Our vision is that every child will move and communicate successfully in order to achieve their full potential throughout life. Our mission is to use engineering principles to understand childhood movement and to discover new treatments and enabling devices that will improve motor function in children with developmental disorders of movement
Ongoing research in the Sanger Lab is divided along these four primary tracks:
(1) phenomenology of movement and movement disorders.
(2) retraining methods for motor learning
(3) computational modeling of movement
(4) electrophysiology of abnormal movement and DBS
Ongoing research in the Sanger Lab is divided along these four primary tracks:
(1) phenomenology of movement and movement disorders.
(2) retraining methods for motor learning
(3) computational modeling of movement
(4) electrophysiology of abnormal movement and DBS
Investigation of Kinematic and Muscle Pattern of Upper Arm Movements
Children with movement disorders are often characterized by unwanted postures and suboptimal execution during goal-directed tasks, in which jerky and imprecise movements are extrinsic performance features. Therefore, we aim to understand the neurophysiological underlying mechanisms that lead to poor control of multi-joint upper arm movements, such as reaching and pointing, in childhood movement disorders.
We use three-dimensional motion-capture technology to record the kinematic characteristic of movement and EMG to measure the activity of the upper arm muscles. Analyses are processed using biomechanical techniques of kinematic and EMG data in the time domain, as well as multivariate statistical models (i.e. Principal Component Analysis) and matrix factorization techniques (i.e. Non-Negative Matrix Factorization) in order to extract fundamental elements that can be used to describe the movements.
Children with movement disorders are often characterized by unwanted postures and suboptimal execution during goal-directed tasks, in which jerky and imprecise movements are extrinsic performance features. Therefore, we aim to understand the neurophysiological underlying mechanisms that lead to poor control of multi-joint upper arm movements, such as reaching and pointing, in childhood movement disorders.
We use three-dimensional motion-capture technology to record the kinematic characteristic of movement and EMG to measure the activity of the upper arm muscles. Analyses are processed using biomechanical techniques of kinematic and EMG data in the time domain, as well as multivariate statistical models (i.e. Principal Component Analysis) and matrix factorization techniques (i.e. Non-Negative Matrix Factorization) in order to extract fundamental elements that can be used to describe the movements.
Myocontrol for Children with Cerebral Palsy
For children with cerebral palsy (CP), prosthetic devices can provide mobility, manipulation, and functional communication. However, children need flexible interfaces that allow them to develop their own movements. Myocontrol, the use of muscle signals to control prosthetics, may accomplish this, however, we must provide the appropriate interface in order to extract the intended movement. We use EMG to record signals from the upper arm and test different algorithms to filter and map those signals in order for children to control external devices. We also study how subjects change their behavior in response to the feedback from the output of the external device. This study will help us understand how to determine the underlying intent of the movements of children with CP. It will also help in developing rehabilitative devices to train their muscle use.
For children with cerebral palsy (CP), prosthetic devices can provide mobility, manipulation, and functional communication. However, children need flexible interfaces that allow them to develop their own movements. Myocontrol, the use of muscle signals to control prosthetics, may accomplish this, however, we must provide the appropriate interface in order to extract the intended movement. We use EMG to record signals from the upper arm and test different algorithms to filter and map those signals in order for children to control external devices. We also study how subjects change their behavior in response to the feedback from the output of the external device. This study will help us understand how to determine the underlying intent of the movements of children with CP. It will also help in developing rehabilitative devices to train their muscle use.
Stochastic Dynamic Operators
The stochastic dynamic operator (SDO) theory has been introduced by Dr. Sanger to model the time-varying effect of neurons as multiple stochastic controllers in motor control. The theory links the dynamics of individual neurons to the dynamics of a full neuro-mechanical system in a linear probabilistic framework. In neural based motor control, there are several phenomena beyond the realm of classical control theory, which include uncertainty in neurons firing and also in the effect of firing as well as presence of non-Gaussian noise. The SDO framework is a probabilistic model that is capable of addressing these phenomena. The linearity of this framework tremendously facilitates modeling of highly nonlinear dynamics during movement. Moreover, SDO is a neural analysis technique that statistically describes the dynamic effect of a single and populations of neurons on the neural activities across levels. In order to validate this theory, we have used the neural data from a known simulated network based on Hodgkin-Huxley neural model. The results demonstrated that the SDO framework is able to capture dynamic effects of neurons and some of the network structure. Beside that as a part of a control model, the framework provides prediction of simulated EMG activities based on neural recordings. Currently, we are using spinal frog experimental data to validate the SDO framework. Additionally, we apply SDO techniques on the human brain data collected in our deep brain stimulation project to understand neural dynamics in dystonia.
The stochastic dynamic operator (SDO) theory has been introduced by Dr. Sanger to model the time-varying effect of neurons as multiple stochastic controllers in motor control. The theory links the dynamics of individual neurons to the dynamics of a full neuro-mechanical system in a linear probabilistic framework. In neural based motor control, there are several phenomena beyond the realm of classical control theory, which include uncertainty in neurons firing and also in the effect of firing as well as presence of non-Gaussian noise. The SDO framework is a probabilistic model that is capable of addressing these phenomena. The linearity of this framework tremendously facilitates modeling of highly nonlinear dynamics during movement. Moreover, SDO is a neural analysis technique that statistically describes the dynamic effect of a single and populations of neurons on the neural activities across levels. In order to validate this theory, we have used the neural data from a known simulated network based on Hodgkin-Huxley neural model. The results demonstrated that the SDO framework is able to capture dynamic effects of neurons and some of the network structure. Beside that as a part of a control model, the framework provides prediction of simulated EMG activities based on neural recordings. Currently, we are using spinal frog experimental data to validate the SDO framework. Additionally, we apply SDO techniques on the human brain data collected in our deep brain stimulation project to understand neural dynamics in dystonia.
Identifying the Mechanisms and Efficacy of Deep Brain Stimulation (DBS) in Childhood Dystonia
Deep brain stimulation (DBS) therapy is a treatment for severe childhood-onset dystonia that usually takes weeks or months to have its full effect. The therapy, which involves brain surgery to implant a “pacemaker for the brain”, can have a profound impact on dystonic symptoms for some individuals, yet for others the changes are subtler.
We are measuring brain activity in relation to DBS (using scalp EEG) to get a better understanding of how the therapy interacts with the brain and to optimize stimulation parameters for a given child. In addition, we are performing recordings during the implant surgery to record signals from brain areas that are otherwise inaccessible. Such recordings will help us get a deeper understanding of abnormalities and dystonia and may help improve DBS therapy.
Deep brain stimulation (DBS) therapy is a treatment for severe childhood-onset dystonia that usually takes weeks or months to have its full effect. The therapy, which involves brain surgery to implant a “pacemaker for the brain”, can have a profound impact on dystonic symptoms for some individuals, yet for others the changes are subtler.
We are measuring brain activity in relation to DBS (using scalp EEG) to get a better understanding of how the therapy interacts with the brain and to optimize stimulation parameters for a given child. In addition, we are performing recordings during the implant surgery to record signals from brain areas that are otherwise inaccessible. Such recordings will help us get a deeper understanding of abnormalities and dystonia and may help improve DBS therapy.