Processing of laser Doppler flowmetry signals from healthy subjects and patients with varicose veins: Information categorisation approach based on intrinsic mode functions and entropy computation

• Blood flowmetry data of healthy people and patients with varicose veins are studied.
• The goal is to categorize information content of flow data in the two populations.
• A framework based on empirical mode decomposition and entropy computation is proposed.
• The results show different dynamical patterns for different pathological states.

The diagnosis of pathologies from signal processing approaches has shown to be of importance. This can provide noninvasive information at the earliest stage. In this work, the problem of categorising – in a quantifiable manner – information content of microvascular blood flow signals recorded in healthy participants and patients with varicose veins is addressed. For this purpose, laser Doppler flowmetry (LDF) signals – that reflect microvascular blood flow – recorded both at rest and after acetylcholine (ACh) stimulation (an endothelial-dependent vasodilator) are analyzed. Each signal is processed with the empirical mode decomposition (EMD) to obtain its intrinsic mode functions (IMFs). An entropy measure of each IMFs is then computed. The results show that IMFs of LDF signals have different complexity for different physiologic/pathological states. This is true both at rest and after ACh stimulation. Thus, the proposed framework (EMD + entropy computation) may be used to gain a noninvasive understanding of LDF signals in patients with microvascular dysfunctions.