Fetal Heart Rate Variability
Conditions
Keywords
fetal heart rate variability, entropy, multifractal analysis, scattering transform, sparse vector machine, acidosis
Brief summary
Background The fetal heart rate (FHR) is commonly monitored during labor to detect early fetal acidosis. FHR variability is traditionally investigated using Fourier transform, however, fetal conditions differ from adults (higher frequency, widely instable signal) modifying spectrum repartition along frequencies, making this approach inappropriate. New mathematic approach, including Multifractal analysis, adaptative multiscale complexity analysis, nearest-neighbor based wavelet entropy rate measures, scattering transform, may identify parameters associated with fetal acidosis. Modern classification system, driven by the data, based on these relevant parameters should help to dsiccriminate the fetus at risk for acidosis and help the clinician in decision making for the management and termination of labor. Analysing 20 to 30 minutes sliding windows with these different methods in a limited number of cases (foetus born with acidosis) and controls (foetus without acidosis) should lead to identify relevant parameters, which will be submitted to these classification tools. For external validation, a large database, including more than 4500 FHR recording documented with foetal outcomes, will be used to evaluate the performance of the combination of parameters identify and classification system used to discriminate foetuses at risk for short term acidosis or not, in order to give a relevant information to the clinician before acidosis occurs, to make the appropriate decision.
Interventions
None listed
Sponsors
Hospices Civils de Lyon
Study design
Observational model
COHORT
Time perspective
PROSPECTIVE
Eligibility
Sex/Gender
FEMALE
Age
18 Years to 45 Years
Healthy volunteers
No
Inclusion criteria
\--- Inclusion Criteria: * women with singleton pregnancy and fetal ECG recording using a scalp electrode, with recording lasting more than 60 minutes, with less than 20 % of missing data and less than 20 minutes between the end of recording and birth. *
Exclusion criteria
* Women with recirdin lasting less than 60 minutes, with more than 20 % of missing data, or more than 20 minutes between the end of recording and birth.
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| P-leader comes from multifractal analysis | Day 0 (during 20 to 30 minutes sliding windows during labor) | Hurst parameter and P-leader comes from multifractal analysis of fetal ECG that quantify fetal heart rate variability. |
| Hurst parameter comes from multifractal analysis | Day 0 (during 20 to 30 minutes sliding windows during labor) | Hurst parameter and P-leader comes from multifractal analysis of fetal ECG that quantify fetal heart rate variability. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Scattering transform | Day 0 (during 20 to 30 minutes sliding windows during labor) | The scattering transform is a nonlinear extension of the underlying wavelet transform used for multifractal analysis and is another way to quantify fetal heart rate variability |
| Entropy rate | Day 0 (during 20 to 30 minutes sliding windows during labor) | A k-nearest neighbor procedure yields estimates for entropy rates is another method to quantify fetal heart rate variability |
| adaptative multiscale complexity analysis | Day 0 (during 20 to 30 minutes sliding windows during labor) | THE FHR signal is separated in the lower number of lines to fullfile the whole signal. The number and longer of segment is data driven, and is a direct reflect of the signal complexity |
| Support vector machine (SVM) classification | Day 0 (during 20 to 30 minutes sliding windows during labor) | SVM permits to select a small number of relevant features from the different method of analysis previously described and to discriminate fetuses with and without acidosis |
Outcome results
None listed