NeuroStep: How smart socks could help with diabetic nerve pain

How does the calibration of the prosthesis work?

Raspopovic: The calibration of neurostimulation parameters is highly participant-specific since the perceived sensation relies on the relative position of the superficial electrodes to the underlying nerve and is sensitive to different nerve conditions. In addition, this may change over time due to neural adaptation or electrode displacement.

To address this, we carried out a personalized calibration phase to determine the optimal electrode positions and stimulation settings for each participant. The NeuroStep sock was designed to target superficial nerves in the foot and includes three electrode arrays tailored to different foot anatomies.

Using our interactive software, different combinations of electrodes were tested until a natural-feeling, localized sensation was achieved. Each electrode can be configured as cathodic, anodic, or inactive, allowing for flexible stimulation patterns. Once the optimal setup is found, we define personal thresholds for each user.

How is patient feedback from the first study integrated into the development process?

Raspopovic: Participants rated the stimulation intensity on a 0–10 scale. We adjusted parameters until a clear but non-painful 5/10 sensation was reached in the area of sensory loss. Otherwise, electrode settings were adjusted. After the electrodes were fixed, a calibration procedure was conducted to find the perceptual (minimum, 2/10 intensity) and maximum threshold (below pain, 8/10 intensity) for each nerve in each participant.

How flexible can "NeuroStep" be regarding different anatomies or degrees of diabetic neuropathy?

Raspopovic: We developed different sock and insole sizes to accommodate various foot shapes and nerve positions. NeuroStep remains effective until nerve conduction velocity (NCV) drops below ~33 m/s. Since neuropathy progresses slowly – around 0.55 m/s per year – the device can remain effective for several years.

What has been the biggest surprise in the study so far?

Raspopovic: It is possible to elicit sensations, even years after the disease onset, by stimulating the nerves that have neuropathic damage. This was previously not known. Also, when we measured the correlations from the brain by fMRI it appeared that these sensations are close-to-natural. This resulted in the pain diminishment which is an important health benefit.

What long-term neuroplastic effects are imaginable?

Raspopovic: We do believe that with long-term use of the device the pain could potentially be eliminated thanks to beneficial neuroplastic changes.

What is a realistic roadmap to broad availability?

Raspopovic: The project is patented, but there must be an industrial development to make the product robust and CE-approved. There are industrial interests and efforts in this direction. Therefore, I believe that in approximately three years we will have this device available.

In your opinion: How far away are we from a "digital skin" or an intelligent foot prosthesis that not only stimulates but also learns to adapt?

Raspopovic: This is a logical next step and involves novel AI techniques, which are becoming widely available. We are developing adaptive AI algorithms to reduce expert calibration time. These can help the system adjust to swelling or hardware issues – bringing us closer to learning prostheses. These advancements are happening today.