Q&A: Investigating brain plasticity after limb amputation

Key takeaways:

• The cortical region of the brain does not reorganize after amputation of a limb.
• Machine learning is essential to develop brain-computer interfaces that restore lost sensory input.

A collaboration between scientists from the NIH and University College London found that the brain does not automatically recognize when a limb has been amputated, according to data published in Nature Neuroscience.

The researchers utilized functional MRI to compare brain activity of individuals before and after amputation to further investigate the phenomenon, called “phantom limb syndrome.” They found that circuits in the brain’s outermost region, the cortex, did not respond as such to other parts of the body such as lips and feet.

Further, the researchers employed a machine learning algorithm to try and determine if it could detect which part of an extremity the brain perceived as moving, following an amputation.

Healio spoke with the study’s first author, Hunter Schone, PhD, formerly of the NIH and currently a postdoctoral associate at Rehab Neural Engineering Labs at the University of Pittsburgh, who offered additional insight into these phenomena and the ways in which evolving technology can assist with future research.

Healio: What made this particular facet of neurology appealing and worthy of investigation?

Schone: For decades, scientists believed that when a body part is amputated, the brain’s map of the body dramatically reorganizes itself, with neighboring body parts taking over the area once represented by the missing limb. Further, cortical reorganization after amputation has been assumed to be one of the primary drivers of phantom limb pain.

Behavioral therapies using mirror boxes and virtual reality have operated on this rationale that we need to reverse reorganization or “normalize” the amputated body-parts map in the brain. However, these therapies are consistently failing to outperform placebo treatments in clinical trials, yet they remain our front-line treatments for treating phantom limb pain.

So, we’ve been at this crossroads where older scientific theories suggested the brain’s body map is “broken” after amputation, despite amputees themselves consistently reporting they still feel their missing limb. We aimed to resolve this (re)organization debate by tracking the brain’s body map for the first time both before and up to 5 years after amputation in patients preparing to undergo planned arm amputations. This work was led by myself and supervised by Prof. Tamar Makin at the University of Cambridge with additional supervision from Chris Baker at the NIH.

Healio: What makes the brain’s plasticity different in the wake of amputation compared to any other bodily or neurologic injury?

Schone: Unlike stroke or traumatic brain injury, limb amputation doesn’t damage the brain directly. It removes the brain’s sensory input from the missing limb. This creates a unique test case to test for plasticity: the brain region receiving sensory inputs is intact, but deprived of normal input signals.

After testing these patients before and after amputation, asking them to move their fingers before the amputation and phantom fingers after the amputation, we show that even though the somatosensory cortex receives lost or altered peripheral input for the amputated limb and intact somatosensation for the rest of the body, this alone is not sufficient to drive neurons to alter their functional tuning. The cortical body map remains stable, despite the amputation of a body part.

Healio: How do you envision specialists may be able to rewire the cortex to speed up its recognition of an amputated limb?

Schone: Our results suggest the cortex doesn’t rewire itself. The map of the missing limb is still there, so, in the clinic, we don’t need to develop therapies that attempt to fix any “broken” maps. Instead, we need to reprioritize research on the severed peripheral nerves: understanding how atypical firing patterns arise and their relationship with pain; how chronic pain impacts hyperexcitability of the nerve and the spinal cord; and how novel amputation procedures that reinnervate the severed nerves into a muscle or dermal graft can help stabilize the peripheral input.

Healio: How may brain-computer interfaces be improved based on your research?

Schone: There are already many brain-computer interface (BCI) systems implanted in individuals with spinal cord injuries. Like amputees, they’ve lost sensory input from their bodies, yet these systems work remarkably well for controlling cursors and robotic limbs. We can even restore somatosensation by directly stimulating their sensory-deprived hand representations. Notably, no one has ever reported feeling stimulation on the supposedly reorganized face. Our results show that these systems can expect a stable source of signals, despite altered sensory inputs, which is reassuring for future BCI development.

Healio: How do you envision AI or machine learning could impact researchers’ ability to pinpoint areas of concern in the future?

Schone: In this study, machine learning helped us directly test whether the hand map changed after amputation. We trained a decoder on patients’ brain activity before surgery and tested its ability to classify phantom finger movements years after amputation, and it worked remarkably well. More broadly, machine learning gives us unique tools to ask these fine-grained questions, and it is essential for developing next-generation BCIs to restore lost sensorimotor function for patients.

Reference:

Phantom limb study rewires our understanding of the brain. https://www.eurekalert.org/news-releases/1095412. Published Aug. 21, 2025. Accessed Aug. 26, 2025.

For more information:

Hunter Schone, PhD, can be reached at [email protected] and on LinkedIn here.