Computer vision-based assessment of cyclist-tram track interactions for predictive modeling of crossing success.

INTRODUCTION: Single Bicycle Brashes (SBCs) are common, and underreported in official statistics. In urban environments, light rail tram tracks are a frequent factor, however, they have not yet been the subject of engineering analysis. METHOD: This study employs video-based analysis at nine Dublin city centre locations and introduces a predictive model for crossing success on tram tracks, utilising cyclist crossing angles within a Surrogate Measure of Safety (SMoS) framework. Additionally, Convolutional Neural Networks (CNNs) were explored for automatic estimation of crossing angles. RESULTS: Modelling results indicate that cyclist crossing angle is a strong predictor of crossing success, and that cyclist velocity is not. Findings also highlight the prevalence of external factors which limit crossing angles for cyclists. In particular, kerbs are a common factor, along with passing/approaching vehicles or other cyclists. Furthermore, results indicate that further training on a relatively small sample of 100 domain-specific examples can achieve substantial accuracy improvements for cyclist detection (from 0.31AP0.5 to 0.98AP0.5) and crossing angle inference from traffic camera footage. CONCLUSIONS: Ensuring safe crossing angles is important for cyclist safety around tram tracks. Infrastructural planners should aim for intuitive, self-explainable road layouts that allow for and encourage crossing angles of 60° or more - ideally 90°. PRACTICAL APPLICATIONS: The SMoS framework and the open-source SafeCross1 application offer actionable insights and tools for enhancing cyclist safety around tram tracks.

Assessment of deep learning pose estimates for sports collision tracking.

Injury assessment during sporting collisions requires estimation of the associated kinematics. While marker-based solutions are widely accepted as providing accurate and reliable measurements, setup times are lengthy and it is not always possible to outfit athletes with restrictive equipment in sporting situations. A new generation of markerless motion capture based on deep learning techniques holds promise for enabling measurement of movement in the wild. The aim of this work is to evaluate the performance of a popular deep learning model "out of the box" for human pose estimation, on a dataset of ten staged rugby tackle movements performed in a marker-based motion capture laboratory with a system of three high-speed video cameras. An analysis of the discrepancy between joint positions estimated by the marker-based and markerless systems shows that the deep learning approach performs acceptably well in most instances, although high errors exist during challenging intervals of heavy occlusion and self-occlusion. In total, 75.6% of joint position estimates are found to have a mean absolute error (MAE) of less than or equal to 25 mm, 17.8% with MAE between 25 and 50 mm and 6.7% with MAE greater than 50 mm. The mean per joint position error is 47 mm.

A Systematic Review of the Educational Effectiveness of Simulation Used in Open Surgery.

STATEMENT: The role of simulation to teach and access open surgical skills has become more prevalent in recent years. This systematic review synthesizes the totality of evidence with respect to the educational effectiveness of simulators used in open surgical training. A systematic literature search was conducted in PubMed, Embase, CINAHL, Scopus, and Web of Science. Only randomized controlled trials were included that explored the educational efficacy of theses simulators. Six randomized controlled trials were included from the 9934 studies found. The methodological quality of the included studies was variable. Overall, the use of the simulators was more educationally effective compared with standard teaching of the skill without a simulator (P < 0.05). Two studies showed that the simulator was as good as an animal model of much higher fidelity. Further studies are needed to secure higher evidence for the educational value, validity, and transferability of the skills to the hospital setting for all simulators.

Does ball carrier technique influence tackler head injury assessment risk in elite rugby union?

The aim of this study was to use video evidence of tackles in elite level rugby union to identify ball carrier proficiency characteristics, for both lower and upper body tackles, that have a higher propensity to result in Head Injury Assessments(HIA) for the tackler. HIA (n = 74) and non-HIA tackles (n = 233) were categorised as either front-on or side-on upper or lower body tackles and scored for ball carrying proficiency characteristics. Side-on tackles included tackles from behind. A Chi-Square test (p < 0.05) and Cramer's V were calculated to compare proficiency characteristics in HIA and non-HIA cases. For front-on upper body tackles, the ball carrier "fending into contact" (p < 0.01;ES = Moderate) and "explosiveness on contact" (p = 0.04;ES = Moderate) had a higher propensity to result in a HIA for the tackler. Fending into contact was exhibited in 47% of all upper body Tackle front-on HIA cases. The fending arm contacted the tackler's head in 67% of these cases. Fending into contact can potentially be dangerous and therefore emphasis should be placed on safe fending during tackle-based training drills. Referees should also be alert to arm-to-head contact during the fend. Given the low number of ball carrier characteristics identified, focus should be placed on tackler characteristics for HIA prevention strategies.

Can tackle height influence head injury assessment risk in elite rugby union?

