Estimating musculoskeletal loading and muscular adaptations to hypogravity using an optimal control approach

Abstract

In this thesis a computational framework was developed to prole the
musculoskeletal loading during exercise in hypogravity and to model muscular
adaptations to disuse. The aims were i) to create a Biomechanical Handbook
of normative internal musculoskeletal loading proles when exercising in
hypogravity, and ii) to assess how muscular adaptations to unloading can be
replicated with a Hill-type muscle model. A direct collocation framework was
used to estimate muscle and joint reaction forces, validated against the Knee
Grand Challenge dataset. The framework was then used to estimate lower-limb
joint reaction forces during single-leg hopping at ve hypogravity levels, and
predict exercise volume to avoid detrimental adaptations. Joint reaction forces
were estimated within 0.62 - 0.85 BW relative to the Knee Grand Challenge
data, with a peak error of 1.24 0.17 BW. The framework was also able to
detect the increase in peak joint reaction force as walking speed increased. The
hypogravity case-study revealed an increased quadriceps muscle forces and a shift
in rectus femoris force as gravity approached 1 g. When quadriceps muscle forces
were input into a muscle adaption model, predicted exercise volumes needed to
combat muscle adaptations decreased substantially with gravity. The framework
allows for the comparison between dierent movements and gravity levels needed
to create a Biomechanical Handbook. An experimental protocol, which expands
on the handbook vision, is presented to provide a blueprint for the analysis of
a catalogue of gait and jumping exercises in hypogravity to provide reference
values to the handbook. Finally, a Monte Carlo sampling technique was used to
perturb Hill-type muscle model parameters during an isokinetic knee extension
task. The results highlighted the Hill-type muscle model can replicate muscular
adaptations to unloading as long as optimal bre length is adjusted appropriately..
This information is key for future research to adjust musculoskeletal models
to achieve appropriate simulation results, which will improve application of
simulation methods to space science contexts.

Date of Award	12 Oct 2022
Original language	English
Awarding Institution	University of Bath
Supervisor	Dario Cazzola (Supervisor), Steffi Colyer (Supervisor) & Aki Salo (Supervisor)