Welcome to the MUSED-I dataset repository, a valuable resource for research and development in the field of Rehabilitation Engineering. This dataset was generated as part of the "REHABOTICS" project, aiming to provide interactive, cost-effective, and patient-directed rehabilitation to stroke patients through the use of Machine Learning and Robotics.

• Title: MUSED-I: Multi-Gesture Surface Electromyography (sEMG) Dataset for Stroke Rehabilitation
• Abstract: This dataset comprises raw surface electromyography (sEMG) signals recorded under the supervision of trained physiotherapists and doctors at Mayo Hospital Lahore and the National University of Sciences & Technology. The dataset includes six hand gestures: wrist flexion, wrist extension, hand close, wrist radial deviation, wrist ulnar deviation, and rest.

• Ethical Approval: Data collection was carried out following the World Medical Association's Standard Operating Procedure and Declaration of Helsinki, with approval from the Research Ethical Committee of the National University of Science and Technology (approval no: M-20221206).
• Participants: Ten healthy subjects (with no prior history of upper limb pathology) and two upper limb stroke patients participated in the study.

The dataset is organized into two main folders:

1. Healthy Subjects Data:

   • Contains three .csv files under the names: 3_DOF, 4_DOF, and 6_DOF, where DOF stands for the degree of freedom (number of gestures utilized for a given data recording setting).

2. Stroke Patients Data:

   • Contains subfolders for each stroke patient:
     • patient_1: Data for the first stroke patient.
     • patient_2: Data for the second stroke patient, who was only able to perform a subset of gestures.

• Three sessions of data collection were conducted on each working day, with each session consisting of 5 repetitions of 6 dissimilar hand gestures.
• Data recording for each gesture lasted for 5 seconds.
• A 10-minute break was provided between two consecutive sessions to prevent muscle fatigue.

You can download the dataset from the following sources:

• IEEE Dataport
• Kaggle

If you use this dataset in your research, please cite it as follows: [Muhammad Mustafa Khan , Muhammad Farhan, Ammar Shahzad, Hamza Suhaib Qarni, Asim Waris, Omer Gillani, August 5, 2023, "MUSED-I: "Multi-Gesture Surface Electromyography (sEMG) Dataset for Stroke Rehabilitation”", IEEE Dataport, doi: https://dx.doi.org/10.21227/04zq-yz45.]

I extend my sincere thanks to my group members, Muhammad Farhan, Ammar Shahzad, and Hamza Suhaib Qarni, for their unwavering and resilient support in successfully accomplishing this impactful project.

• Stroke Rehabilitation
• sEMG
• Biomedical Engineering
• Signal Processing
• Machine Learning
• Robotics