Neural Noise

Integration of learned artificial sensation (ICMS) with vision during freely moving navigation

2 min read

  • Intracortical microstimulation (ICMS) elicits artificial sensation that can supplement vision during navigation.1
  • Multisensory integration of visual and proprioceptive information supports accurate movement; ICMS-evoked perceptions could supplement visual information for patients controlling neural prostheses.1
  • PNAS Neuroscience (March 2026); directly addresses sensory neuroprosthetics and core BCI/neuroprosthetics with clear path for patient applications.1
  • Study used freely moving mice implanted with a 16-channel microwire array in primary somatosensory cortex; mice navigated to randomly selected floor targets encoded by visual and/or patterned multichannel ICMS feedback.2
  • Mice received multimodal feedback from the start of training and achieved 75% proficiency on multimodal trials after approximately 1,000 training trials; they also quickly learned to use ICMS alone to locate invisible targets.2
  • Critically, ICMS-only performance matched or exceeded natural vision performance, and multimodal (ICMS + vision) performance significantly exceeded either unimodal condition — demonstrating rapid multisensory integration.2
  • Target location was encoded via patterned multichannel ICMS, enabling multi-variable spatial encoding; this demonstrates a viable algorithm strategy for encoding artificial proprioception in next-generation prostheses.2
  • Lead authors Samuel J. Senneka and Maria C. Dadarlat are from the Weldon School of Biomedical Engineering at Purdue University; the work establishes a rapid behavioral protocol for testing new ICMS encoding algorithms.3
  • The freely moving paradigm is a methodological advance over head-fixed tasks: subjects navigate in an ecologically valid context, making the integration result more directly relevant to real-world prosthetic deployment.1
  • Builds on prior ICMS navigation work in which rats learned to use goal-direction ICMS with the same proficiency as natural vision (PNAS 2019), extending the paradigm to multimodal integration and mice with multichannel arrays.1
  • Clinical significance: prosthetic limb users lose proprioceptive signals; ICMS-evoked artificial proprioception combined with residual vision could restore closed-loop sensorimotor control, reducing cognitive burden on the user.1

Footnotes

  1. https://www.pnas.org/doi/abs/10.1073/pnas.2521769123?af=R ↩ ↩2 ↩3 ↩4 ↩5 ↩6

  2. https://pubmed.ncbi.nlm.nih.gov/41758662/ ↩ ↩2 ↩3 ↩4

  3. https://engineering.purdue.edu/BME/AboutUs/News/2026/purdue-bme-researchers-dadarlat-and-senneka-published-in-pnas ↩

Graph View