1
|
Football Game Video Analysis Method with Deep Learning. COMPUTATIONAL INTELLIGENCE AND NEUROSCIENCE 2022; 2022:3284156. [PMID: 35720915 PMCID: PMC9200495 DOI: 10.1155/2022/3284156] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Figures] [Subscribe] [Scholar Register] [Received: 04/08/2022] [Accepted: 05/17/2022] [Indexed: 11/18/2022]
Abstract
Football is a beloved sport, and its wide audience makes football video one of the most analytically valuable types of video. Researchers have achieved certain research results in football video content analysis. How to locate interesting event clips from a complete long video is an urgent issue to be addressed in football game video analysis. The granularity of sports event detection results with traditional machine learning is relatively coarse, and the types of events that can be detected are limited. In recent years, deep learning has made good progress in the research of video single-person events and action detection, but there are few achievements in the detection of sports video events. In response to this problem, this work uses a deep learning method to build an event detection model to detect events contained in football videos. The whole model is divided into two stages, in which the first stage is utilized to generate candidate event fragments. It divides the football video to be detected into a sequence of frames of a certain length and scans using a sliding window. Multiple frame sequences within a sliding window form a segment, and each segment is a prediction unit. The frame sequence features within the segment are obtained through a three-dimensional convolutional neural network, which is used as the input of each time point of the bidirectional recurrent neural network and further integrated to generate the event prediction of the segment. The second stage is to further process the above results to remove all segments predicted as nonevents. The thresholds are set according to the detection effect of various events to filter out event fragments with higher probability values, obtain the start and end positions of the events through merging, classify and mark them, and finally output complete event fragments. This work has carried out comprehensive and systematic experiments to verify correctness of the proposed method.
Collapse
|
2
|
Pankow MP, Syrydiuk RA, Kolstad AT, Hayden AK, Dennison CR, Mrazik M, Hagel BE, Emery CA. Head Games: A Systematic Review and Meta-analysis Examining Concussion and Head Impact Incidence Rates, Modifiable Risk Factors, and Prevention Strategies in Youth Tackle Football. Sports Med 2021; 52:1259-1272. [PMID: 34894348 DOI: 10.1007/s40279-021-01609-4] [Citation(s) in RCA: 6] [Impact Index Per Article: 2.0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Accepted: 11/21/2021] [Indexed: 10/19/2022]
Abstract
OBJECTIVES The aims were to (1) examine the rates and mechanisms of concussion and head impact in youth football (high school level or younger); (2) identify modifiable risk factors for concussion and head impact; and (3) evaluate the effectiveness of prevention strategies in tackle football at any level. METHODS Nine databases (CINAHL Plus with Full Text; Cochrane Central Register of Controlled Trials; EMBASE; ERIC; Ovid MEDLINE(R) and Epub Ahead of Print, In-Process & Other Non-Indexed Citations and Daily; ProQuest Dissertations & Theses Global Database; PsycINFO; Scopus; and SPORTDiscus with Full Text) were searched using the search strategy focusing on four main concepts: concussion/head impact, tackle football, modifiable risk factors, and primary prevention. Two reviewers completed title, abstract, and full-text screening as well as risk of bias assessment (using the Downs and Black checklist), with a third author available to resolve any disagreements. MAIN RESULTS After removing duplicates, 1911 articles were returned. Fifty-eight articles were included in the review and 20 in the meta-analysis. The overall combined rates of concussion (including game and practice-related concussion) based on the meta-analysis were 0.78 concussions/1000 athlete exposures [95% confidence interval (CI) 0.67-0.89] for high school football (ages 13-19) and 1.15 concussions/1000 athlete exposures (95% CI 0.89-1.41) for minor football players (ages 5-15). There is evidence that contact training and practice contact restrictions have reduced the rate of head impacts and concussion. Heads Up Football (an intervention focused on coach education and contact training) has been shown to reduce the rate of concussion by 32% and head impacts by 38% amongst high school football players. Limiting contact practices in high schools to 2 days per week reduced practice head impacts per player-season by 42%, and limiting full contact in practice to 75 min per week in the second week of the season and 60 min in week 3 and beyond resulted in a 54% decrease in the practice-related concussion rate (p = 0.003). CONCLUSIONS This review identified a critical need for interventions to address the high rates of concussion and head impact in youth football. To date, contact training and contact restrictions have the strongest evidence supporting their effectiveness at reducing these rates. Future research should use consistent concussion definitions and validated injury surveillance systems, and ensure complete reporting of participant characteristics and sampling details. Prospero ID CRD42020193775.
