RPC-Net Dataset. Simultaneous HD-sEMG Recordings on the Forearm and angles of a 29-DOF Hand Kinematic Model

This data set refers to the two journal papers:

"Developing RPC-Net: Leveraging High-Density electromyography and Machine Learning for Improved Hand Position Estimation" (published, doi: 10.1109/TBME.2023.3346192) and "HDE-Array: Development and Validation of a New Dry Electrode Array Design to Acquire HD-sEMG for Hand Position Estimation" currently undergoing peer-review. 

The dataset includes data acquired in the two papers.

Part 1) We report the data acquisition process as defined in "Developing RPC-Net: Leveraging High-Density electromyography and Machine Learning for Improved Hand Position Estimation". Please refer to the updated figures (fig1.png and fig2.png) and to the original paper for additional information about the acquisition process:

High-Density surface EMG data:EMG was recorded on the surface of the forearm using the MEACS system, the EMG amplifier developed at LISiN (Politecnico di Torino, Turin, Italy). The system is made up of multiple Sensor Units (SU), each measuring 34 mm x 30 mm x 15 mm and sampling 32 channels at fs=2.048 kHz (192 V/V gain, 16 bit resolution, 2.4 V dynamic range). Three SUs were used, each connected to an anisotropic electrode array (2 rows and 16 columns, with 10 mm and 15 mm inter-electrode distance respectively) for a total of N=96 acquired monopolar electromyographic channels. The electrodes were arranged in 6 rows and 16 columns around the circumference of the forearm, covering approximately a third of its length (Fig. 1). The proximal row of electrodes (row 1) was positioned at 20 \% of the distance between the medial epicondyle and the pisiform bone. The reference electrode was positioned on the lateral epicondyle. The electromyographic signal was used as input for RPC-Net during the phases of training and testing.

Hand position data:Hand position data were acquired using a motion capture system (VICON Motus; VICON Motion Systems, Centennial, Oxford, UK) sampling at 100 samples/s. The setup included 12 infrared cameras (Vero v2.2). A total of Mh=21 infrared reflective markers (diameter of 6 mm) were positioned on the dominant hand of the subject, embedded in a glove. Additionally, 12 markers were placed on the arm, chest and back, resulting in M=33 markers in total (Fig. 1). The hand position data, translated to joint angles using the Inverse Kinematic Algorithm (IKA) defined below, was used both as input and to provide the target values for the training phase of RPC-Net. 

Part 2) We report the data acquisition process as defined in "HDE-Array: Development and Validation of a New Dry Electrode Array Design to Acquire HD-sEMG for Hand Position Estimation". Please refer to the original paper for additional information about the acquisition process:

High-Density surface EMG data:EMG was recorded on the surface of the dominant forearm using the MEACS system, an EMG amplifier developed at LISiN (Politecnico di Torino, Turin, Italy) [30] [31]. The system comprises multiple Sensor Units (SUs), each measuring 34 mm × 30 mm × 15 mm. The SUs sample 32 channels at 2.048 kHz (192 V/V gain, 16- bit resolution, 2.4 V dynamic range). The modular system can connect to various electrode arrays (HDE-Array in this case) HDE-Array is manufactured as a 2- dimensional polyimide array that is assembled using elastic bands (Fig. 1). Before assembling, the array is a 632 mm by 48 mm wide rectangular array. The matrix features 32 resin stiffeners positioned perpendicularly along its longest side, spaced 20 mm apart. Two electrodes are soldered on each stiffener, 10 mm apart, resulting in a total of 64 electrodes arranged in 2 rows and 32 columns. The silver electrodes are disc-shaped with a 4 mm diameter (12.57 mm2 area). The perimeter of the assembled HDE-Array is 200 mm at rest

Hand position data:Hand position data were acquired using a VICON Motus motion capture system (VICON Motion Systems, Centennial, Oxford, United Kingdom) sampling at 100 samples per second. The setup included 14 infrared cameras (Vero v2.2). Twenty-one infrared reflective markers (6 mm diameter) were positioned on the dominant hand of thesubject and embedded in a glove. Additionally, 12 markers were placed on the upper limb and trunk, resulting in 33 markers in total (Fig. 2). The hand position data, translated to joint angles using the Inverse Kinematic Algorithm (IKA) defined below, was used both as input and to provide the target values for the training phase of RPC-Net.