A marker-less human motion analysis system for motion-based biomarker identification and quantification in knee disorders.

In recent years the healthcare industry has had increased difficulty seeing all low-risk patients, including but not limited to suspected osteoarthritis (OA) patients. To help address the increased waiting lists and shortages of staff, we propose a novel method of automated biomarker identification and quantification for the monitoring of treatment or disease progression through the analysis of clinical motion data captured from a standard RGB video camera. The proposed method allows for the measurement of biomechanics information and analysis of their clinical significance, in both a cheap and sensitive alternative to the traditional motion capture techniques. These methods and results validate the capabilities of standard RGB cameras in clinical environments to capture clinically relevant motion data. Our method focuses on generating 3D human shape and pose from 2D video data via adversarial training in a deep neural network with a self-attention mechanism to encode both spatial and temporal information. Biomarker identification using Principal Component Analysis (PCA) allows the production of representative features from motion data and uses these to generate a clinical report automatically. These new biomarkers can then be used to assess the success of treatment and track the progress of rehabilitation or to monitor the progression of the disease. These methods have been validated with a small clinical study, by administering a local anaesthetic to a small population with knee pain, this allows these new representative biomarkers to be validated as statistically significant (p-value <0.05). These significant biomarkers include the cumulative acceleration of elbow flexion/extension in a sit-to-stand, as well as the smoothness of the knee and elbow flexion/extension in both a squat and sit-to-stand.

The effects of concurrent biomechanical biofeedback on novel skill acquisition.

The aim of this study was to assess the effects of concurrent biomechanical biofeedback on the ability of novices to modify relative knee, spine, and elbow motions during a rowing-type task. After six non-instructed practice sessions, novices were assigned to a biofeedback (BFb; n = 7) or control group (Con; n = 7), before six, ten-minute sessions of continuous rowing were performed over 2 weeks. The BFb group received concurrent, visual biofeedback for developing sequential timing of knee, spine, and elbow motions during the pull. Following the intervention, the BFb group demonstrated delayed elbow flexion initiation (pre-intervention, 46 ± 11% pull; post-intervention, 78 ± 3% pull; p = 0.001). The biofeedback further promoted the consecutive ending of joint rotations (BFb: knee, 69 ± 4% pull; spine, 73 ± 7% pull; elbow, 85 ± 3% pull; Con: knee, 79 ± 8% pull; spine, 28 ± 6% pull; elbow, 79 ± 4% pull) and a move towards the sequential sequencing pattern. Concurrent biomechanical biofeedback during short-term training altered technique, possibly by providing guidance towards the desired movement pattern and increasing error detection and correction capabilities.

The effects of concurrent biomechanical biofeedback on rowing performance at different stroke rates.

The aims of this study were to assess the effects of stroke rate (SR) on the ability of trained rowers to: a) comply with concurrent biomechanical biofeedback on knee-back-elbow joint sequencing; and b) transfer any changes to competition-intensity conditions (maximal rowing task). Following a five-minute maximal rowing task (Baseline), 30 trained rowers were randomised to four groups. Two groups rowed at high SRs (90% maximum SR with biofeedback (BFb90) or control), while others rowed at low SRs (60% maximum SR with biofeedback (BFb60) or control) for 3 sessions. All rowers then completed another maximal rowing task (Transfer). Rowers complied with the biofeedback at both SRs, which promoted coordinative changes to knee-elbow motions during the pull. During Transfer, control rowers did not improve whereas those receiving biofeedback covered significantly greater distances (increase from Baseline: BFb60 = 6 ± 5%; BFb90 = 5 ± 4%; p < 0.05). However, movement adaptations were temporally different between SRs and were better maintained into Transfer by those that rowed at higher rates. This indicated biofeedback specificity, as transference of modified movement patterns appeared better when acquisition and transfer conditions were similar. These findings have practical implications for assimilating biofeedback into training programmes.

Self-Selected Versus Standardised Warm-Ups; Physiological Response on 500 m Sprint Kayak Performance.

