Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Modèles réduits statiques appliqués à l'industrie automobile

Modélisation, simulation, automobile --- Ingénierie, informatique & technologie > Ingénierie mécanique

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Normal gait, but also more complex gait modalities (e.g. slow walking and backwards walking) are important activities of daily life. Nevertheless, they require a complex interaction between neural and mechanical factors, which becomes even more apparent when this interaction is impaired. For instance, a reduced walking speed and asymmetric gait patterns are often observed after stroke. In the past, experimental studies have contributed insights in both the mechanical and neural mechanisms underlying muscle coordination patterns during complex and pathological gait. However, to determine causal relationships between neural control and muscle activity on the one hand, and the resulting motion on the other hand, simulation studies are required. Therefore this thesis used (neuro)-musculo-skeletal models and dynamic simulations to further define these causal relationships during complex forms of gait and during hemiparetic gait after stroke.In this thesis we addressed four research topics:- Identification of differences in muscle contributions to anterior-posterior (AP) and up-down (UD) center of mass (COM) acceleration during forward walking and backward walking.- Identification of contribution of muscles and gravity to the mediolateral (ML) control of the COM during very slow walking.- Identification of adaptations in muscle contribution to COM acceleration during split-belt walking (i.e. asymmetric, limping-like walking) compared to tied-belt walking.- Identification of the contribution of an increased length and velocity feedback and altered modulation patterns to hemiparetic gait impairments. The first study examined how a gait direction reversal influences the muscle contributions to COM accelerations in AP and vertical direction. Previous studies showed that leg kinematics during backward walking is very similar to the time-reversed kinematics during forward walking. Experimental studies found that in some muscles a time-reversal in electromyography (EMG) activity is present in backward walking compared to forward walking, as well as a reversal of function at joint level (i.e. from energy generator to energy absorber, and vice versa). However, a function reversal at the level of COM control has not been investigated before. In current study, simulations of forward and backward walking at 4 km/h were generated for 10 healthy subjects. A perturbation analysis determined the contribution of the individual muscles to AP and UD COM acceleration in both conditions. Our results showed indeed both time and function reversal in muscle contributions in AP direction, i.e. muscles that accelerate the COM in forward walking become decelerators of the COM during backward walking and vice versa. In vertical direction, muscle contributions only showed a time-reversal, but their function, i.e. opposing gravity, remains the same in both walking directions. Some muscles showed direction and phase specific adjustments in contributions to COM acceleration. The fact that similar muscles can be used to achieve the majority of functional and even opposing demands during both forward and backward walking supports the idea of a common spinal mechanism (e.g. central pattern generator, CPG) responsible for both forward and backward walking. For some muscles however a time-reversal is insufficient, and additional (supraspinal) modulation is required to meet direction-specific demands.The second study analyzed how COM movement is controlled during normal speed and very slow walking, and if sagittal and ML plane control is coupled. Controlling ML COM motion during gait is very important to maintain balance, especially at very slow walking speeds. Previous simulation studies already investigated the effect of walking speed on muscle contributions in both the sagittal plane and ML direction. However, these studies did not consider speeds comparable with gait speeds seen after stroke. It therefore remains unclear how the ML COM motion is controlled and how it is coupled to sagittal plane control at these very slow speeds. In our study, simulations (n=12) were generated of very slow (1 km/h) and normal (4 km/h) walking speeds. Individual contributions of muscles and gravity to COM accelerations in AP, UD and ML direction were calculated. Our results showed that most muscle contributions decrease when walking speed decreased from normal to very slow walking. This decrease coincides in both the sagittal plan and ML direction, therefore confirming a coupled control. However, gluteus medius (GMED), the prime contributor to control in ML direction, showed an uncoupling: GMED contributions to medial accelerations of the COM decreased, while contributions to sagittal plane accelerations increased. This difference was mainly due to the direction-specific changes of the acceleration potential of GMED. While GMED contributions to ML COM control decreased at slow speed, a tight balance in ML direction is still needed to counteract the increased destabilizing effect of gravity. When walking speed increases, GMED has a unique role in maintaining ML balance, while increased contributions from other muscle are important for both propulsion, support and balance.In the third study, we investigated how asymmetric gait changes muscle contributions to COM accelerations compared to symmetric gait in healthy subjects. To induce an asymmetric gait pattern in healthy subjects, split-belt walking, with two belts running at a different speed, was used. Previous experimental studies showed that EMG activity during split-belt walking was modulated compared to symmetric, tied-belt walking. However, no studies so far documented the adaptations of the slow and fast leg muscle contributions to COM accelerations during asymmetric, limping like walking in the absence of pathology. Our study generated simulations of split-belt walking (slow leg at 1 km/h, fast leg at 4 km/h) and the corresponding tied-belt conditions in 12 healthy subjects. Subsequently, a perturbation analysis determined the individual muscle contributions to AP, UD and ML accelerations of the COM. Our results showed that the different contributions in the slow and the fast leg cannot be attributed to a simple difference in speed as most contributions differed from the corresponding tied-belt condition. Instead, we found muscle-specific alterations in contributions that increased, decreased or even reversed the asymmetry expected, based on the asymmetry between the tied-belt conditions at the corresponding speeds. These modulations induce limping-like walking in healthy subjects. Moreover, most of the asymmetry patterns observed in healthy subjects agreed with reported adaptation patterns in hemiparetic subjects after stroke. We therefore conclude that these are related with the limping itself. Exceptions were the reorganizations in GMED and biceps femoris short head, which are thought to be related to concomitant deficits rather than directly to limping.In the last study, a simulation framework was developed to investigate the contribution of increased muscle spindle feedback in combination with altered feedback modulation to hemiparetic gait patterns after stroke. A range of impairments (e.g. muscle weakness, balance problems) are thought to contribute to hemiparetic gait deficits and amongst these spasticity is commonly mentioned as an underlying cause. However, the exact contribution of spasticity is often disputed. To investigate the effect of altered neural control to gait kinematics, forward simulations were generated. To this aim, the standard musculoskeletal model was extended with a neural control model. The neural control model included muscle length and velocity feedback of plantarflexors (soleus and gastrocnemius) and quadriceps (rectus femoris and vasti). We found that both increased length and velocity feedback induced gait deviations that are often reported in hemiparetic gait. Moreover the altered modulation pattern even reinforced the effects of the increased feedback, especially during swing. Overall, this work contributes to a better fundamental understanding of how COM movement is controlled during complex modalities of gait. More specifically, we acquired a better insight in the causal relationships between individual muscle contributions and the resulting motion of the COM, and this during backward walking, very slow walking and asymmetric walking in healthy subjects. Comparing our results with hemiparetic gait patterns increased the understanding to what extent mechanical and neural mechanisms contribute to hemiparetic gait. Furthermore, the implementation of a neuro-musculo-skeletal model provides opportunities for future research on how pathological neural control influences gait.

