Sensors dataset

Sensors dataset: A subset from the MIMIC-III Waveform Database Matched Subset provided by Aguirre et al. [1]. There were 5821 PPG and Arterial Blood Pressure (ABP) records of 1196 subjects, with a sampling rate of 125 Hz from the downloaded dataset. The authors kept only two 15 s segments spaced 5 min apart per record. We preprocessed the Sensors dataset with our data preparation pipeline, resulting in 9013 segments, which is approximately 13 hours of recording. As a result, the Sensors dataset has a medium to large number of segments and subjects have a high sample variability, a decent ratio of segments per subject (7.70), and a discrete data segmentation. Later, we applied a 5-fold Cross Validation (CV) to partition the data. [1] Aguirre, N., Grall-Maës, E., Cymberknop, L. J. & Armentano, R. L. Dataset corresponding to "blood pressure morphology assessment from photoplethysmogram and demographic information using deep learning with attention mechanism" (2021). Zenodo https://doi.org/10.5281/zenodo.4598938.