After a stroke, deficits in the control of standing balance contribute to a decreased quality of life, decreased functional mobility, and an increased risk of falls. Unfortunately, rehabilitation methods that have successfully reduced fall incidence in older adults (e.g., targeted strengthening exercises, perturbation training) have not yet shown the same benefits among people with chronic stroke (PwCS). This lack of success is likely due in part to these interventions not including a component that targets the somatosensory deficits that contribute to losses of balance and are quite common among PwCS. The objective of this proposal is to design a novel multisensory augmentation approach to improve the control of standing balance in PwCS. With sensory augmentation, artificial feedback provides the nervous system with information about the dynamic state of the body, which can be used to prevent losses of balance. Such methods have been used to improve balance among individuals with vestibular deficits, primarily by applying vibratory stimuli to the trunk that provide information about body sway. However, three primary barriers have prevented the effective use of this approach among PwCS. First, these methods rely on cognitive processing to interpret and respond to the novel source of feedback, while many PwCS have cognitive deficits. Second, the focus on a single source of augmented feedback does not account for the extensive variability in how PwCS respond to sensory stimulation. Finally, it is presently unclear whether sensory augmentation would be more effectively applied among PwCS as a training tool (i.e., used temporarily during rehabilitation) or as an assistive tool (i.e., a device worn during daily activities in the community). The proposed study will address each of these barriers, centered around the hypothesis that targeting somatosensory augmentation toward the feedback sources that are most useful for an individual patient will improve post-stroke balance performance. This hypothesis will be addressed through three Specific Aims. The first Specific Aim is to characterize changes in balance performance with augmented sensory feedback among PwCS. Four distinct sources of somatosensory feedback will be augmented (trunk tactile sense, hip proprioception, ankle proprioception, foot sole cutaneous sense), each of which can contribute to the control of mediolateral balance. The relative effectiveness of each type of augmentation will be quantified for individual participants, as will the ability to predict this effectiveness from brief sensory perturbations – which would be of great value in quickly identifying an appropriate augmentation approach. The second Specific Aim is to assess the feasibility and effectiveness of human-in-the-loop optimization of multisensory augmentation. Given the many sources of sensory feedback that can contribute to balance control, this use of novel optimization methods to identify the best stimulation paradig...