Collapse
Affiliation(s)
- Mark Patrick Pankow
- Faculty of Kinesiology, Sport Injury Prevention Research Centre, University of Calgary, 2500 University Dr NW, Calgary, AB, T2N 1N4, Canada. .,O'Brien Institute of Public Health, University of Calgary, Calgary, Canada. .,Alberta Children's Hospital Research Institute, University of Calgary, Calgary, Canada. .,Hotchkiss Brain Institute, University of Calgary, Calgary, Canada.
| | - R A Syrydiuk
- Faculty of Kinesiology, Sport Injury Prevention Research Centre, University of Calgary, 2500 University Dr NW, Calgary, AB, T2N 1N4, Canada.,O'Brien Institute of Public Health, University of Calgary, Calgary, Canada.,Alberta Children's Hospital Research Institute, University of Calgary, Calgary, Canada.,Hotchkiss Brain Institute, University of Calgary, Calgary, Canada
| | - A T Kolstad
- Faculty of Kinesiology, Sport Injury Prevention Research Centre, University of Calgary, 2500 University Dr NW, Calgary, AB, T2N 1N4, Canada.,O'Brien Institute of Public Health, University of Calgary, Calgary, Canada.,Alberta Children's Hospital Research Institute, University of Calgary, Calgary, Canada.,Hotchkiss Brain Institute, University of Calgary, Calgary, Canada
| | - A K Hayden
- Libraries and Cultural Resources, University of Calgary, Calgary, Canada
| | - C R Dennison
- Department of Mechanical Engineering, University of Victoria, Victoria, Canada.,Biomedical Instrumentation Lab, Department of Mechanical Engineering, University of Alberta, Edmonton, Canada
| | - M Mrazik
- Department of Educational Psychology, University of Alberta, Edmonton, AB, Canada
| | - B E Hagel
- Faculty of Kinesiology, Sport Injury Prevention Research Centre, University of Calgary, 2500 University Dr NW, Calgary, AB, T2N 1N4, Canada.,O'Brien Institute of Public Health, University of Calgary, Calgary, Canada.,Alberta Children's Hospital Research Institute, University of Calgary, Calgary, Canada.,Community Health Sciences, Cumming School of Medicine, University of Calgary, Calgary, Canada.,Department of Pediatrics, Cumming School of Medicine, University of Calgary, Calgary, Canada
| | - C A Emery
- Faculty of Kinesiology, Sport Injury Prevention Research Centre, University of Calgary, 2500 University Dr NW, Calgary, AB, T2N 1N4, Canada.,O'Brien Institute of Public Health, University of Calgary, Calgary, Canada.,Alberta Children's Hospital Research Institute, University of Calgary, Calgary, Canada.,Hotchkiss Brain Institute, University of Calgary, Calgary, Canada.,Community Health Sciences, Cumming School of Medicine, University of Calgary, Calgary, Canada.,Department of Pediatrics, Cumming School of Medicine, University of Calgary, Calgary, Canada
| |
Collapse
|
3
|
Kent JB, Sasser P, Austin AV, MacKnight JM. Are you missing a concussion by watching American football? Video analysis of ball location in a Division 1 football program. Brain Inj 2021; 35:880-885. [PMID: 33896298 DOI: 10.1080/02699052.2021.1917658] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 10/21/2022]
Abstract
OBJECTIVE Action in a football game occurs quickly. Medical staff can miss a sport-related concussion (SRC) if they do not observe it directly. The objective of this study is to determine if SRCs occur more frequently at ball than away from ball during gameplay. Game-specific concussion statistics can enhance medical provider care of athletes. METHODS We used gameplay videos of an NCAA Division I football program to analyze SRCs and determine the primary tackler or ball handler during each concussion play. We compared the relative risk of SRCs for the primary ball handlers/tacklers to that of the other 10 players on the same team during that play. RESULTS Over 10 seasons, 26 SRCs occurred at ball for the primary ball handler/tackler position (0.22 SRCs/game) versus 16 away from ball (0.13 SRCs/game). The relative risk of an SRC according to exposure (at ball) vs. no exposure (away from ball) was 16.2 (CI 8.7-30.2, P < .05). Special teams had more SRCs away from ball than at ball, but this was not significant (relative risk 3.32, CI 0.90-12.3, P > .05). CONCLUSIONS The study provides medical staff guidance to more efficiently identify in-game SRCs and supports evidence for rules changes.
