Biology subjects

Bezodis, N.

Publications and source records attributed to Bezodis, N..

3 recordsLinked to original sources

bioRxiv2021

Determining jumping performance from a single body-worn accelerometer using machine learning

External peak power in the countermovement jump is frequently used to monitor athlete training. The gold standard method uses force platforms, but they are unsuitable for field-based testing. However, alternatives based on jump flight time or Newtonian methods applied to inertial sensor data have not been sufficiently accurate for athlete monitoring. Instead, we developed a machine learning model based on characteristic features (functional principal components) extracted from a single body-worn accelerometer. Data were collected from 69 male and female athletes at recreational, club or national levels, who performed 696 jumps in total. We considered vertical countermovement jumps (with and without arm swing), sensor anatomical locations, machine learning models and whether to use resultant or triaxial signals. Using a novel surrogate model optimisation procedure, we obtained the lowest errors with a support vector machine when using the resultant signal from a lower back sensor in jumps without arm swing. This model had a peak power RMSE of 2.3 W{middle dot}kg-1 (5.1% of the mean), estimated using nested cross validation and supported by an independent holdout test (2.0 W{middle dot}kg-1). This error is lower than in previous studies, although it is not yet sufficiently accurate for a field-based method. Our results demonstrate that functional data representations work well in machine learning by reducing model complexity in applications where signals are aligned in time. Our optimisation procedure also was shown to be robust can be used in wider applications with low-cost, noisy objective functions.

bioinformatics↗

bioRxiv2021

Animal lifestyle changes acceptable mass limits for attached tags

O_LIAnimal-attached devices have transformed our understanding of vertebrate ecology. To minimize tag-related harm for these studies, researchers have long advocated that tag masses should not exceed 3% of the animals body mass. However, this proposition ignores tag forces generated as a result of animal movement. C_LIO_LIUsing data from collar-attached accelerometers on diverse free-ranging terrestrial animals, we detail a tag-based acceleration method (TbAM) in which we quantify animal athleticism in terms of fractions of animal movement time devoted to different collar-recorded accelerations. The varying accelerations are converted to forces imposed on the animals based on the acceleration and tag mass and allow derivation of defined force limits, including those amounting to 3% of the animals mass, for specified fractions of any animals active time. C_LIO_LIWe demonstrate how species athleticism is the principal determinant of tag forces, whereas body mass is of little importance. Forces exerted by 3% tags were mostly equivalent to 4-19% of the animals masses during moving, with a maximum of 54% in a hunting cheetah. Cumulative frequency curves of tag acceleration for periods when animals were active, all showed a characteristic sigmoid pattern, which was displaced further to the right as higher acceleration activities accounted for an increasing proportion of any animals time. Specifying that tags should exert forces that are less than 3% of the animals body mass for 95% of the time led to corrected tag masses constituting between 1.6% and 2.98% of our study animals masses, with values depending on animal athleticism. C_LIO_LIRecognition that animal athleticism affects tag forces of their carriers fundamentally changes how acceptable tag mass limits should be determined by ethics bodies. In order to have a scientifically robust acceptable threshold to limit the forces experienced by an animal carrier, we suggest practitioners derive a similar cumulative acceleration profile for their study species and use a minimum of the 95% limits on the plot (although higher limits may be more appropriate). C_LI

zoology↗

bioRxiv2020

Forces experienced by instrumented animals depend on lifestyle

Animal-attached devices have transformed our understanding of vertebrate ecology. However, to be acceptable, researchers must minimize tag-related harm. The long-standing recommendation that tag masses should not exceed 3% of the animals body mass ignores tag forces generated by movement. We used collar-attached accelerometers on four free-ranging carnivores, spanning two orders of magnitude in mass, to reveal that during movement, forces exerted by 3% tags were generally equivalent to 4-19% of the animals masses, with a record of 54% in a hunting cheetah. Controlled studies on domestic dogs revealed how the tag forces are dictated by animal gait and speed but appear largely invariant of body mass. This fundamentally changes how acceptable tag mass limits should be determined, requiring cognizance of animal athleticism. One Sentence SummaryThere can be no universal rule for collar-tag masses as a percentage of carrier mass since tag forces depend on lifestyle.

ecology↗