Movement artifact reduction in laser Doppler blood flowmetry : myocardial perfusion applications

Abstract: Laser Doppler perfusion monitoring (LDPM) and imaging (LDPI) enable assessment of tissue microvascular perfusion. The techniques are based on the Doppler broadening of the optical spectrum occurring when coherent laser light is scattered by moving red blood cells (RBC). However, if tissue motion not related to moving RBCs is present, artifacts arise in the derived perfusion estimate. The aim of this thesis was to develop and evaluate methods to reduce tissue motion influence on the perfusion estimate in general and for the specific purpose of enabling myocardial perfusion monitoring in the beating heart.An LDPM system, based on digital signal processing, was developed for myocardial perfusion assessment. To achieve an accurate estimate of the local microvascular perfusion, the varying myocardial tissue motion during the cardiac cycle is taken into account. By means of ECG-triggering, periods of minimum myocardial tissue motion can be pinpointed. The system and proposed methods were successfully evaluated both in an animal model (3 calves) and during coronary artery bypass grafting (CABG) on 13 humans. Animal studies showed the importance of processing during minimum tissue motion, at late diastole and/or late systole, to reduce movement artifacts. The human evaluation confirmed earlier animal findings and revealed low flow situations in the intraoperative phase. Influence of mechanical ventilation on the myocardial blood flow was found. The results justify investigation postoperative of CABG, where myocardial perfusion monitoring may give a rapid response to potential ischemia.The influence of tissue motion on LDPI was studied in model measurements and on the skin. A relatively large tissue velocity, compared to microvascular flow velocities, was needed to significantly influence the perfusion signal. Movement artifact magnitude depended on the movement direction and the surface structure. An LDPI system utilizing a polarization technique that blocks specularly reflected light is proposed. The method was evaluated in a flow model and on the skin of 12 subjects and was found to significantly reduce influence from tissue motion. Finally, a theoretical explanation of the origin of LDPI movement artifacts is proposed. Reduction of movement artifacts makes measurements more reliable and increases the potential of LDPI as a clinical tool.