Breathing Rate Asymmetry for Early Detection of Severe Preeclampsia Using Machine Learning Classifiers
Breathing Rate Asymmetry for Early Detection of Severe Preeclampsia Using Machine Learning Classifiers
No Thumbnail Available
Files
Date
2025
Authors
Ximena Gonzalez Reyes, 0009-0001-3003-0126
Hector Romero Morales, 0000-0002-9038-4591
Jenny Muñoz-Montes de Oca, 0000-0002-6376-0344
José Javier Reyes-Lagos, 0000-0001-5361-5007
Eric Alonso Abarca-Castro, 0000-0002-2029-3790
Journal Title
Journal ISSN
Volume Title
Publisher
Universidad Autónoma Metropolitana- Unidad Lerma
Abstract
Description
This dataset contains breath-to-breath (BB) interval signals in miliseconds, captured from parturient women at the "Monica Pretelini Saenz" Maternal-Perinatal Hospital in Toluca de Lerdo, State of Mexico, during 2021-2022. The dataset includes signals from two distinct groups: 210 normotensive control segments (C) and 210 severe preeclampsia segments (SP). Each original respiratory signal (RESP) was recorded for 3 minutes with a sampling frequency of 1000 Hz using a Mobi amplifier system equiped with a respiration sensor. The pre-processing steps applied to the respirograms involved band-pass filtering within 0.2 to 0.5 Hz to isolate the breath-to-breath intervals, followed by peak detection and manual selection of reliable three minute BB interval segments. Additionally, the results of asymmetric indices from Porta, Guzik, Ehlers, HDR, HAR, HNO, MFE, and multiscale versions of Porta, Guzik, and Ehlers, as well as the mean breathing rate derived from this BB interval signals were included. These data were used to train and validate the classifiers.
Keywords
MEDICINA Y CIENCIAS DE LA SALUD