Data from: Application of kilohertz-frequency block to mitigate off-target motor effects of vagus nerve stimulation
Data files
Jan 15, 2026 version files 152.99 GB
-
HFBlock_Analysis_Metadata.xlsx
24.77 KB
-
Ludwig_HFBlock_NatComm2025.zip
152.99 GB
-
README.md
5.58 KB
Abstract
Bioelectronic medicines are a growing medical device sector with numerous applications spanning neurologic, rehabilitation, cardiopulmonary, and pain applications. In many cases, therapeutic efficacy is limited by off-target activation of motor fibers which can cause unwanted, and sometimes painful, muscle activation. Owing to their large diameter and myelinated nature, motor fibers have lower activation thresholds than the small diameter fibers that are typically targeted for therapeutic applications. To address this limitation, achieving net selective activation of small diameter fibers has been an active area of neuromodulation research with numerous techniques developed to overcome the threshold differences between off-target and on-target fiber types. One such technique is the usage of high-frequency (HF) waveforms which has been previously demonstrated to block neural fibers in a fiber-size dependent manner in rodent models. Here, we investigated HF block during vagus nerve stimulation (VNS) in a large (pig) animal model while recording the neural (evoked compound action potential), muscle (electromyogram), and heart rate (electrocardiogram) responses to standard, low-frequency (LF) VNS when varying amplitudes of HF stimulation are applied. We found that HF was able to selectively block fibers, and their resultant physiologic effect, in an amplitude and fiber-size dependent manner in a large diameter mixed nerve, comparable in size and complexity to the human vagus. We characterize the block/activation thresholds needed to achieve block of the large diameter motor fibers responsible for off-target activation and demonstrate that they can be reliably blocked without significant effects on small diameter fiber conduction or physiologic outcomes. This potentially allows for much higher stimulation amplitudes to be applied in bioelectronic medicines by mitigating the issue of off-target motor activation during the application of treatment.
Description
Raw electrophysiologic and physiologic data for Nature Communications article, "Application of kilohertz-frequency block to mitigate off-target motor effects of vagus nerve stimulation" by Cheng et al. at the University of Wisconsin-Madison. For accessing/reading the data we strongly recommend downloading and using the accompanying software package (https://doi.org/10.5281/zenodo.17781109) which contains code for aligning the separate neural, muscle, and cardiac data streams based on the specific stimulation trial parameters using the stimulation indexing inherent to the TDT data format.
File/Folder Structure:
HFBlock_Analysis_Metadata.xlsx: Excel file containing experimental metadata
- Worksheets:
- Pig Info: Contains weight, sex, date info for subjects.
- metaDF: Maps raw data folder to experimental trial blocks
- IntegrationWindows: Integration windows used in analysis of muscle and neural signals.
- Channels: Subject specific channel identifiers for raw data channel located within the data recordings for that subject.
- Cuff_Distances: Relative cuff electrode distances in each subject. Can be used for calculation of conduction velocity.
Ludwig_HFBlock_NatComm2025.zip: Zip file containing raw data with subject specific sub-directories. Folder structure is organized to match the accompanying analysis code.
- Subject specific folders ((i.e. HF01, HF02, etc.,))
- TDT subfolder: Directory containing individual TDT tanks (folders) containing the raw electrophysiological data for each subject. Sub-folders correspond to individual experimental blocks or tanks (i.e. single recording which contains multiple stimulation trials) collected over the course of the day and are named using TDT's default auto-naming convention and date/time format. Each of these tanks (subfolders) contains the raw data from an experimental trial block containing the neural (StoreID: ECAP) and muscle (StoreID: EMGG) responses to a combination of LF-VNS and HF stimulation (refer to accompanying Excel spreadsheet 'metaDF' worksheet for the mapping of parameters and experimental blocks to TDT subfolder names).
- These data were recorded using TDT's Synapse software and are stored in TDT's DataTank format (https://www.tdt.com/docs/synapse/before-you-begin/#file-types). They can be opened/viewed using TDT's offline analysis tools (i.e. TDT OpenScope or OpenExplorer) or via MATLAB/Python using TDT's provided API's (https://www.tdt.com/docs/sdk/offline-data-analysis/overview/).
- Note: the .tbk, .tdx, .tev, .tin, .tnt, .tsq files are proprietary TDT data files that collectively makeup the data "Tank". They are not intended to be opened individually by other software (i.e. Excel) or without the use of TDT's software/API's (see above). The StoresListing.txt file contains metadata about each recording including stores listings (i.e. data streams), time of creation, and sampling frequency.
- _PhysData.mat: Physiology file containing heart rate and blood pressure (when collected) data that was recorded using an ADInstruments PowerLab bioamplifier and LabChart 8 software and exported in .mat format using LabChart 8's export function. A single physiology data file was used over the entire course of the day for each individual subject. Files can be opened using MATLAB or any software that can access the .mat file type. The structures contained within the .mat are simple arrays of the heart rate, blood pressure, and time synchronization signal with stimulation. Mapping between the physiologic data and specific stimulation trials from the TDT tanks is performed within the pyeCAP software package.
- We highly recommend using the accompanying pyeCAP software package for loading these files as the software package aligns the physiological data with individual electrical stimulation events. The .mat files can be loaded independently using the 'loadmat' function from the scipy.io library.
- The .mat files generated by LabChart 8 are auto-formatted by their export function and include ADI specific metadata/variables used by other ADI software for viewing/analyzing data. For further information please refer to: https://www.adinstruments.com/support/knowledge-base/using-labchart-8-matlab-python-and-other-software.
- The raw physiologic data is contained within the 'data' subfield of each .mat file with the variable mapping to the raw data contained within the 'titles' subfield of the .mat file. Variable names differed slightly across experiments, but generally followed the conventions below:
- 'EKG Raw' or 'BP Raw': The raw voltage signal recorded by the ADInstruments bio-amplifier. Volts (V)
- 'Stim Trigger': A synchronization TTL-pulse sent from the electrical stimulation system to the ADInstruments amplifier to pair stimulation with physiologic data. Volts (V)
- 'HR from EKG'/'HR ECG': The calculated heart rate using the default 'Rate' measurement setting within LabChart 8. Beats per minute (BPM)
- 'HR Data': The calculated heart rate following an artifact identification and removal process of the raw ECG/BP data. Beats per minute (BPM)
- 'HR Plotting': Smoothed heart rate data for visualization and plotting. Beats per minute (BPM)
- TDT subfolder: Directory containing individual TDT tanks (folders) containing the raw electrophysiological data for each subject. Sub-folders correspond to individual experimental blocks or tanks (i.e. single recording which contains multiple stimulation trials) collected over the course of the day and are named using TDT's default auto-naming convention and date/time format. Each of these tanks (subfolders) contains the raw data from an experimental trial block containing the neural (StoreID: ECAP) and muscle (StoreID: EMGG) responses to a combination of LF-VNS and HF stimulation (refer to accompanying Excel spreadsheet 'metaDF' worksheet for the mapping of parameters and experimental blocks to TDT subfolder names).