Collapse
Affiliation(s)
- Jeremy B Kent
- Department of Family Medicine, Team Physician, UVa Sports Medicine, University of Virginia Health System, Charlottesville, VA, USA
| | - Phillip Sasser
- Department of Pediatrics, University of Wisconsin Health, Madison, WI, USA
| | - Ashley V Austin
- Department of Family Medicine, University of Washington School of Medicine, Seattle, WA, USA
| | - John M MacKnight
- Team Physician and Medical Director, UVA Sports Medicine, University of Virginia Health System, Charlottesville, VA, USA
| |
Collapse
|
4
|
Saw AE, Howard DJ, Kountouris A, McIntosh AS, Orchard JW, Saw R, Hill T. Situational factors associated with concussion in cricket identified from video analysis. JOURNAL OF CONCUSSION 2020. [DOI: 10.1177/2059700220947197] [Citation(s) in RCA: 2] [Impact Index Per Article: 0.5] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/17/2022] Open
Abstract
Video analysis of situational factors associated with head impacts and concussion has been completed in several sports, however has yet to be completed in cricket. This study aimed to identify situational factors associated with concussion in elite Australian male and female cricket. Match video of head impacts were coded for player position, impacting object, source of ball, location of impact, and where the ball went after impact. Head impacts were then categorised as either concussion or no concussion based on clinical diagnosis. Data for 197 head impacts included 35 (18%) which were diagnosed as concussion. Head impacts typically occurred to an on-strike batter facing a pace bowler (84%). If the ball stopped or rebounded towards the source, 21% were diagnosed as concussion (13% if the ball deflected away from the source). If impact was to an unprotected head, 38% were diagnosed as concussion (16% if impact was to a helmet). If impact was to the back of the helmet or head, 40% were diagnosed as concussion (11–21% for other areas of the head or helmet). The combination of situational factors most consistent with concussion were impact from ball that hit the back of helmet or head and stopped or rebounded towards the source (PPV 80%, p = 0.002). Consideration of the situational factors of a head impact may improve the speed and accuracy of clinical decision making on whether to remove a player from the field for further assessment, particularly if clinical signs are unclear. Video may be used as a tool to support this process. Improved impact attenuation of cricket helmets, particularly at the back, may reduce the risk of concussion.
Collapse
Affiliation(s)
- Anna E Saw
- Cricket Australia, Jolimont, VIC, Australia
| | - David J Howard
- Sunshine Coast University Hospital Health Service, Birtinya, QLD, Australia
| | - Alex Kountouris
- Cricket Australia, Jolimont, VIC, Australia
- La Trobe Sport and Exercise Medicine Research Centre, La Trobe University, Bundoora, VIC, Australia
| | - Andrew S McIntosh
- McIntosh Consultancy and Research, Sydney, NSW, Australia
- ACRISP and School of Engineering, Edith Cowan University, Joondalup, WA, Australia
| | | | | | | |
Collapse
|
5
|
Kent JB, Wood CL, Pugh K, Statuta SM, MacKnight JM. The medical observer in American football: a survey of use and efficacy. Brain Inj 2020; 34:1100-1105. [DOI: 10.1080/02699052.2020.1782473] [Citation(s) in RCA: 2] [Impact Index Per Article: 0.5] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 10/24/2022]
Affiliation(s)
- Jeremy B. Kent
- Department of Family Medicine, Team Physician, UVa Sports Medicine, University of Virginia Health System, Charlottesville, Virginia, Virginia, USA
| | - Colton L. Wood
- Primary Care Sports Medicine Fellow, Department of Family Medicine, University of North Carolina, Chapel Hill, North Caroline, USA
| | - Kelli Pugh
- Associate Athletics Director for Sports Medicine, University of Virginia, Charlottesville, Charlottesville, Virginia, USA
| | - Siobhan M. Statuta
- Family Medicine and Physical Medicine and Rehabilitation, University of Virginia Health System, Charlottesville, Virginia, USA
| | - John M. MacKnight
- Internal Medicine & Orthopaedic Surgery, Team Physician and Medical Director, UVA Sports Medicine, University of Virginia Health System
| |
Collapse
|