Tackle and impact detection in elite Australian football using wearable microsensor technology

Cost, time savings and effectiveness of wearable devices for remote monitoring of patient rehabilitation after total knee arthroplasty: study protocol for a randomized controlled trial

BACKGROUND

Total knee arthroplasty (TKA) is a surgical procedure primarily used to treat patients with end-stage knee osteoarthritis (KOA). Postoperative physical exercise is a critical part of the overall treatment of KOA and can bring significant benefits to the patients' recovery. Wearable devices can monitor patients' exercise data and upload it to the physician's workstation. This allows the rehabilitation physician to make timely adjustments based on the patients' movement feedback, and the surgeon can be informed of the patients' functional status. Overall, this study aims to evaluate the effectiveness of using wearable monitoring devices for rehabilitation exercise after TKA, with a focus on cost, time savings, and patient outcomes.

METHOD/DESIGN

This is a single-center, single-blinded, parallel randomized controlled trial conducted at Xi'an Honghui Hospital, a regional orthopedic medical center. Eligible patients will be recruited to participate in the study, and baseline data collection and clinical assessments will be conducted at the time of admission. Using the principle of random allocation, recruited patients will be divided into either the experimental or control group. Both groups will undergo a standard, widely promoted rehabilitation program. The patients in the experimental group will wear equipment to detect and track mobility in the lower limbs. All patients will return to the outpatient clinic for follow-up assessments at 2 weeks, 12 weeks, and 24 weeks after discharge, where outcome indicators will be measured. The primary outcome will be the cost and time after discharge, while secondary outcomes will include the 6-min walk test (6MWT), range of motion (ROM), visual analog scale (VAS), American Knee Society Score (KSS), the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC).

DISCUSSION

We should encourage the adoption of novel, easy-to-use, supervised devices if they prove to be beneficial for patients in terms of cost, time, and effectiveness after TKA. This type of device is particularly important for people in remote rural areas, those with limited financial resources, and those who are reluctant to return to hospitals for follow-up care. Trial registration Chinese Clinical Trial Registry ChiCTR2300068418. Registered on 17 February 2023.

Effects of Physical Exertion on Early Changes in Blood-Based Brain Biomarkers: Implications for the Acute Point of Care Diagnosis of Concussion

Blood-based brain biomarkers (BBM) such as glial fibrillary acidic protein (GFAP) and ubiquitin carboxy-terminal hydrolase L1 (UCH-L1) have potential to aid in the diagnosis of concussion. Recently developed point-of-care test devices would enable BBMs to be measured in field settings such military and sport environments within minutes of a suspicious head hit. However, head hits in these environments typically occur in the setting of vigorous physical exertion, which can itself increase BBMs levels. Thus, efforts to develop BBMs as acute concussion aids in field settings need to account for the effects of physical exertion. To determine the acute effects of physical exertion on the BBMs, we measured GFAP, UCH-L1, tau, and neurofilament light chain (NF-L) immediately before, immediately after, and 45 min after a single workout session consisting of aerobic and resistance exercises in 30 collegiate football players. Subjects wore body sensors measuring several aspects of exertion and underwent diffusion tensor imaging 24 h before and 48 h after exertion. All subjects were male with a mean age of 19.5 ± 1.2 years. The mean duration of activity during the workout session was 94 ± 31 min. There was a significant decrease in serum GFAP immediately after (median decrease of 27.76%, p < 0.0001) and a significant increase in serum UCH-L1 45 min after (median increase of 37.11%, p = 0.016) exertion, compared with pre-exertion baseline. No significant changes in tau or NF-L were identified. The duration of exertion had a significant independent linear correlation to the increase in serum UCHL1 from pre-exertion to 45 min after exertion (r = 0.68, p = 0.004). There were no significant pre- to post-exertional changes in any of the 39 examined brain white matter regions, and biomarker changes did not correlate to variation in white matter integrity in any of these regions. Thus, exertion appeared to be associated with immediate decreases in serum GFAP and very acute (45 min) increases in UCH-L1. These changes were related to the duration of exertion, but not to changes in brain white matter integrity. Our results have important implications for how these BBMs might be used to aid in the on-scene diagnosis of concussion occurring in the setting of physical exertion.

Prototype Machine Learning Algorithms from Wearable Technology to Detect Tennis Stroke and Movement Actions

This study evaluated the accuracy of tennis-specific stroke and movement event detection algorithms from a cervically mounted wearable sensor containing a triaxial accelerometer, gyroscope and magnetometer. Stroke and movement data from up to eight high-performance tennis players were captured in match-play and movement drills. Prototype algorithms classified stroke (i.e., forehand, backhand, serve) and movement (i.e., "Alert", "Dynamic", "Running", "Low Intensity") events. Manual coding evaluated stroke actions in three classes (i.e., forehand, backhand and serve), with additional descriptors of spin (e.g., slice). Movement data was classified according to the specific locomotion performed (e.g., lateral shuffling). The algorithm output for strokes were analysed against manual coding via absolute (n) and relative (%) error rates. Coded movements were grouped according to their frequency within the algorithm's four movement classifications. Highest stroke accuracy was evident for serves (98%), followed by groundstrokes (94%). Backhand slice events showed 74% accuracy, while volleys remained mostly undetected (41-44%). Tennis-specific footwork patterns were predominantly grouped as "Dynamic" (63% of total events), alongside successful linear "Running" classifications (74% of running events). Concurrent stroke and movement data from wearable sensors allows detailed and long-term monitoring of tennis training for coaches and players. Improvements in movement classification sensitivity using tennis-specific language appear warranted.

Integrating wearables and modelling for monitoring rehabilitation following total knee joint replacement

BACKGROUND AND OBJECTIVE

Wearable inertial devices integrated with modelling and cloud computing have been widely adopted in the sports sector, however, their use in the health and medical field has yet to be fully realised. To date, there have been no reported studies concerning the use of wearables as a surrogate tool to monitor knee joint loading during recovery following a total knee joint replacement. The objective of this study is to firstly evaluate if peak tibial acceleration from wearables during gait is a good surrogate metric for computer modelling predicted functional knee loading; and secondly evaluate if traditional clinical patient related outcomes measures are consistent with wearable predictions.

