01 / The question
Can a weak rhythm be separated from a stronger one?
The project explored extraction of a fetal ECG component from abdominal recordings dominated by other signal contributions.
02 / The intuition
A mechanism worth testing.
An adaptive filter updates its estimate from the data. Its behaviour depends on what the reference signal actually contains.
03 / The work
Inside the method.
Explore each part of the approach.
01Model the mixture
Signal components and interference were considered before selecting the filtering approach.
02Adapt the filter
Least-mean-squares updates were used to estimate and reduce an interfering component.
03Examine the residual
Parameter changes and residual signals were used to study the separation behaviour.
04 / The observations
What emerged.
The notes document LMS implementation and parameter exploration for biomedical signal separation.
This account is based on recorded project notes. Original reports, figures and datasets are not embedded here.
05 / The limits
Where the evidence stops.
Without independent references, a cleaner-looking residual does not prove faithful recovery of a fetal signal. Clinical utility is not established here.
The academic foundations