abstract
- This study introduces the Adaptive Transform, a novel mathematical tool designed to analyze the frequency spectrum of filtered signals, offering an innovative integration of adaptive filtering and Fourier-based analysis. This technique dynamically adjusts its parameters based on the characteristics of the analyzed signal. Its main advantages include increased sensitivity to specific events, reduced interference from crossterms, and improved accuracy in detecting transient events. The researchers applied this new transform to abdominal electrocardiogram (AECG) signals from pregnant women to directly obtain the fetal heart rate (FHR). Using public data from the PhysioNet abdominal and direct fetal ECG database, the Adaptive Transform was applied to the AECG signals, achieving highly accurate FHR estimation. The extracted signal exhibited a spectral peak at 2.18097 Hz, with an absolute error of 9.7 × 10-4 Hz compared to the reference direct fetal electrocardiogram (FECG). These results position the Adaptive Transform as a promising tool for robust spectral analysis in fetal monitoring applications and other signal processing domains.