METHODS

Following ethical approval, four healthy participants were used to establish the relationship between computer modelling predicted knee joint loading and wearable measured tibial acceleration. Following this, ten patients who had total knee joint replacements were then followed during their 6-week rehabilitation. Gait analysis, wearable acceleration, computer models of knee joint loading, and patient related outcomes measures including the Oxford knee score and range of motion were recorded.

RESULTS

A linear correlation (R² of 0.7-0.97) was observed between peak tibial acceleration (from wearables) and musculoskeletal model predicted knee joint loading during gait in healthy participants first. Whilst patient related outcome measures (Oxford knee score and patient range of motion) were observed to improve consistently during rehabilitation, this was not consistent with all patient's tibial acceleration. Only those patients that exhibited increasing peak tibial acceleration over 6-weeks rehabilitation were positively correlated with the Oxford knee score (R² of 0.51 to 0.97). Wearable predicted tibial acceleration revealed three patients with a consistent knee loading, five patients with improving knee loading, and two patients with declining knee loading during recovery. Hence, 20% of patients did not present with satisfactory joint loading following total knee joint replacement and this was not detected with current patient related outcome measures.

CONCLUSIONS

The use of inertial measurement units or wearables in this study provided additional insight into patients who were not exhibiting functional improvements in joint loading, and offers clinicians an 'off-site' early warning metric to identify potential complications during recovery and provide the opportunity for early intervention. This study has important implications for improving patient outcomes, equity, and for those who live in rural regions.

Tackle Technique and Changes in Playerload™ During a Simulated Tackle: An Exploratory Study

In collision sports, the tackle has the highest injury incidence, and is key to a successful performance. Although the contact load of players has been measured using microtechnology, this has not been related to tackle technique. The aim of this study was to explore how PlayerLoad™ changes between different levels of tackling technique during a simulated tackle. Nineteen rugby union players performed twelve tackles on a tackle contact simulator (n = 228 tackles). Each tackle was recorded with a video-camera and each player wore a Catapult OptimEyeS5. Tackles were analysed using tackler proficiency criteria and split into three categories: Low scoring(≤5 Arbitrary units (AU), medium scoring(6 and 7AU) and high scoring tackles(≥8AU). High scoring tackles recorded a higher PlayerLoad™ at tackle completion. The PlayerLoad™ trace was also less variable in the high scoring tackles. The variability in the PlayerLoad™ trace may be a consequence of players not shortening their steps before contact. This reduced their ability to control their movement during the contact and post-contact phase of the tackle and increased the variability. Using the PlayerLoad™ trace in conjunction with subjective technique assessments offers coaches and practitioners insight into the physical-technical relationship of each tackle to optimise tackle skill training and match preparation.

The Use of Wearable Sensors for Preventing, Assessing, and Informing Recovery from Sport-Related Musculoskeletal Injuries: A Systematic Scoping Review

Wearable technologies are often indicated as tools that can enable the in-field collection of quantitative biomechanical data, unobtrusively, for extended periods of time, and with few spatial limitations. Despite many claims about their potential for impact in the area of injury prevention and management, there seems to be little attention to grounding this potential in biomechanical research linking quantities from wearables to musculoskeletal injuries, and to assessing the readiness of these biomechanical approaches for being implemented in real practice. We performed a systematic scoping review to characterise and critically analyse the state of the art of research using wearable technologies to study musculoskeletal injuries in sport from a biomechanical perspective. A total of 4952 articles were retrieved from the Web of Science, Scopus, and PubMed databases; 165 were included. Multiple study features-such as research design, scope, experimental settings, and applied context-were summarised and assessed. We also proposed an injury-research readiness classification tool to gauge the maturity of biomechanical approaches using wearables. Five main conclusions emerged from this review, which we used as a springboard to propose guidelines and good practices for future research and dissemination in the field.

Validating an algorithm from a trunk-mounted wearable sensor for detecting stroke events in tennis

This study analysed the accuracy of a prototype algorithm for tennis stroke detection from wearable technology. Strokes from junior-elite tennis players over 10 matches were analysed. Players wore a GPS unit containing an accelerometer, gyroscope and magnetometer. Manufacturer-developed algorithms determined stoke type and count (forehands, backhands, serves and other). Matches were video recorded to manually code ball contacts and shadow swing events for forehands, backhands and serves and further by stroke classifications (i.e., drive, volley, slice, end-range). Comparisons between algorithm and coding were analysed via ANOVA and Bland-Altman plots at the match-level and error rates for specific stroke-types. No significant differences existed for stroke count between the algorithm and manual coding (p > 0.05). Significant (p < 0.0001) overestimation of "Other" strokes were observed from the algorithm, with no difference in groundstrokes and serves (p > 0.05). Serves had the highest accuracy of all stroke types (≥98%). Forehand and backhand "drives" were the most accurate (>86%), with volleys mostly undetected (58-60%) and slices and end-range strokes likely misclassified (49-51%). The prototype algorithm accurately quantifies serves and forehand and backhand "drives" and serves. However, underestimations of shadow swings and overestimations of "other" strokes suggests strokes with reduced trunk rotation have poorer detection accuracy.

Predicting Temporal Gait Kinematics From Running Velocity

ABSTRACT

Gray, A, Price, M, and Jenkins, D. Predicting temporal gait kinematics from running velocity. J Strength Cond Res 35(9): 2379-2382, 2021-The manner in which stride frequency (f) changes in response to running velocity (v) is well established. Notably, as running velocity increases, duty factor (d, the % of the stride in stance) decreases, concomitantly with higher stride frequencies. Mathematical descriptions of this relationship do not exist, limiting our ability to reasonably predict gait-based metrics from wearable technologies. Therefore, the purpose of this study was to establish prediction equations for stride frequency and duty factor from running velocity. On 2 occasions, 10 healthy men (aged, 21.1 ± 2.2 years) performed constant pace running efforts at 3, 4, 5, 6, 7, and 8 m·s-1 over a 10-m segment on a tartan athletics track. Running efforts were filmed using a digital video camera at 300 frames per second, from which stride duration, support duration, and swing duration were determined. Regression equations to predict stride frequency and duty factor from running velocity were established by curve fitting. Acceptable test-retest reliability for the video-based determination of stride frequency (intraclass correlation = 0.87; typical error of the measurement [TEM] = 0.01 Hz; coefficient of variation [CV] = 2.9%) and duty factor (r = 0.93; TEM = 1%; CV = 3.9%) were established. The relationship between stride frequency and running velocity was described by the following quadratic equation: f = 0.026·v2 - 0.111·v + 1.398 (r2 = 0.903). The relationship between duty factor and running velocity was described by the quadratic equation d = 0.004·v2 - 0.061·v + 0.50 (r2 = 0.652). The relationships between v and f and between v and d are consistent with previous observations. These equations may contribute broader locomotor models or serve as input variables in data fusion algorithms that enhance outputs from wearable technologies.

