Intra-abdominal pressure during walking, fast walking and running: A dataset for Machine Learning

General information This dataset consists of Motion capture recordings acquired using Captury's markerless MoCap technology (The Captury GmbH, Saarbrücken, Germany) with a recording frequency of 60 Hz. A total of 211 subjects (height: 177.8 ± 9.8 cm; weight: 74.3 ± 13.9 kg; age: 24.5 ± 5.3 years; 121 male, 90 female) were assessed during walking and running activities. Each participant performed three exercises: walking, fast walking and running on a 7-meter track, traversing it four times back and forth at self-selected speeds. The average speeds recorded were 4.7 km/h for walking, 6.1 km/h for fast walking and 9.1 km/h for running. Musculosceletal modeling was conducted using the AnyBodyTM Modeling System (AnyBodyTM Technology A/S, Aalborg, Denmark, version 8.0). For this dataset, the joint angle data of 28 major physiological joints, the subjects' height and weight and the intra-abdominal pressure values were extracted for each timestep. Pre-processing Elimination of first and last frames of each recording to remove any potential irregularities at the start and end of the Motion Capture sessions. Segmentation into individual gait cycles: From heel strike right foot to heel strike right foot. Selection of suitable gait cycles and exclusion of outliers in terms of duration, muscle activity and proximo-distal hip reaction forces. Interpolation of each gait cycle to a length of 50 timesteps with cubic cpline Split into training (data of 169 subjects), test (21 subjects) and validation (21 subjects) datasets Z-score normalisation of training data, normalisation of test and validation data with same mean and standard deviation values. File descriptions x_train.csv Training dataset: Feature values with column names and index y_train.csv Training dataset: Label values with column names and index x_test.csv Test dataset: Feature values with column names and index y_test.csv Test dataset: Label values with column names and index x_val.csv Validation dataset: Feature values with column names and index y_val.csv Validation dataset: Label values with column names and index z_norm_spec.csv Mean and standard deviation of the training data. Basis for z-score normalisation of the features of all datasets z_norm_spec_label.csv Mean and standard deviation of the training data. Basis for z-score normalisation of the labels of all datasets