UC Davis

Recent advances have allowed robotic prostheses to become increasingly sophisticated in the movements they perform and the functions they offer. However, these devices are still perceived by users as separate machines, rather than true limb replacements. Therefore, an important next step in the evolution of robotic prosthesis is to promote the sense that they are a part of the body. Here, the concept of embodiment represents a cognitive connection between the patient and their prosthesis. Embodiment is the experience of owning and controlling your body. It emerges when our intentions, body’s actions, and their sensory outcomes align with what our brains predict we will see and feel. The objective of this study is to develop a research platform (robotic experimental setup) and protocol that allows us to explore embodiment, by studying the relationships between its three constituent components: ownership (the feeling our bodies belong to ourselves), agency (the sense of authoring our actions), and peripersonal space (the perceived space surrounding one’s body in which one can interact with their environment). Although well investigated individually, there is still little research in how these components interact to form a cohesive sense embodiment. This information is vital to designing and tuning control and sensory systems for prosthesis so that they are more closely perceived as body parts.

I have designed a research platform that strategically manipulates the sensory and motor information provided to participants during an experimental task that requires participants to operate a robotic hand. The hand is controlled by EMG signals from an armband that monitors muscle activity, as a measure of motor intentions. The participant will see the robotic hand moving and we manipulated sensory outcomes of the hand’s actions through three possible outcomes: an LED light will illuminate when the robotic hand contacts objects, a beeping tone that may play, and/or vibration motors in a haptic glove worn by participants may activate. We positioned the hand at different locations in front of the user to manipulate peripersonal space. Together this allowed us to measure user intentions, control motor actions, and manipulate sensory feedback; all the sensory motor channels involved in forming a sense of embodiment. We used multiple validated measures such as ownership and agency questionnaires, intentional binding, and proprioceptive drift to capture the formation of the senses of agency, ownership, and changes in peripersonal space. Finally, I performed preliminary testing with a small cohort of N=7 able-bodied participants. This allowed us to assess the effectiveness of the experimental design, capture the variability of the embodiment measure, identify future refinements, and prepare planned data collection activities with larger cohorts of participants. Early testing established that providing constant vibration in the haptic glove to indicate when the robotic fingertips touch an object offers stronger ownership and agency values compared to using pulsing vibration for feedback. These tests also revealed that roughly 18 grasps (repeated experimental trials) is the optimal number for the participants to perform the task, as ownership and agency values increases when the participants approaches 18 grasps and then plateau after this point.