Remote Patient Monitoring with Wearable Sensors Following Knee Arthroplasty

(Background) Inertial Measurement Units (IMUs) provide a low-cost, portable solution to obtain functional measures similar to those captured with three-dimensional gait analysis, including spatiotemporal gait characteristics. The primary aim of this study was to determine the feasibility of a remote patient monitoring (RPM) workflow using ankle-worn IMUs measuring impact load, limb impact load asymmetry and knee range of motion in combination with patient-reported outcome measures. (Methods) A pilot cohort of 14 patients undergoing primary knee arthroplasty for osteoarthritis was prospectively enrolled. RPM in the community was performed weekly from 2 up to 6 weeks post-operatively using wearable IMUs. The following data were collected using IMUs: mobility (Bone Stimulus and cumulative impact load), impact load asymmetry and maximum knee flexion angle. In addition, scores from the Oxford Knee Score (OKS), EuroQol Five-dimension (EQ-5D) with EuroQol visual analogue scale (EQ-VAS) and 6 Minute Walk Test were collected. (Results) On average, the Bone Stimulus and cumulative impact load improved 52% (p = 0.002) and 371% (p = 0.035), compared to Post-Op Week 2. The impact load asymmetry value trended (p = 0.372) towards equal impact loading between the operative and non-operative limb. The mean maximum flexion angle achieved was 99.25° at Post-Operative Week 6, but this was not significantly different from pre-operative measurements (p = 0.1563). There were significant improvements in the mean EQ-5D (0.20; p = 0.047) and OKS (10.86; p < 0.001) scores both by 6 weeks after surgery, compared to pre-operative scores. (Conclusions) This pilot study demonstrates the feasibility of a reliable and low-maintenance workflow system to remotely monitor post-operative progress in knee arthroplasty patients. Preliminary data indicate IMU outputs relating to mobility, impact load asymmetry and range of motion can be obtained using commercially available IMU sensors. Further studies are required to directly correlate the IMU sensor outputs with patient outcomes to establish clinical significance.

Does Site Matter? Impact of Inertial Measurement Unit Placement on the Validity and Reliability of Stride Variables During Running: A Systematic Review and Meta-analysis

BACKGROUND

Inertial measurement units (IMUs) are used for running gait analysis in a variety of sports. These sensors have been attached at various locations to capture stride data. However, it is unclear if different placement sites affect the derived outcome measures.

OBJECTIVE

The aim of this systematic review and meta-analysis was to investigate the impact of placement on the validity and reliability of IMU-derived measures of running gait.

METHODS

Online databases SPORTDiscus with Full Text, CINAHL Complete, MEDLINE (EBSCOhost), EMBASE (Ovid) and Scopus were searched from the earliest record to 6 August 2020. Articles were included if they (1) used an IMU during running (2) reported spatiotemporal variables, peak ground reaction force (GRF) or vertical stiffness and (3) assessed validity or reliability. Meta-analyses were performed for a pooled validity estimate when (1) studies reported means and standard deviation for variables derived from the IMU and criterion (2) used the same IMU placement and (3) determined validity at a comparable running velocity (≤ 1 m·s^-1 difference).

RESULTS

Thirty-nine articles were included, where placement varied between the foot, tibia, hip, sacrum, lumbar spine (LS), torso and thoracic spine (TS). Initial contact, toe-off, contact time (CT), flight time (FT), step time, stride time, swing time, step frequency (SF), step length (SL), stride length, peak vertical and resultant GRF and vertical stiffness were analysed. Four variables (CT, FT, SF and SL) were meta-analysed, where CT was compared between the foot, tibia and LS placements and SF was compared between foot and LS. Foot placement data were meta-analysed for FT and SL. All data are the mean difference (MD [95%CI]). No significant difference was observed for any site compared to the criterion for CT (foot: - 11.47 ms [- 45.68, 22.74], p = 0.43; tibia: 22.34 ms [- 18.59, 63.27], p = 0.18; LS: - 48.74 ms [- 120.33, 22.85], p = 0.12), FT (foot: 11.93 ms [- 8.88, 32.74], p = 0.13), SF (foot: 0.45 step·min^-1 [- 1.75, 2.66], p = 0.47; LS: - 3.45 step·min^-1 [- 16.28, 9.39], p = 0.37) and SL (foot: 0.21 cm [- 1.76, 2.18], p = 0.69). Reliable derivations of CT (coefficient of variation [CV] < 9.9%), FT (CV < 11.6%) and SF (CV < 4.4%) were shown using foot- and LS-worn IMUs, while the CV was < 7.8% for foot-determined stride time, SL and stride length. Vertical GRF was reliable from the LS (CV = 4.2%) and TS (CV = 3.3%) using a spring-mass model, while vertical stiffness was moderately (r = 0.66) and nearly perfectly (r = 0.98) correlated with criterion measures from the TS.

CONCLUSION

Placement of IMUs on the foot, tibia and LS is suitable to derive valid and reliable stride data, suggesting measurement site may not be a critical factor. However, evidence regarding the ability to accurately detect stride events from the TS is unclear and this warrants further investigation.