OBJECTIVES: Tackle height laws are an area of controversy in rugby union. It is reported that the tackler is at most risk of a Head Injury Assessment (HIA). Therefore, the aim of this study was to use match video evidence of tackles in elite level rugby union to examine the effect of tackle heights on HIA risk for the tackler. DESIGN: Qualitative observational case-control study. METHODS: Each HIA (n=74) and control tackle (n=965) was categorised based on tackle direction (front- or side-on), tackle type (arm, shoulder or smother) and tackle height (upper trunk, mid-trunk, lower trunk, upper leg or lower leg). The Relative Risk (RR), 95% Confidence Interval (CI) and probability (p) values were calculated for each tackle height. RESULTS: Intended primary contact at the upper trunk of the ball carrier had a greater propensity to result in a HIA for the tackler for front-on upper body shoulder tackles (RR=1.48; 95%CI=1.16-1.90; p<0.01) and side-on upper body smother tackles (RR=2.30; 95%CI=1.82-2.92; p<0.01). Intended primary contact at the upper leg of the ball carrier had a greater propensity to result in a HIA for the tackler for front-on (RR=2.60; 95%CI=1.70-3.97; p<0.01) and side-on (RR=3.34; 95%CI=1.65-6.79; p<0.01) lower body shoulder tackles. CONCLUSIONS: To reduce tackler HIA risk, the results suggest tackling below the upper trunk for upper body tackles. The results also suggest tackling at the lower trunk for lower body tackles and avoiding the upper legs. Prevention strategies should place emphasis on tackling lower risk body regions such as the mid- and lower trunk.

Assessment of model-based image-matching for future reconstruction of unhelmeted sport head impact kinematics.

Player-to-player contact inherent in many unhelmeted sports means that head impacts are a frequent occurrence. Model-Based Image-Matching (MBIM) provides a technique for the assessment of three-dimensional linear and rotational motion patterns from multiple camera views of a head impact event, but the accuracy is unknown for this application. The goal of this study is to assess the accuracy of the MBIM method relative to reflective marker-based motion analysis data for estimating six degree of freedom head displacements and velocities in a staged pedestrian impact scenario at 40 km/h. Results showed RMS error was under 20 mm for all linear head displacements and 0.01-0.04 rad for head rotations. For velocities, the MBIM method yielded RMS errors between 0.42 and 1.29 m/s for head linear velocities and 3.53-5.38 rad/s for angular velocities. This method is thus beneficial as a tool to directly measure six degree of freedom head positional data from video of sporting head impacts, but velocity data is less reliable. MBIM data, combined in future with velocity/acceleration data from wearable sensors could be used to provide input conditions and evaluate the outputs of multibody and finite element head models for brain injury assessment of sporting head impacts.

Detailed assessment of pedestrian ground contact injuries observed from in-depth accident data.

Most pedestrians struck by vehicles receive injuries from contact with the vehicle and also from the subsequent ground contact. However, ground related pedestrian injuries have received little focus. This paper uses 1221 German pedestrian collision cases occurring between 2000 and 2015 to assess the distribution and risk factors for pedestrian ground related injuries. Results show that for MAIS 2, the ground accounted for 24% of cases, for MAIS 3 the ground accounted for 20% of cases and for MAIS 4-5, the ground accounted for 14% of cases. There were no AIS 6 ground related injuries, though there were several fatal cases where the ground was coded as the most serious injury. The head, thorax and spine dominate AIS 4-5 ground contact injuries. Vehicle impact speeds were higher for ground related AIS 4-5 compared to AIS 2 injury cases and the average impact speed for ground related injuries to the upper and lower extremities was lower than for body regions like head, thorax and spine. There was a significant age effect on pedestrian ground related injury outcome, with older pedestrians suffering more severe injuries and the median age for thorax injuries was higher than for all other body regions. There was no significant difference in the proportions of AIS 2+ head injuries produced by ground contact for more recent vehicles (model year since 2005) compared to older vehicles (model year before 2005). However, logistic regression analysis showed that the normalised bonnet leading-edge height is a risk factor for adult pedestrian AIS2+ ground related head injuries, and this provides empirical support for recent computational modelling predictions which implied a relationship between vehicle shape and pedestrian ground contact injuries. Considering the potential benefits of preventing pedestrian ground contact, for collisions below 40km/h two thirds of the injury costs would be eliminated if ground contact could be prevented, and even higher benefits are likely at lower speeds (20 and 30km/h). These data demonstrate the importance of ground related pedestrian injuries and show that vehicle shape influences pedestrian injury outcome in ground contact. The data therefore provides significant motivation for countermeasures to prevent or moderate pedestrian ground related injuries.

The Effect of Tackler Technique on Head Injury Assessment Risk in Elite Rugby Union.

PURPOSE: This study aimed to use match video evidence of tackles in elite-level rugby union to identify tackler proficiency characteristics, for both lower body and upper body tackles, that result in head injury assessments (HIA) for the tackler. METHODS: A review of international rugby union matches (2013-2017) and Pro 12/European Rugby Champions Cup matches (2014-2017) from a professional rugby union club was conducted. HIA (n = 74) and non-HIA tackles (n = 233) were categorized as either front-on or side-on upper body or lower body tackles and were scored for tackling proficiency characteristics. A chi-square test (P < 0.05) and phi and Cramer's V were calculated to compare HIA and non-HIA tackling proficiency characteristics. RESULTS: In both front- and side-on upper body and lower body tackles, "head up and forward/face up" and "head placement on correct side of ball carrier" were identified as having a lower propensity to result in an HIA for the tackler. For both front-on and side-on upper body tackles, "identify/track ball carrier onto shoulder" and "shortening steps" were identified. In addition, "straight back, centre of gravity forward of support base" and "identify/track ball carrier onto shoulder" were identified for front-on and side-on lower body tackles, respectively. CONCLUSIONS: This study identified tackle characteristics that had a lower propensity to result in an HIA for the tackler in both front-on and side-on upper body and lower body tackles.