Techniques to derive and clean acceleration and deceleration data of athlete tracking technologies in team sports: A scoping review

Ellens, Susanne; Middleton, Kane; Gastin, Paul B.; Varley, Matthew C.

doi:10.1080/02640414.2022.2054535

Cited by 4 publications

(6 citation statements)

References 233 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For example, Lacome et al (37) and Akenhead et al (1) observed that MSS and maximum acceleration biases were larger with raw data vs. smoothed (moving average) data during linear sprinting tasks. In accordance with previous literature (21), a low-pass Butterworth filter was the most frequently reported filtering technique with both radar and GNSS data in this review. However, as mentioned, and important to note, only 2 GNSS studies reported all specific filtering specifications (e.g., digital, low-pass, dual-pass, or 1 Hz cutoff Butterworth; Table 3).…”

Section: Discussionsupporting

confidence: 84%

“…More recent studies have since shown improvements in agreement outcomes (CVs <5%) across metrics with improved intrasystem and intersystem reliability using 10 Hz GNSSs (11,14,22). Although measures of speed are achieving good agreement with higher sampling rates, present results suggest that GNSS is still suboptimal for quantifying acceleration metrics (21,58). It has been suggested that low sampling rates limit the ability to determine changes in velocity during the initial portions of the sprint where acceleration is likely the highest.…”

Section: Discussionmentioning

confidence: 70%

“…Primarily, sport science practitioners are encouraged to open dialog with sport technology companies to better understand GNSS units' proprietary signal processing algorithms. If this information cannot be shared, sport scientists should consider the postmanufacturer processing, sprint model processing, and reporting of sprint protocol and specification of the GNSS technology used (21,38). This will allow for better comparisons of the findings across studies and improvements in future research.

Practical Applications

Although there are some methodological challenges to sprint modeling using GNSS technology, validity and reliability studies suggest that sprint modeling can be performed and show promise for sprint monitoring in the field.…”

Section: Discussionmentioning

confidence: 99%

“…There are existing recommendations that sport scientists should evaluate when planning studies and disseminating findings from GNSS-derived outcomes (Malone et al (38) and Ellen et al (21)). This review adds to these recommendations by highlighting the need for standardization and transparent reporting of data processing to support best practice of sprint modeling.…”

Section: Discussionmentioning

confidence: 99%

“…System sensors sampling at 5 Hz showed larger bias than sensors sampling at 10 Hz and above when compared to radar or laser. Two studies used sensor fusion which resulted in a minimal mean bias of 20.01 m•s 21 for MSS, 20.17 m•s 21 for V 0 , 20.02 N•kg 21 for relative F 0 , and 20.31 W•kg 21 for relative P max compared with radar or timing gates (3,50). Filtering or smoothing the GNSS velocity signal resulted in better estimates than the use of raw data (1,37).…”

Section: Narrative Assessmentmentioning

confidence: 99%

See 4 more Smart Citations

A Systematic Review and Meta-Analysis of Wearable Satellite System Technology for Linear Sprint Profiling: Technological Innovations and Practical Applications

Cormier,

Meylan,

Agar-Newman

et al. 2023

Journal of Strength and Conditioning Research

View full text Add to dashboard Cite

Cormier, P, Meylan, C, Agar-Newman, D, Geneau, D, Epp-Stobbe, A, Lenetsky, S, and Klimstra, M. A systematic review and meta-analysis of wearable satellite system technology for linear sprint profiling: technological innovations and practical applications. J Strength Cond Res XX(X): 000–000, 2023—An emerging and promising practice is the use of global navigation satellite system (GNSS) technology to profile team-sports athletes in training and competition. Therefore, the purpose of this narrative systematic review with meta-analysis was to evaluate the literature regarding satellite system sensor usage for sprint modeling and to consolidate the findings to evaluate its validity and reliability. Following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, an electronic search of the databases, PubMed and SPORTDiscus (EBSCO), was conducted. Concurrent validity and reliability studies were considered, and 16 studies were retained for the review from the initial 1,485 studies identified. The effects on outcomes were expressed as standardized mean differences (SMDs, Cohen's d) for each outcome (i.e., maximal sprint speed [MSS], the acceleration constant [τ], maximal theoretical velocity [V 0], relative force [F 0], and relative power [Pmax]). Effect magnitudes represented the SMD between GNSS-derived and criterion-derived (i.e., radar and laser) and resulted in the following estimates: small for MSS (d = 0.22, 95% CI 0.01 to 0.42), τ (d = −0.18, 95% CI −0.60 to 0.23), V 0 (d = 0.14, 95% CI −0.08 to 0.36), relative F 0 (d = 0.15, 95% CI −0.25 to 0.55), and relative Pmax (d = 0.21, 95% CI −0.16 to 0.58). No publication bias was identified in meta-analyzed studies and moderator analysis revealed that several factors (sampling rate and sensor manufacturer) influenced the results. Heterogeneity between studies was considered moderate to high. This highlighted the differences between studies in sensor technology differences (i.e., sampling rate, sensor fusion, and satellite network acquisition), processing techniques, criterion technology used, sprint protocols, outcome reporting, and athlete characteristics. These findings may be useful in guiding improvements in sprint modeling using GNSS technology and enable more direct comparisons in future research. Implementation of all-out linear sprint efforts with GNSS technology can be integrated into sport-specific sessions for sprint modeling when robust and consistent data processing protocols are performed, which has important implications for fatigue monitoring, program design, systematic testing, and rehabilitation in individual and team sports.

show abstract

Section: Discussionsupporting

confidence: 84%

Section: Discussionmentioning

confidence: 70%

Practical Applications

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%