Wearable Technology and Analytics as a Complementary Toolkit to Optimize Workload and to Reduce Injury Burden

Wearable sensors enable the real-time and non-invasive monitoring of biomechanical, physiological, or biochemical parameters pertinent to the performance of athletes. Sports medicine researchers compile datasets involving a multitude of parameters that can often be time consuming to analyze in order to create value in an expeditious and accurate manner. Machine learning and artificial intelligence models may aid in the clinical decision-making process for sports scientists, team physicians, and athletic trainers in translating the data acquired from wearable sensors to accurately and efficiently make decisions regarding the health, safety, and performance of athletes. This narrative review discusses the application of commercial sensors utilized by sports teams today and the emergence of descriptive analytics to monitor the internal and external workload, hydration status, sleep, cardiovascular health, and return-to-sport status of athletes. This review is written for those who are interested in the application of wearable sensor data and data science to enhance performance and reduce injury burden in athletes of all ages.

The Validity and Reliability of Wearable Microtechnology for Intermittent Team Sports: A Systematic Review

BACKGROUND

Technology has long been used to track player movements in team sports, with initial tracking via manual coding of video footage. Since then, wearable microtechnology in the form of global and local positioning systems has provided a less labour-intensive way of monitoring movements. As such, there has been a proliferation in research pertaining to these devices.

OBJECTIVE

A systematic review of studies that investigate the validity and/or reliability of wearable microtechnology to quantify movement and specific actions common to intermittent team sports.

METHODS

A systematic search of CINAHL, MEDLINE, and SPORTDiscus was performed; studies included must have been (1) original research investigations; (2) full-text articles written in English; (3) published in a peer-reviewed academic journal; and (4) assessed the validity and/or reliability of wearable microtechnology to quantify movements or specific actions common to intermittent team sports.

RESULTS

A total of 384 studies were retrieved and 187 were duplicates. The titles and abstracts of 197 studies were screened and the full texts of 88 manuscripts were assessed. A total of 62 studies met the inclusion criteria. Additional 10 studies, identified via reference list assessment, were included. Therefore, a total of 72 studies were included in this review.

CONCLUSION

There are many studies investigating the validity and reliability of wearable microtechnology to track movement and detect sport-specific actions. It is evident that for the majority of metrics, validity and reliability are multi-factorial, in that it is dependent upon a wide variety of factors including wearable technology brand and model, sampling rate, type of movement performed (e.g., straight line, change of direction) and intensity of movement (e.g., walk, sprint). Practitioners should be mindful of the accuracy and repeatability of the devices they are using when making decisions on player training loads.

The use of GPS and inertial devices for player monitoring in team sports: A review of current and future applications

Player-worn devices, combining global positioning system and inertial monitors, are being used increasingly by professional sports teams. Recent interest focusses on using the data generated to track trainingload and whether this may lead to more effective training prescription with better management of injury risk. The aim of this review is to summarize the development and current use of this technology alongside proposed future applications. PubMed and Medline searches (2000-2017) identified all relevant studies involving use in team sports or comparative studies with other accepted methods. Our review determined that the latest devices are valid and reliably track activity levels. This technology is both accurate and more efficient than previous methods. Furthermore, recent research has shown that measurable changes in trainingload (the acute-to-chronic load ratio) are related to injury risk. However, results remain very sport specific and generalization must be done with caution. Future uses may include injury-prevention strategies and return-to-play judgement.

Deceleration, Acceleration, and Impacts Are Strong Contributors to Muscle Damage in Professional Australian Football

Gastin, PB, Hunkin, SL, Fahrner, B, and Robertson, S. Deceleration, acceleration, and impacts are strong contributors to muscle damage in professional Australian football. J Strength Cond Res 33(12): 3374-3383, 2019-The purpose of this study was to investigate the relationships between serum creatine kinase [CK], an indirect marker of muscle damage, and specific indices of match load in elite Australian football. Twenty-six professional players were assessed during a competitive Australian Football League (AFL) season. [CK] was collected 24-36 hours before match and 34-40 hours after match during 8 in-season rounds. An athlete-tracking technology was used to quantify match load. Generalized estimating equations and random forest models were constructed to determine the extent to which match-load indices and pre-match [CK] explained post-match [CK]. There was a 129 ± 152% increase in [CK] in response to AFL competition. Generalized estimating equations found that number of impacts >3g (p = 0.004) and game time (p = 0.016) were most strongly associated with post-match [CK]. Random forest, with considerably lower errors (130 vs. 316 U·L), found deceleration, acceleration, impacts >3g, and sprint distance to be the strongest predictors. Pre-match [CK] accounted for 11% of post-match [CK], and considerable interindividual and intraindividual variability existed in the data. Creatine kinase, an indicator of muscle damage, was considerably elevated as a result of AFL competition. Parametric and machine-learning analysis techniques found several indices of physical load associated with muscle damage during competition, with impacts >3g and high-intensity running variables as the strongest predictors. [CK] may be used as a global measure of muscle damage in field team sports such as AF, yet with some caution given cost, invasiveness, and inherent variability. Quantifying physical load and the responses to that load can guide athlete management decision-making and is best undertaken within a suite of practical, sport-specific measures, where data are interpreted individually and with an understanding of the limitations.

Development of a Human Activity Recognition System for Ballet Tasks

Background

Accurate and detailed measurement of a dancer’s training volume is a key requirement to understanding the relationship between a dancer’s pain and training volume. Currently, no system capable of quantifying a dancer’s training volume, with respect to specific movement activities, exists. The application of machine learning models to wearable sensor data for human activity recognition in sport has previously been applied to cricket, tennis and rugby. Thus, the purpose of this study was to develop a human activity recognition system using wearable sensor data to accurately identify key ballet movements (jumping and lifting the leg). Our primary objective was to determine if machine learning can accurately identify key ballet movements during dance training. The secondary objective was to determine the influence of the location and number of sensors on accuracy.

Results

Convolutional neural networks were applied to develop two models for every combination of six sensors (6, 5, 4, 3, etc.) with and without the inclusion of transition movements. At the first level of classification, including data from all sensors, without transitions, the model performed with 97.8% accuracy. The degree of accuracy reduced at the second (83.0%) and third (75.1%) levels of classification. The degree of accuracy reduced with inclusion of transitions, reduction in the number of sensors and various sensor combinations.

