Senneka, S. | Search bioRxiv

bioRxiv2026

A redundant encoding algorithm for artificial sensory information speeds learning and improves multisensory-guided navigation

Intracortical microstimulation (ICMS) directly modulates cortical activity, providing an artificial sensory stream to guide accurate control of prosthetic limbs. Yet, only a fraction of reported sensations evoked by ICMS are proprioceptive (i.e., describing the position and movement of the body). Taking a learning-based approach to encoding an artificial proprioceptive signal bypasses this limitation. For example, animals trained with multi-channel ICMS in association with natural vision learn to both decode the ICMS signal to localize an invisible target and learn to integrate artificial sensation with natural sensation. The limitation of a learning-based approach, however, is that it requires training. We hypothesized that changing the algorithm used to encode artificial sensory information could both reduce learning time and improve plateau performance on ICMS-guided navigation. To test this idea, we trained eight mice on a sensory-guided navigation task: mice were required to locate a target in a training cage, the position of which was encoded by a red circle (vision), by multi-channel ICMS (ICMS trials), or by both vision and ICMS (multimodal). ICMS was encoded using one of two algorithms: sparse (fewer simultaneously stimulating electrodes) or redundant (more simultaneously stimulating electrodes). We found that redundant encoding sped learning of the ICMS signal relative to sparse encoding. Further, mice guided by redundant ICMS ran faster and completed trials in less time than when guided by sparse ICMS. Having redundant encoding also facilitated multisensory integration of ICMS with natural vision, improving success rates, path efficiency, movement speed, and movement time. We conclude that optimized ICMS encoding algorithms could overcome the current limitations of learning-based sensory encoding, facilitating learning and integration of artificial sensory information into existing sensorimotor neural circuits, paving the way to restoring both the sensory and motor streams of information flow in damaged sensorimotor systems.

neuroscience↗

bioRxiv2025

Rapid learning and integration of artificial sensation

1Humans rely on both proprioceptive and visual feedback during reaching, integrating these two sensory streams to improve movement accuracy and precision [1, 2]. Patients using a Brain-Machine interface (BMI) will similarly require artificial proprioceptive feedback in addition to vision to finely control a prosthesis [3, 4]. Intracortical microstimulation (ICMS) elicits sensory perceptions that could replace the lost proprioceptive signal. However, some learning may be required for encoding artificial sensation [5], as current technology does not give access to neurons with all of the desired encoding properties [6]. We developed a freely-moving mouse behavioral task in which to test learning and integration of artificial sensory information. Five mice were implanted with a 16-channel microwire array in primary somatosensory cortex. Mice were trained to navigate to randomly-selected targets upon the floor of a custom behavioral training cage. Target location was encoded with visual and/or patterned multi-channel ICMS feedback. Mice received multi-modal feedback from the beginning of training of the behavioral task, achieving 75% on multimodal trials after approximately 1000 training trials. Mice also quickly learned to use the ICMS signal to locate invisible targets, achieving 75% proficiency on ICMS-only trials when tested. Critically, we found that performance on multimodal trials significantly exceeded unimodal performance (vision or ICMS), demonstrating that animals rapidly learned to integrate natural vision with artificial sensation. 2 SignificanceMultisensory integration of visual and proprioceptive information facilitates accurate and precise movements. Intracortical microstimulation (ICMS) elicits perceptions that could supplement visual information for patients controlling a prostheses. Here, we developed a freely-moving mouse behavioral task to examine how ICMS can be used to encode multi-variable task-relevant information. Mice implanted with a cortical microwire array were trained to interpret patterned multi-channel ICMS to navigate to targets upon the floor of a custom behavioral training cage. Mice quickly learned to use the ICMS signal to locate invisible targets and integrated the artificial signal with natural vision, improving task performance. This protocol can be applied to efficiently develop and test algorithms to encode artificial proprioception for neural prostheses.

neuroscience↗