Falls are one of the most serious problems among the elderly, resulting in fatal physical injuries. Early identification of people at a high risk of falling is needed to facilitate rehabilitation to reduce future fall risk. The overall goal of this dissertation was to develop biomechanical models that identify dynamic limits of balance control in daily functional activities associated with falling, including sit-to-stand (STS) movement, standing (stance perturbation), and walking. Poor performance of STS movement has been identified as one of the risk factors of falls among elderly individuals. We proposed a novel method to identify dynamic limits of balance control during STS movement using whole body center of mass (COM) acceleration and assessed its feasibility to differentiate individuals with difficulty in STS movement from healthy individuals. The results demonstrated that our model with COM acceleration could better differentiate individuals with difficulty in STS movement from healthy individuals than the traditional model with COM velocity. Poor postural control ability is also a risk factor of falls. Postural recovery responses to backward support surface translations during quiet standing were examined for healthy young and elderly adults. The results demonstrated that functional base of support (FBOS) and ankle dorsiflexor strength could be sensitive measures to detect elderly individuals with declined balance control. Our biomechanical model, which determines a set of balance stability boundaries, showed a better predictive capability than the statistical model for identifying unstable balance recovery trials, while the statistical model better predicted stable recovery trials. Lastly, walking requires a fine momentum control where COM acceleration could play an important role. Differences in control of dynamic stability during walking were examined with our proposed boundaries of dynamic stability. Elderly fallers adapted a more conservative gait strategy than healthy individuals, demonstrating significantly slower forward COM velocity and acceleration with their COM significantly closer to the base of support at toe-off, which could be indicative of a poor momentum control ability. Overall, this study demonstrated that COM acceleration would provide further information on momentum control, which could better reveal underlying mechanisms causing imbalance and provide an insightful evaluation of balance dysfunction. This dissertation includes unpublished co-authored material.

Spinal cord injury and stroke affect many people each year and can result in the loss of muscle function. Current research attempts to correct muscle paralysis through the use of mechanical braces or through open-loop stimulation methods. However, prosthetic systems that use closed-loop control strategies can offer improved functionality by accounting for the changing dynamics associated with the human body and external disturbances. In particular, closed-loop functional electrical stimulation (FES) offers the possibility of moving paralyzed muscles in a predetermined manner, allowing a paraplegic individual to regain the ability to perform some tasks. An experimental setup was designed for the development and testing of a closed-loop FES control system, as well as the characterization of muscle properties. Due to the complexities associated with using a human subject, an inverted pendulum model is utilized for this preliminary study. This model is a basic engineering control problem often used when studying postural control in humans. In particular, electrical stimuli will be applied to the gastrocnemius muscle of a frog in order to produce a contraction force that will drive an inverted pendulum and maintain its desired angle. The stimulation signal will be determined by control algorithms applied through the use of Matlab® and implemented in real-time with a data acquisition system. This setup will help provide an understanding of the muscle behavior and can be used to establish the validity of proposed controller methods.

Bakgrund: Transfemoralt amputerade har nedsatt balans och ökad fallrisk, men väldigt lite forskning är gjord om detta. Inverted pendulum model (IPM) är en balansmodell för icke-amputerade som bygger på ett känt samband mellan Center of Pressure (CoP) och Center of Mass (CoM). Syftet med denna studie är att kinetiskt validera den inverterade pendelmodellen för transfemoralt amputerade. Metod: I studien deltog amputerade (n=5) och en matchande kontrollgrupp (n=5). Man samlade in data genom att deltagarna fick stå stilla på två stycken kraftplattor under tre styckern villkor; öppna ögon, stängda ögon och weight-bearing feedback. Man undersökte sedan korrelationen mellan avståndet från CoM till CoP och CoMacc i både anterioposterior (A/P) riktning samt medio-lateral (M/L) riktning och analyserade datan med trevägsvariansanalys (ANOVA). Resultat: Resultatet visade att det fanns en signifikant interaktionseffekt mellan villkor och position i M/L-riktning. I A/P-riktning fanns det en signifikant interaktionseffekt mellan grupp och position samt villkor och position. Slutsats: Resultatet innebär att IPM är kinetiskt validerat i A/P-riktning sett till hela kroppen men inte på den amputerade sidan. Hur det intakta benet förhåller sig till IPM för amputerade är tvivelaktigt. / Background: It is known that transfemoral prosthesis users lack normal balance control and are more likely to fall. Research on this topic is insufficient. The Inverted Pendulum Model (IPM) is a commonly used biomechanical model for assessment of balance and postural control for healthy individuals based on an assumption that Center of Pressure (CoP) and Center of Mass (CoM) are inter-dependent. The aim of the study is to validate IPM kinetically for transfemoral prosthesis users. Method: Amputees (n=5) and a control group (n=5) participated. During data collection, participants stood on two force plates with eyes open, eyes closed and with weight-bearing feedback. Correlation of the distance CoP-CoM and CoMacc were calculated for anteroposterior and mediolateral directions and evaluated with three-way ANOVA. Result: Results showed significant interaction effects between condition and position plus group and position in anteroposterior direction and condition and group in mediolateral direction. Conclusion: Results indicate kinetic validity of IPM for transfemoral amputees when looking at the whole body but not at the amputated side in the A/P direction. Kinetic validity of IPM for the intact leg is questionable.