Conclusion

The models developed were robust enough to identify jumping and leg lifting tasks in real-world exposures in dancers. The system provides a novel method for measuring dancer training volume through quantification of specific movement tasks. Such a system can be used to further understand the relationship between dancers’ pain and training volume and for athlete monitoring systems. Further, this provides a proof of concept which can be easily translated to other lower limb dominant sporting activities

Wearables for Integrative Performance and Tactic Analyses: Opportunities, Challenges, and Future Directions

Micro-electromechanical systems (MEMS) have reduced drastically in size, cost, and power consumption, while improving accuracy. The combination of different sensor technologies is considered a promising step in the monitoring of athletes. Those "wearables" enable the capturing of relevant physiological and tactical information in individual and team sports and thus replacing subjective, time-consuming and qualitative methods with objective, quantitative ones. Prior studies mainly comprised sports categories such as: targeting sports, batting and fielding games as well as net and wall games, focusing on the detection of individual, non-locomotive movements. The increasing capabilities of wearables allow for more complex and integrative analysis expanding research into the last category: invasion sports. Such holistic approaches allow the derivation of metrics, estimation of physical conditions and the analysis of team strategic behavior, accompanied by integrative knowledge gains in technical, tactical, physical, and mental aspects of a sport. However, prior and current researchers find the precise measurement of the actual movement within highly dynamic and non-linear movement difficult. Thus, the present article showcases an overview of the environments in which the wearables are employed. It elaborates their use in individual as well as team-related performance analyses with a special focus on reliability and validity, challenges, and future directions.

Use of Machine Learning and Wearable Sensors to Predict Energetics and Kinematics of Cutting Maneuvers

Changes of directions and cutting maneuvers, including 180-degree turns, are common locomotor actions in team sports, implying high mechanical load. While the mechanics and neurophysiology of turns have been extensively studied in laboratory conditions, modern inertial measurement units allow us to monitor athletes directly on the field. In this study, we applied four supervised machine learning techniques (linear regression, support vector regression/machine, boosted decision trees and artificial neural networks) to predict turn direction, speed (before/after turn) and the related positive/negative mechanical work. Reference values were computed using an optical motion capture system. We collected data from 13 elite female soccer players performing a shuttle run test, wearing a six-axes inertial sensor at the pelvis level. A set of 18 features (predictors) were obtained from accelerometers, gyroscopes and barometer readings. Turn direction classification returned good results (accuracy > 98.4%) with all methods. Support vector regression and neural networks obtained the best performance in the estimation of positive/negative mechanical work (coefficient of determination R² = 0.42-0.43, mean absolute error = 1.14-1.41 J) and running speed before/after the turns (R² = 0.66-0.69, mean absolute error = 0.15-018 m/s). Although models can be extended to different angles, we showed that meaningful information on turn kinematics and energetics can be obtained from inertial units with a data-driven approach.

Video corroboration of player incurred impacts using trunk worn sensors among national ice-hockey team members

PURPOSE

Video corroboration of player incurred impacts (PII) using trunk-worn wearable sensors (WS) among national ice-hockey team members.

METHODS

23 members of the U.S. National (NTDP) U18 team consented to procedures approved by EMU Human Subjects Committee. Bioharness-3 (Zephyr, MD) WS recorded occurrences of PII during games and impacts were generated using Impact Processor (Zephyr, MD). Eight players with the top activity levels each game determined by WS, were observed using video and synchronized with game video collected by NTDP staff. Impacts identified by WS of 6-7.9 g (Z3), 8-9.9 g (Z4) and 10+ g (Z5) were used to corroborate PII. Magnitude and duration of each identified impact were compared by category using MANOVA with Tukey post hoc (α = 0.05; SPSS 22.0, IBM, NY).

RESULTS

Of 419 on-ice impacts, 358 were confirmed true PII (85.5%), 60 as other non-PII (14.3%) and 1 false positive (0.2%). For 358 PII, 17 (4.1%) were 1) Board contact/no check, 74 (17.7%), 2) Board contact/check, 202 (48.2%), 3) Open ice check, 65 (15.5%), 4) Player fall. Of 60 Non-PII, 19 (4.5%) as 5) other form of player to player event, 16 (3.8%) as 6) Hard Stop, 19 (4.5%) as 7) Slapshots and 6 (1.4%) as 8) other identifiable player events. 160 of the 200 Z3 events were PII (80%), 103 of 110 Z4 events (93.6%) and 95 of 109 Z5 events were PII (87.2%). The magnitude of impacts was not different by category, but the duration of category 6 (Hard stop; .058 s) was lower than categories 2, 4 and 7 (.112, .112, .133 s, respectively, p < .05).

CONCLUSION

These data show that using some limited criteria (e.g. impact magnitude and duration), PII can be identified with relatively high accuracy in ice hockey using trunk-worn wearable sensors.

Training and Competition Activity Profiles of Australian Football Field Umpires

Jessiman, SW, Harvey, B, Corrigan, SL, and Gastin, PB. Training and competition activity profiles of Australian football field umpires. J Strength Cond Res 34(10): 2956-2964, 2020-The purpose of this study was to determine the activity profiles of Australian football (AF) field umpires during training and competition, and subsequently assess the specificity to competition of locomotor training sessions. Microtechnology incorporating a 5-Hz (interpolated to 15 Hz) global positioning system sensor tracked the movements of 24 field umpires during matches at the Melbourne Cricket Ground and fitness and skill-based training sessions before competition. Paired t-tests or Wilcoxon signed-rank tests determined whether significant differences existed between single training session and competition paired samples, with Cohen's d effect size and percent differences describing the magnitude of the training-competition differences. Absolute measures of total (d = 5.4; percent difference = 85.8%) and high-speed distance (>14.4 km·h) (1.0; 36.9%), as well as accelerations (3.3; 106.3%) and decelerations (3.2; 107.5%) were significantly greater during competition compared with training (p < 0.001). When standardized for time, high-speed distance (1.4; 52.0%), sprint distance (>23.0 km·h) (0.9; 121.5%), sprint efforts (1.0; 107.4%), and high acceleration (1.1; 114.3%) and deceleration (0.6; 66.7%) events (≥3 m·s) were greater during training (p < 0.001). No difference between training and competition was observed for relative distance. A single training session did not match the volume of work during competition, due primarily to large differences in duration. By contrast, training sessions were higher in intensity, likely to compensate for the shorter duration of training. Further research is required to determine whether the total weekly training load is sufficient to maintain and develop the competition-specific fitness of AF field umpires.