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Het leren van motorische vaardigheden is een belangrijke dimensie binnen de normale ontwikkeling van elk kind of volwassene. Heel wat dagdagelijkse activiteiten, al dan niet tijdens het uitvoeren van een beroep of sport zoals typen, autorijden, fietsen of hoogspringen vergen immers een nauwkeurige coördinatie van de onderste en/of bovenste ledematen. Het leren van deze interledematen coördinatiepatronen vereist dat de uitvoerder een representatie opbouwt van de manier waarop de juiste beweging wordt uitgevoerd om dezelfde beweging later opnieuw te kunnen uitvoeren. Je moet met andere woorden in je hersenen opslaan welke beweging je hebt geleerd en hoe je die moet uitvoeren, net zoals je leert lezen of schrijven en dat blijft onthouden. Zowel vanuit onze eigen ervaring, als vanuit wetenschappelijk onderzoek weten we dat het frequent herhalen van een beweging leidt tot verbetering van de prestatie en automatisering en dus tot het scherper stellen van deze interne bewegingsrepresentatie. Ondanks het feit dat (motorisch) leren een dagelijks fenomeen is, weten we nog steeds niet eenduidig waaruit de representatie is opgebouwd: namelijk hoe en waar een beweging wordt opgeslagen. De doelstelling van dit doctoraatsproject is daarom het bestuderen van de kenmerken van de bewegingsrepresentatie die in het centraal zenuwstelsel opgeslagen wordt na het leren van coördinatiepatronen. Een manier om een bewegingsrepresentatie te bestuderen is transfer onderzoek. Onder transfer verstaat men de positieve of negatieve invloed van het verwerven van een (motorische) taak A op het uitvoeren van deze taak in een andere context (A') enerzijds of op het leren van een nieuwe vaardigheid (B) anderzijds. Bijvoorbeeld: je leert als kind volleyballen (taak A) en de vaardigheden die je hierbij leert kan je transfereren naar beachvolleybal (A' zelfde taak, andere context) of naar tennis (taak B). Het leren van je eerste taak A kan het aanleren van andere taak A' of B bevorderen of juist verstoren. Door het aanleren van een nieuw bewegingspatroon en de transfer van deze geleerde vaardigheid naar andere ledematen, wensen we na te gaan of de reorganisatie van de hersenen het toelaat om de aangeleerde skill te generaliseren naar andere condities. Positieve transfer resultaten suggereren dat de geleerde vaardigheid A op een relatief abstracte manier wordt opgeslagen in het geheugen, waarna de representatie later terug kan aangesproken worden om de taak te sturen als die door andere ledematen wordt uitgevoerd. Multiledematen coördinatiepatronen zijn taken waarbij twee of meer ledematen met elkaar gecoördineerd moeten worden en die dus een intensief leerproces vergen. Onze resultaten tonen aan dat deze bewegingen deels effector-onafhankelijk, deels effector-specifiek wordt opgeslagen. Dit betekent dat je een bewegingspatroon wel kan transfereren van de ene lichaamszijde naar de andere: bijvoorbeeld je leert een taak met je rechterarm en transfereert die naar je linkerarm omdat je het globale bewegingspatroon onafhankelijk van de ledematen hebt opgeslagen. Anderzijds betekent dat ook dat er interferentie optreedt tussen twee gelijkaardige taken als die met dezelfde effector worden uitgevoerd. Deze negatieve transfer treedt op omdat de opgeslagen effector-specifieke representatie interfereert met de opbouw van de nieuwe taak. De bewegingsrepresentatie wordt opgeslagen in onze hersenen in een parietaal-premotorisch netwerk, met lateralisatie van de activatie in de linker hemisfeer. Een betere kennis omtrent de manier waarop een beweging wordt gerepresenteerd is nuttig om trainingsprogramma's samen te stellen voor (top)sporters, maar eveneens voor het opstellen van efficiënte revalidatieprogramma's.