This study investigated the effectiveness of a self-selected (SS) warm-up on 500 m sprint kayak performance (K500) compared to continuous (CON) and intermittent high intensity (INT)-type warm-ups. Twelve nationally ranked sprint kayakers (age 17.7 ± 2.3 years, mass 69.2 ± 10.8 kg) performed CON (15 min at the power at 2 m·mol-1), INT (10 min at 2 m·mol-1, followed by 5 × 10 s sprints at 200% power at VO2max with 50 s recovery at 55% power at VO2max), and SS (athlete's normal competition warm-up) warm-ups in a randomised order. After a five-minute passive recovery, K500 performance was determined on a kayak ergometer. Heart rate and blood lactate (BLa) were recorded before and immediately after each warm-up and K500 performance. Ratings of perceived exertion (RPE) were recorded at the end of the warm-up and K500. BLa, heart rate, and RPE were generally higher after the INT than CON and SS warm-ups (p < 0.05). No differences in these parameters were found between the conditions for the time trial (p > 0.05). RPE and changes in BLa and heart rate after the K500 were comparable. There were no differences in K500 performance after the CON, SS, or INT warm-ups. Applied practitioners can, therefore, attain similar performance independent of warm-up type.

Lower limb biomechanics before and after anterior cruciate ligament reconstruction: A systematic review.

This review aimed to synthesise the findings of literature that have assessed the changes in lower limb biomechanics following anterior cruciate ligament (ACL) reconstructive surgery. Systematic searches of CINHAL, MEDLINE, SCOPUS, and SPORTDiscus databases were run. All included studies had presented biomechanical variables pre- and post-surgery for the same participants. Articles were categorised by the analysed movement, and effect sizes were calculated. Fifty-four studies met the inclusion criteria, providing data on gait (n = 31), balance (n = 12), joint position sense (n = 5), stair ambulation (n = 4), pivoting (n = 6), and landing (n = 5). Measures of balance performance and joint position sense showed improvements from pre- to post-surgery. Changes in joint kinematics were inconsistent between studies, however increased knee flexion excursion, and reduced tibial anterior translation and internal rotation post reconstruction were identified. Joint kinetics reduced in magnitude in the early stages after surgery (≤5 weeks), then increased later in recovery (≥24 weeks). Risk of bias assessment identified most articles had a moderate or high risk (low = 5; moderate = 21; high = 11) resulting from participant retention and surgical intervention differences. The results of the review identified that although lower limb biomechanics did alter following reconstruction, few variables provided consistent results across studies and tasks. The low methodological quality of some articles may have contributed to these inconsistent findings. Alternatively, differences across studies may have resulted from individual coping strategies of participants that have previously been suggested to be present before reconstructive surgery, and future research should look to explore individual coping strategies to ACL reconstruction.

Direct and indirect effects of joint torque inputs during an induced speed analysis of a swinging motion.

This study proposed a method to quantify direct and indirect effects of the joint torque inputs in the speed-generating mechanism of a swinging motion. Linear and angular accelerations of all segments within a multi-linked system can be expressed as the sum of contributions from a joint torque term, gravitational force term and motion-dependent term (MDT), where the MDT is a nonlinear term consisting of centrifugal force, Coriolis force and gyroscopic effect moment components. Direct effects result from angular accelerations induced by a joint torque at a given instant, whereas indirect effects arise through the MDT induced by joint torques exerted in the past. These two effects were quantified for the kicking-side leg during a rugby place kick. The MDT was the largest contributor to the foot centre of gravity (CG)'s speed at ball contact. Of the factors responsible for generating the MDT, the direct and indirect effects of the hip flexion-extension torque during both the flight phase (from the final kicking foot take-off to support foot contact) and the subsequent support phase (from support foot contact to ball contact) were important contributors to the foot CG's speed at ball contact. The indirect effect of the ankle plantar-dorsal flexion torque and the direct effect of the knee flexion-extension torque during the support phase showed the largest positive and negative contributions to the foot CG's speed at ball contact, respectively. The proposed method allows the identification of which individual joint torque axes are crucial and the timings of joint torque exertion that are used to generate a high speed of the distal point of a multi-linked system.

Modelling error distribution in the ground reaction force during an induced-acceleration analysis of running in rear-foot strikers.