Assessment of a Novel Algorithm to Determine Change-of-Direction Angles While Running Using Inertial Sensors

Balloch, AS, Meghji, M, Newton, RU, Hart, NH, Weber, JA, Ahmad, I, and Habibi, D. Assessment of a novel algorithm to determine change-of-direction angles while running using inertial sensors. J Strength Cond Res 34(1): 134-144, 2020-The ability to detect and quantify change-of-direction (COD) movement may offer a unique approach to load-monitoring practice. Validity and reliability of a novel algorithm to calculate COD angles for predetermined COD movements ranging from 45 to 180° in left and right directions was assessed. Five recreationally active men (age: 29.0 ± 0.5 years; height: 181.0 ± 5.6 cm; and body mass: 79.4 ± 5.3 kg) ran 5 consecutive predetermined COD trials each, at 4 different angles (45, 90, 135, and 180°), in each direction. Participants were fitted with a commercially available microtechnology unit where inertial sensor data were extracted and processed using a novel algorithm designed to calculate precise COD angles for direct comparison with a high-speed video (remotely piloted, position-locked aircraft) criterion measure. Validity was assessed using Bland-Altman 95% limits of agreement and mean bias. Reliability was assessed using typical error (expressed as a coefficient of variation [CV]). Concurrent validity was present for most angles. Left: (45° = 43.8 ± 2.0°; 90° = 88.1 ± 2.0°; 135° = 136.3 ± 2.1°; and 180° = 181.8 ± 2.5°) and Right: (45° = 46.3 ± 1.6°; 90° = 91.9 ± 2.2°; 135° = 133.4 ± 2.0°; 180° = 179.2 ± 5.9°). All angles displayed excellent reliability (CV < 5%) while greater mean bias (3.6 ± 5.1°, p < 0.001), weaker limits of agreement, and reduced precision were evident for 180° trials when compared with all other angles. High-level accuracy and reliability when detecting COD angles further advocates the use of inertial sensors to quantify sports-specific movement patterns.

Validity of a Microsensor-Based Algorithm for Detecting Scrum Events in Rugby Union

PURPOSE

Commercially available microtechnology devices containing accelerometers, gyroscopes, magnetometers, and global positioning technology have been widely used to quantify the demands of rugby union. This study investigated whether data derived from wearable microsensors can be used to develop an algorithm that automatically detects scrum events in rugby union training and match play.

METHODS

Data were collected from 30 elite rugby players wearing a Catapult OptimEye S5 (Catapult Sports, Melbourne, Australia) microtechnology device during a series of competitive matches (n = 46) and training sessions (n = 51). A total of 97 files were required to "train" an algorithm to automatically detect scrum events using random forest machine learning. A further 310 files from training (n = 167) and match-play (n = 143) sessions were used to validate the algorithm's performance.

RESULTS

Across all positions (front row, second row, and back row), the algorithm demonstrated good sensitivity (91%) and specificity (91%) for training and match-play events when the confidence level of the random forest was set to 50%. Generally, the algorithm had better accuracy for match-play events (93.6%) than for training events (87.6%).

CONCLUSIONS

The scrum algorithm was able to accurately detect scrum events for front-row, second-row, and back-row positions. However, for optimal results, practitioners are advised to use the recommended confidence level for each position to limit false positives. Scrum algorithm detection was better with scrums involving ≥5 players and is therefore unlikely to be suitable for scrums involving 3 players (eg, rugby sevens). Additional contact- and collision-detection algorithms are required to fully quantify rugby union demands.

Automatic detection of one-on-one tackles and ruck events using microtechnology in rugby union

OBJECTIVES

To automate the detection of ruck and tackle events in rugby union using a specifically-designed algorithm based on microsensor data.

DESIGN

Cross-sectional study.

METHODS

Elite rugby union players wore microtechnology devices (Catapult, S5) during match-play. Ruck (n=125) and tackle (n=125) event data was synchronised with video footage compiled from international rugby union match-play ruck and tackle events. A specifically-designed algorithm to detect ruck and tackle events was developed using a random forest classification model. This algorithm was then validated using 8 additional international match-play datasets and video footage, with each ruck and tackle manually coded and verified if the event was correctly identified by the algorithm.

RESULTS

The classification algorithm's results indicated that all rucks and tackles were correctly identified during match-play when 79.4±9.2% and 81.0±9.3% of the random forest decision trees agreed with the video-based determination of these events. Sub-group analyses of backs and forwards yielded similar optimal confidence percentages of 79.7% and 79.1% respectively for rucks. Sub-analysis revealed backs (85.3±7.2%) produced a higher algorithm cut-off for tackles than forwards (77.7±12.2%).

CONCLUSIONS

The specifically-designed algorithm was able to detect rucks and tackles for all positions involved. For optimal results, it is recommended that practitioners use the recommended cut-off (80%) to limit false positives for match-play and training. Although this algorithm provides an improved insight into the number and type of collisions in which rugby players engage, this algorithm does not provide impact forces of these events.

Modelling Movement Energetics Using Global Positioning System Devices in Contact Team Sports: Limitations and Solutions

Quantifying the training and competition loads of players in contact team sports can be performed in a variety of ways, including kinematic, perceptual, heart rate or biochemical monitoring methods. Whilst these approaches provide data relevant for team sports practitioners and athletes, their application to a contact team sport setting can sometimes be challenging or illogical. Furthermore, these methods can generate large fragmented datasets, do not provide a single global measure of training load and cannot adequately quantify all key elements of performance in contact team sports. A previous attempt to address these limitations via the estimation of metabolic energy demand (global energy measurement) has been criticised for its inability to fully quantify the energetic costs of team sports, particularly during collisions. This is despite the seemingly unintentional misapplication of the model's principles to settings outside of its intended use. There are other hindrances to the application of such models, which are discussed herein, such as the data-handling procedures of Global Position System manufacturers and the unrealistic expectations of end users. Nevertheless, we propose an alternative energetic approach, based on Global Positioning System-derived data, to improve the assessment of mechanical load in contact team sports. We present a framework for the estimation of mechanical work performed during locomotor and contact events with the capacity to globally quantify the work done during training and matches.

