ECG, IMUs, and Foot Plantar Pressure Signals for Gait and Health Monitoring

Citation Author(s):
MUHAMMAD
TALHA
Department of Electronic Engineering, Quaid-e-Awam University of Engineering, Science and Technology, Nawabshah, Pakistan
Maria
Kyrarini
Department of Electrical and Computer Engineering, Santa Clara University, USA
Ehsan Ali
Buriro
Department of Electronic Engineering, Quaid-e-Awam University of Engineering, Science and Technology, Nawabshah, Pakistan
Submitted by:
Muhammad Talha
Last updated:
Fri, 03/14/2025 - 11:51
DOI:
10.21227/f4jr-k711
Data Format:
License:
0
0 ratings - Please login to submit your rating.

Abstract 

This dataset was collected to support research on the screening and diagnosis of Diabetic Peripheral Neuropathy (DPN) and Cardiac Autonomic Neuropathy (CAN) using wearable sensor technology. It includes synchronized data from gait analysis and physiological signals such as electrocardiogram (ECG), heart rate variability (HRV), and inertial measurement units (IMUs) obtained from individuals with and without DPN and CAN. The data was recorded using a BioSignalsPlux Explorer Kit for ECG recording during motion, thigh and shank-mounted IMU sensors under controlled walking conditions for gait analysis and foot plantar pressure data using in-shoe pressure sensor i.e. ZNX-01. Each subject’s dataset includes relevant feature points that can help in distinguishing  DPN and CAN patients in clinical environment based on standard diagnostic assessments. This dataset can be used to develop and validate machine learning models, feature extraction algorithms, and signal analysis techniques for DPN and CAN screening and classification. The dataset is anonymized and ethically approved by the institutional review board of Peoples University of Medical and Health Sciences, Nawabshah (PUMHs). It is intended for researchers, clinicians, and developers working in biomedical signal processing, digital health, and assistive diagnostic technologies.

Instructions: 

This dataset contains extracted feature vectors derived from raw ECG, IMU, and foot plantar pressure signals. Due to the large size and privacy concerns associated with raw biomedical signal data, only the feature-level data are shared. These features are sufficient for reproducibility of analysis and machine learning modeling. Researchers who require raw data may contact the corresponding author for access under appropriate data-sharing agreements.”