The objective of this study was to develop and evaluate a methodology for quantifying the contributions of modelling error terms, as well as individual joint torque, gravitational force and motion-dependent terms, to the generation of ground reaction force (GRF), whose true value can be measured with high accuracy using a force platform. Dynamic contributions to the GRF were derived from the combination of (1) the equations of motion for the individual segments, (2) the equations for constraint conditions arising from the connection of adjacent segments at joints, and (3) the equations for anatomical constraint axes at certain joints. The contribution of the error term was divided into four components caused by fluctuation of segment lengths, geometric variation in the constraint joint axes, and residual joint force and moment errors. The proposed methodology was applied to the running motion of thirteen rear-foot strikers at a constant speed of 3.3 m/s. Modelling errors arose primarily from fluctuations in support leg segment lengths and rapid movement of the virtual joint between the foot and ground during the first 20% of stance phase. The magnitudes of these error contributions to the vertical and anterior/posterior components of the GRF are presented alongside the non-error contributions, of which the joint torque term was the largest.

Kicking foot swing planes and support leg kinematics in rugby place kicking: Differences between accurate and inaccurate kickers.

Place kicking is a complex whole-body movement that contributes 45% of the points scored in international Rugby Union. This study compared the kicking foot swing plane characteristics of accurate and inaccurate kickers, underpinned by differences in their support leg and pelvis kinematics at support foot contact, to identify key technique characteristics. Motion capture data (240 Hz) were collected from 33 experienced kickers, and distinct groups of accurate (n = 18) and inaccurate (n = 8) kickers were identified based on their performance characteristics. All accurate kickers were capable of kicking successfully from at least 33.3 m, whereas all inaccurate kickers would have missed left from distances greater than 30.7 m. The accurate group exhibited a moderately shallower swing plane inclination (50.6 ± 4.8° vs. 54.3 ± 2.1°) and directed the plane moderately further to the right of the target (20.2 ± 5.4° vs. 16.7 ± 4.1°). At support foot contact, the accurate group placed their support foot moderately less far behind the ball (0.08 ± 0.08 m vs. 0.12 ± 0.04 m) and positioned their centre of mass moderately further to the support leg side (0.77 ± 0.07 m vs. 0.72 ± 0.01 m) due to a moderately greater stance leg lean (29.3 ± 4.1° vs. 26.8 ± 3.2°). The kicking foot swing plane is highly planar in rugby place kicking but its orientation differs between accurate and inaccurate kickers. These plane characteristics may be controlled by support foot placement and support leg and pelvis kinematics at support foot contact.

Age-congruency and contact effects in body expression recognition from point-light displays (PLD).

Recognition of older people's body expressions is a crucial social skill. We here investigate how age, not just of the observer, but also of the observed individual, affects this skill. Age may influence the ability to recognize other people's body expressions by changes in one's own ability to perform certain action over the life-span (i.e., an own-age bias may occur, with best recognition for one's own age). Whole body point light displays of children, young adults and older adults (>70 years) expressing six different emotions were presented to observers of the same three age-groups. Across two variations of the paradigm, no evidence for the predicted own-age bias (a cross-over interaction between one's own age and the observed person's age) was found. Instead, experience effects were found with children better recognizing older actors' expressions of 'active emotions,' such as anger and happiness with greater exposure in daily life. Together, the findings suggest that age-related changes in one own's mobility only influences body expression categorization in young children who interact frequently with older adults.

The planarity of the stickface motion in the field hockey hit.

The field hockey hit is an important but poorly understood stroke. In this study, we investigated the planarity of the stickface motion during the downswing to better characterize the kinematics and to assess the suitability of planar pendulum models for simulating the hit. Thirteen experienced female field hockey players were filmed executing hits with a single approach step, and the kinematics of the centre of the stickface were measured. A method was developed for identifying how far back from impact the stickface motion was planar. Orthogonal regression was used to fit least-squares planes to the stickface path during sections of the downswing of varying length, with each section ending at impact. A section was considered planar if the root mean square residual between the stickface path and the fitted plane was less than 0.25% of the distance travelled by the stickface during that period. On average, the stickface motion was planar for the last 83 ± 12% of its downswing path, with the length of the planar section ranging from 1.85 m to 2.70 m. The suitability of a planar model for the stickface motion was supported, but further investigation of the stick and arm kinematics is warranted.