Machine and deep learning for sport-specific movement recognition: a systematic review of model development and performance

Objective assessment of an athlete's performance is of importance in elite sports to facilitate detailed analysis. The implementation of automated detection and recognition of sport-specific movements overcomes the limitations associated with manual performance analysis methods. The object of this study was to systematically review the literature on machine and deep learning for sport-specific movement recognition using inertial measurement unit (IMU) and, or computer vision data inputs. A search of multiple databases was undertaken. Included studies must have investigated a sport-specific movement and analysed via machine or deep learning methods for model development. A total of 52 studies met the inclusion and exclusion criteria. Data pre-processing, processing, model development and evaluation methods varied across the studies. Model development for movement recognition were predominantly undertaken using supervised classification approaches. A kernel form of the Support Vector Machine algorithm was used in 53% of IMU and 50% of vision-based studies. Twelve studies used a deep learning method as a form of Convolutional Neural Network algorithm and one study also adopted a Long Short Term Memory architecture in their model. The adaptation of experimental set-up, data pre-processing, and model development methods are best considered in relation to the characteristics of the targeted sports movement(s).

Reliability of Wearable Inertial Measurement Units to Measure Physical Activity in Team Handball

PURPOSE

To assess the reliability and sensitivity of commercially available inertial measurement units to measure physical activity in team handball.

METHOD

Twenty-two handball players were instrumented with 2 inertial measurement units (OptimEye S5; Catapult Sports, Melbourne, Australia) taped together. They participated in either a laboratory assessment (n = 10) consisting of 7 team handball-specific tasks or field assessment (n = 12) conducted in 12 training sessions. Variables, including PlayerLoad™ and inertial movement analysis (IMA) magnitude and counts, were extracted from the manufacturers' software. IMA counts were divided into intensity bands of low (1.5-2.5 m·s^-1), medium (2.5-3.5 m·s^-1), high (>3.5 m·s^-1), medium/high (>2.5 m·s^-1), and total (>1.5 m·s^-1). Reliability between devices and sensitivity was established using coefficient of variation (CV) and smallest worthwhile difference (SWD).

RESULTS

Laboratory assessment: IMA magnitude showed a good reliability (CV = 3.1%) in well-controlled tasks. CV increased (4.4-6.7%) in more-complex tasks. Field assessment: Total IMA counts (CV = 1.8% and SWD = 2.5%), PlayerLoad (CV = 0.9% and SWD = 2.1%), and their associated variables (CV = 0.4-1.7%) showed a good reliability, well below the SWD. However, the CV of IMA increased when categorized into intensity bands (2.9-5.6%).

CONCLUSION

The reliability of IMA counts was good when data were displayed as total, high, or medium/high counts. A good reliability for PlayerLoad and associated variables was evident. The CV of the previously mentioned variables was well below the SWD, suggesting that OptimEye's inertial measurement unit and its software are sensitive for use in team handball.

Trends Supporting the In-Field Use of Wearable Inertial Sensors for Sport Performance Evaluation: A Systematic Review

Recent technological developments have led to the production of inexpensive, non-invasive, miniature magneto-inertial sensors, ideal for obtaining sport performance measures during training or competition. This systematic review evaluates current evidence and the future potential of their use in sport performance evaluation. Articles published in English (April 2017) were searched in Web-of-Science, Scopus, Pubmed, and Sport-Discus databases. A keyword search of titles, abstracts and keywords which included studies using accelerometers, gyroscopes and/or magnetometers to analyse sport motor-tasks performed by athletes (excluding risk of injury, physical activity, and energy expenditure) resulted in 2040 papers. Papers and reference list screening led to the selection of 286 studies and 23 reviews. Information on sport, motor-tasks, participants, device characteristics, sensor position and fixing, experimental setting and performance indicators was extracted. The selected papers dealt with motor capacity assessment (51 papers), technique analysis (163), activity classification (19), and physical demands assessment (61). Focus was placed mainly on elite and sub-elite athletes (59%) performing their sport in-field during training (62%) and competition (7%). Measuring movement outdoors created opportunities in winter sports (8%), water sports (16%), team sports (25%), and other outdoor activities (27%). Indications on the reliability of sensor-based performance indicators are provided, together with critical considerations and future trends.

Application of Global Positioning System and Microsensor Technology in Competitive Rugby League Match-Play: A Systematic Review and Meta-analysis

BACKGROUND

The use of global positioning system (GPS) devices with the inclusion of microsenor technology in rugby league enables measurement of player activity profiles to understand the demands of match-play and optimise on-field performance.

OBJECTIVE

The aim of this review was to systematically review the use of GPS and microsensor technology to quantify player activity profiles in match-play, and conduct a meta-analysis of relevant movement variables in order to present the contemporary and emerging themes within rugby league.

METHODS

A systematic search of electronic databases (MEDLINE, SPORTDiscus, CINAHL, Web of Science, Scopus, ScienceDirect, EMBASE, and Google Scholar) from the earliest record to February 2015 was conducted. Permutations of key words included GPS, microtechnology, activity profiles, match demands (movement or physical demands), and rugby league. A meta-analysis was conducted to provide a pooled mean and confidence intervals on comparable data from at least three studies.

RESULTS

Twenty-seven studies met the eligibility criteria and included 1270 male participants. The studies reported on GPS use in elite competition (n = 16) with limited representation of other competition standards: sub-elite (n = 6), amateur (n = 1) and junior (n = 3). All studies reported on movement variables (distance, relative distance, speed and accelerations), with studies analysing movement at high speed (n = 18, 66.7%), evaluating collision or impact variables (n = 15, 55.6%) and determining the metabolic energy (n = 2, 7.4%) associated with rugby league match-play. Activity profiles of varying positions, positional groups and levels of rugby league competition were described. Meta-analysis indicated that the total distance covered by backs and adjustables were both greater than forward positions, but adjustables covered greater relative distance than forwards and backs. Speed zones were typically categorised into six speed zones ranging from 0 to 36 km·h(-1), or into low- and high-intensity movement. Vast inconsistencies were apparent across studies in categorising movement at high speed, posing difficulties for comparison. Meta-analysis indicated that, although the number of repeated high-intensity effort (RHIE) bouts in elite players were similar to sub-elite (and both greater than juniors), the number of efforts per RHIE were significantly greater in elite than sub-elite players. Differential pacing strategies were adopted according to player selection (whole-match vs. interchange), time period within match-play and match outcome, in order to maintain high-intensity performance or to challenge for a win. Sizeable inconsistencies were also identified in the definitions of reported collisions (classified as mild, moderate and heavy) and impacts (six zone categories provided by manufacturer), making comparisons across studies difficult. Collision profiles were different between competition standard and position where elite players and forwards sustained more moderate- and high-intensity collisions than sub-elite players and backs, respectively. The recent inclusion of GPS-derived metabolic indices to activity profiles has also accentuated the distinctive workloads of positional groups during match-play where adjustables demonstrate the highest energy expenditure and metabolic power.

