A New Weighted Time Window-based Method to Detect B-point in Impedance Cardiogram

TL;DR

This paper introduces an adaptive weighted time window method for detecting the B-point in Impedance Cardiogram signals, significantly improving delineation accuracy especially in noisy conditions.

Contribution

The study presents a novel weighted time window technique for B-point detection in ICG, with extensive validation on a large, openly shared dataset of over 21,000 annotated points.

Findings

Achieved >99.4% detection accuracy within ±150 ms tolerance.

Outperformed existing methods in noisy and clean ICG segments.

Largest publicly available annotated B-point dataset used for evaluation.

Abstract

Background: Impedance Cardiogram (ICG) has impressive number of applications in healthcare. However, its wider adoption is excessively limited due to the well recognized challenges in ICG delineation. We present a simple, adaptive, and efficient method for the most demanding ICG delineation task - detection of less distinct B-point that marks the onset of the left ventricular ejection. Method: The core of the new method is transformation of ICG time series by weighted time window of an ICG segment preceding the maximal ICG peak (the C-point) aiming at the B-point enhancement. The resulting Modified B-point (MB-point) is then easily delineated. To evaluate the proposed workflow, the dataset comprising 20 healthy participants and 21065 B-points are manually annotated and openly shared with the software code. To the best of our knowledge, our ICG dataset has the largest number of annotated B-points. Detector performance is evaluated on two recorded segments with distinct noises, as well as on an open dataset comprising ICG recorded in another set of 24 healthy subjects. Results: The results showed that the method was superior to the competing methods when the tolerance for B-point detection was set to +/-150 ms in all cases and for both datasets (>99.4%). Conclusions: Proposed approach based on the weighted time windows presents a promising technique for reliable ICG delineation that is instrumental for appropriate feature extraction and ICG characterization. Further customization could incorporate labeling of other pathological ICG segments, other biomedical signals, as well as additional enhancements for improving robustness to noise.