003e\n \u003cp\u003e253[9.3]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.050651230101302%\"\u003e\n \u003cp\u003e12[11,14]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"21.852387843704776%\"\u003e\n \u003cp\u003e390, 180\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"20.549927641099856%\"\u003e\n \u003cp\u003e-\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"7.2507552870090635%\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.459214501510575%\"\u003e\n \u003cp\u003e21.6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.178247734138973%\"\u003e\n \u003cp\u003e250\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"13.141993957703928%\"\u003e\n \u003cp\u003e252[8.3]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.709969788519638%\"\u003e\n \u003cp\u003e13[10,14]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"22.809667673716014%\"\u003e\n \u003cp\u003e270, 360\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"21.45015105740181%\"\u003e\n \u003cp\u003e-\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"4.196816208393632%\" rowspan=\"2\" valign=\"top\"\u003e\n \u003cp\u003eB\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"6.9464544138929085%\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.104196816208393%\"\u003e\n \u003cp\u003e19\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.709117221418234%\"\u003e\n \u003cp\u003e250\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.590448625180898%\"\u003e\n \u003cp\u003e250]5.5]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.050651230101302%\"\u003e\n \u003cp\u003e12[10,13]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"21.852387843704776%\"\u003e\n \u003cp\u003e300, 270\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"20.549927641099856%\"\u003e\n \u003cp\u003euncomfortable feeling\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"7.2507552870090635%\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.459214501510575%\"\u003e\n \u003cp\u003e15\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.178247734138973%\"\u003e\n \u003cp\u003e240\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"13.141993957703928%\"\u003e\n \u003cp\u003e216[5.2]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.709969788519638%\"\u003e\n \u003cp\u003e12[9,13]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"22.809667673716014%\"\u003e\n \u003cp\u003e360, 90\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"21.45015105740181%\"\u003e\n \u003cp\u003eresidual fatigue\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"4.196816208393632%\" rowspan=\"2\" valign=\"top\"\u003e\n \u003cp\u003eC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"6.9464544138929085%\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.104196816208393%\"\u003e\n \u003cp\u003e19\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.709117221418234%\"\u003e\n \u003cp\u003e245\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.590448625180898%\"\u003e\n \u003cp\u003e252[7.4]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.050651230101302%\"\u003e\n \u003cp\u003e13[10,14]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"21.852387843704776%\"\u003e\n \u003cp\u003e270, 300\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"20.549927641099856%\"\u003e\n \u003cp\u003e-\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"7.2507552870090635%\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.459214501510575%\"\u003e\n \u003cp\u003e15\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.178247734138973%\"\u003e\n \u003cp\u003e230\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"13.141993957703928%\"\u003e\n \u003cp\u003e215[4.1]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.709969788519638%\"\u003e\n \u003cp\u003e11[9,13]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"22.809667673716014%\"\u003e\n \u003cp\u003e420, 30\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"21.45015105740181%\"\u003e\n \u003cp\u003eresidual fatigue\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"4.196816208393632%\" rowspan=\"2\" valign=\"top\"\u003e\n \u003cp\u003eD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"6.9464544138929085%\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.104196816208393%\"\u003e\n \u003cp\u003e19\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.709117221418234%\"\u003e\n \u003cp\u003e245\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.590448625180898%\"\u003e\n \u003cp\u003e243[11.1]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.050651230101302%\"\u003e\n \u003cp\u003e11[9,13]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"21.852387843704776%\"\u003e\n \u003cp\u003e420, 150\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"20.549927641099856%\"\u003e\n \u003cp\u003euncomfortable feeling\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"7.2507552870090635%\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.459214501510575%\"\u003e\n \u003cp\u003e15\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.178247734138973%\"\u003e\n \u003cp\u003e225\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"13.141993957703928%\"\u003e\n \u003cp\u003e215[5.6]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.709969788519638%\"\u003e\n \u003cp\u003e11[9,13]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"22.809667673716014%\"\u003e\n \u003cp\u003e330, 120\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"21.45015105740181%\"\u003e\n \u003cp\u003e-\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"4.196816208393632%\" rowspan=\"2\" valign=\"top\"\u003e\n \u003cp\u003eE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"6.9464544138929085%\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.104196816208393%\"\u003e\n \u003cp\u003e19\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.709117221418234%\"\u003e\n \u003cp\u003e250\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.590448625180898%\"\u003e\n \u003cp\u003e249[18.8]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.050651230101302%\"\u003e\n \u003cp\u003e12[11,14]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"21.852387843704776%\"\u003e\n \u003cp\u003e330, 240\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"20.549927641099856%\"\u003e\n \u003cp\u003eresidual fatigue\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"7.2507552870090635%\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.459214501510575%\"\u003e\n \u003cp\u003e21.6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.178247734138973%\"\u003e\n \u003cp\u003e250\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"13.141993957703928%\"\u003e\n \u003cp\u003e250[11.6]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.709969788519638%\"\u003e\n \u003cp\u003e12[11,13]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"22.809667673716014%\"\u003e\n \u003cp\u003e510, 120\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"21.45015105740181%\"\u003e\n \u003cp\u003e-\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"4.196816208393632%\" valign=\"top\"\u003e\n \u003cp\u003eF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"6.9464544138929085%\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.104196816208393%\"\u003e\n \u003cp\u003e16.8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.709117221418234%\"\u003e\n \u003cp\u003e245\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.590448625180898%\"\u003e\n \u003cp\u003e243[9.1]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.050651230101302%\"\u003e\n \u003cp\u003e11[10,13]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"21.852387843704776%\"\u003e\n \u003cp\u003e420, 90\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"20.549927641099856%\"\u003e\n \u003cp\u003euncomfortable feeling\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"4.196816208393632%\" valign=\"top\"\u003e\n \u003cp\u003eG\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"6.9464544138929085%\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.104196816208393%\"\u003e\n \u003cp\u003e16.8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.709117221418234%\"\u003e\n \u003cp\u003e250\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.590448625180898%\"\u003e\n \u003cp\u003e248[6.3]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.050651230101302%\"\u003e\n \u003cp\u003e11[9,13]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"21.852387843704776%\"\u003e\n \u003cp\u003e390, 120\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"20.549927641099856%\"\u003e\n \u003cp\u003euncomfortable feeling\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"4.196816208393632%\" valign=\"top\"\u003e\n \u003cp\u003eH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"6.9464544138929085%\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.104196816208393%\"\u003e\n \u003cp\u003e16.8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.709117221418234%\"\u003e\n \u003cp\u003e250\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.590448625180898%\"\u003e\n \u003cp\u003e243[11.6]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.050651230101302%\"\u003e\n \u003cp\u003e12[11,13]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"21.852387843704776%\"\u003e\n \u003cp\u003e390, 120\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"20.549927641099856%\"\u003e\n \u003cp\u003e-\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"4.196816208393632%\" valign=\"top\"\u003e\n \u003cp\u003eI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"6.9464544138929085%\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.104196816208393%\"\u003e\n \u003cp\u003e21.6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.709117221418234%\"\u003e\n \u003cp\u003e255\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.590448625180898%\"\u003e\n \u003cp\u003e248[12.2]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.050651230101302%\"\u003e\n \u003cp\u003e12[9,15]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"21.852387843704776%\"\u003e\n \u003cp\u003e360, 270\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"20.549927641099856%\"\u003e\n \u003cp\u003euncomfortable feeling, pain\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"4.196816208393632%\" valign=\"top\"\u003e\n \u003cp\u003eJ\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"6.9464544138929085%\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"8.104196816208393%\"\u003e\n \u003cp\u003e19\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"10.709117221418234%\"\u003e\n \u003cp\u003e255\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.590448625180898%\"\u003e\n \u003cp\u003e250[5.5]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"15.050651230101302%\"\u003e\n \u003cp\u003e12[11,13]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"21.852387843704776%\"\u003e\n \u003cp\u003e360, 210\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"20.549927641099856%\"\u003e\n \u003cp\u003e-\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":true,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-3850139/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-3850139/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Running is a widely-accepted activity among the general public, with runners aspiring to achieve optimal performance. However, established methods for the regular monitoring of running forms is lacking. To address this gap, we explore a versatile visualization method utilizing the widely-adopted Inertial Measurement Unit sensor. The running forms of 17-year-old male high school students were monitored during long-distance running training. Acceleration and angular velocity data were collected from a sensor attached to the lumbar region; data from the left foot contact to the next left foot contact were defined as the running cycle. Fatigue during running was assessed using the Borg Scale. The distribution of principal component scores obtained from functional principal component analysis of the running form data corresponded to changes in fatigue from one measurement session to another. However, no consistent trends or changes were observed across subjects. The running forms of participants who were measured twice exhibited a close distribution and similarity, yet unique features were also observed during each measurement. The findings suggest that changes and characteristics of runners' running forms can be readily visualized using a generic approach based on commonly-used sensors and functional data analysis.","manuscriptTitle":"Visualisation of running form changes measured by wearable sensors for conditioning management, an application of the Functional Data Analysis","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-01-18 15:46:33","doi":"10.21203/rs.3.rs-3850139/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"cec3a4e6-90c2-4362-9a63-e9b3a992f952","owner":[],"postedDate":"January 18th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":28202209,"name":"Health sciences/Health care/Disease prevention"},{"id":28202210,"name":"Health sciences/Health care/Patient education"},{"id":28202211,"name":"Biological sciences/Computational biology and bioinformatics/Statistical methods"},{"id":28202212,"name":"Biological sciences/Zoology/Biomechanics"},{"id":28202213,"name":"Biological sciences/Systems biology/Signal processing"},{"id":28202214,"name":"Biological sciences/Systems biology/Time series"},{"id":28202215,"name":"Health sciences/Biomarkers"},{"id":28202216,"name":"Health sciences/Health care"},{"id":28202217,"name":"Health sciences/Signs and symptoms"}],"tags":[],"updatedAt":"2024-03-18T13:02:12+00:00","versionOfRecord":[],"versionCreatedAt":"2024-01-18 15:46:33","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-3850139","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-3850139","identity":"rs-3850139","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}