CONCLUSIONS

This review and the results of the meta-analysis have demonstrated that positional groups have varied kinematic and metabolic demands. During match play, forwards sustain the greatest number of collisions and RHIE bouts, outside backs participate in more high-speed running and cover the greatest distance, and adjustables work at high intensity covering the greatest relative distance with the highest metabolic cost. Therefore, specific training for each positional group should address their match requirements. In addition, sub-elite players exhibit lower intensity of play compared with elite players, as indicated by lower relative distance and less number of efforts per RHIE bout despite similarities in total distance covered and number of RHIE bouts. To prepare them for elite-level play, their training should incorporate higher intensity drills in which greater relative distance and number of efforts per RHIE bout are performed. Furthermore, the lack of consistency in the definition of speed zones, high-intensity movement, collisions and impacts, underscores the difficulties encountered in meaningful comparisons of player activity profiles between studies. Consensus of these definitions would facilitate direct comparisons within rugby league.

A Review of Instrumented Equipment to Investigate Head Impacts in Sport

Contact, collision, and combat sports have more head impacts as compared to noncontact sports; therefore, such sports are uniquely suited to the investigation of head impact biomechanics. Recent advances in technology have enabled the development of instrumented equipment, which can estimate the head impact kinematics of human subjects in vivo. Literature pertaining to head impact measurement devices was reviewed and usage, in terms of validation and field studies, of such devices was discussed. Over the past decade, instrumented equipment has recorded millions of impacts in the laboratory, on the field, in the ring, and on the ice. Instrumented equipment is not without limitations; however, in vivo head impact data is crucial to investigate head injury mechanisms and further the understanding of concussion.

The Use of Wearable Microsensors to Quantify Sport-Specific Movements

BACKGROUND

Microtechnology has allowed sport scientists to understand the locomotor demands of various sports. While wearable global positioning technology has been used to quantify the locomotor demands of sporting activities, microsensors (i.e. accelerometers, gyroscopes and magnetometers) embedded within the units also have the capability to detect sport-specific movements.

OBJECTIVE

The objective of this study was to determine the extent to which microsensors (also referred to as inertial measurement units and microelectromechanical sensors) have been utilised in quantifying sport-specific movements.

METHODS

A systematic review of the use of microsensors and associated terms to evaluate sport-specific movements was conducted; permutations of the terms used included alternate names of the various technologies used, their applications and different applied environments. Studies for this review were published between 2008 and 2014 and were identified through a systematic search of six electronic databases: Academic Search Complete, CINAHL, PsycINFO, PubMed, SPORTDiscus, and Web of Science. Articles were required to have used athlete-mounted sensors to detect sport-specific movements (e.g. rugby union tackle) rather than sensors mounted to equipment and monitoring generic movement patterns.

RESULTS

A total of 2395 studies were initially retrieved from the six databases and 737 results were removed as they were duplicates, review articles or conference abstracts. After screening titles and abstracts of the remaining papers, the full text of 47 papers was reviewed, resulting in the inclusion of 28 articles that met the set criteria around the application of microsensors for detecting sport-specific movements. Eight articles addressed the use of microsensors within individual sports, team sports provided seven results, water sports provided eight articles, and five articles addressed the use of microsensors in snow sports. All articles provided evidence of the ability of microsensors to detect sport-specific movements. Results demonstrated varying purposes for the use of microsensors, encompassing the detection of movement and movement frequency, the identification of movement errors and the assessment of forces during collisions.

CONCLUSION

This systematic review has highlighted the use of microsensors to detect sport-specific movements across a wide range of individual and team sports. The ability of microsensors to capture sport-specific movements emphasises the capability of this technology to provide further detail on athlete demands and performance. However, there was mixed evidence on the ability of microsensors to quantify some movements (e.g. tackling within rugby union, rugby league and Australian rules football). Given these contrasting results, further research is required to validate the ability of wearable microsensors containing accelerometers, gyroscopes and magnetometers to detect tackles in collision sports, as well as other contact events such as the ruck, maul and scrum in rugby union.

Classification of team sport activities using a single wearable tracking device

Wearable tracking devices incorporating accelerometers and gyroscopes are increasingly being used for activity analysis in sports. However, minimal research exists relating to their ability to classify common activities. The purpose of this study was to determine whether data obtained from a single wearable tracking device can be used to classify team sport-related activities. Seventy-six non-elite sporting participants were tested during a simulated team sport circuit (involving stationary, walking, jogging, running, changing direction, counter-movement jumping, jumping for distance and tackling activities) in a laboratory setting. A MinimaxX S4 wearable tracking device was worn below the neck, in-line and dorsal to the first to fifth thoracic vertebrae of the spine, with tri-axial accelerometer and gyroscope data collected at 100Hz. Multiple time domain, frequency domain and custom features were extracted from each sensor using 0.5, 1.0, and 1.5s movement capture durations. Features were further screened using a combination of ANOVA and Lasso methods. Relevant features were used to classify the eight activities performed using the Random Forest (RF), Support Vector Machine (SVM) and Logistic Model Tree (LMT) algorithms. The LMT (79-92% classification accuracy) outperformed RF (32-43%) and SVM algorithms (27-40%), obtaining strongest performance using the full model (accelerometer and gyroscope inputs). Processing time can be reduced through feature selection methods (range 1.5-30.2%), however a trade-off exists between classification accuracy and processing time. Movement capture duration also had little impact on classification accuracy or processing time. In sporting scenarios where wearable tracking devices are employed, it is both possible and feasible to accurately classify team sport-related activities.