Ultrasonic determination of clot deposition rates in a milk-based, in-vitro procedure for thrombogenicity assessment

Thrombosis and thromboembolism remain the main problems associated with mechanical heart valves. We have devised a milk-based clotting technique to simulate in-vitro clinical incidence of thrombosis. Early results with the technique revealed good correlation between milk clot deposition and clinical thrombosis, but were limited in their ability to predict the course of clot deposition, as deposition could only be measured upon dismantling the apparatus.

Clot deposition was observed ultrasonically for both steady and pulsatile flows of a milk preparation through a cylindrical test chamber containing axisymmetric test bodies similar in shape to those used in earlier studies of thrombosis. Echo ultrasound images were recorded at regular time intervals, depicting the interface between clot deposits and flowing milk. From these images, the thickness of the clot deposit could be determined as a function of time.

Milk clot deposition on the test bodies, while observed, could not be measured accurately due to faint reflections arising from significant differences between the angle of incidence and the angle of reflection of the ultrasound beam. However, measurement was possible at the wall of the test chamber where deposition rates revealed steady growth of clot, following an initial 'lag', with growth continuing until a maximum thickness is reached. In some experiments shedding of parts of the deposit was observed. In pulsatile flow, wall clot deposition rates and final clot thickness attained were significantly lower than in steady flow.

Ultrasonic measurement of clot deposition rates is possible in our thrombogenicity assessment apparatus on surfaces perpendicular to the line of incidence of the ultrasound beam. With suitably designed viewing windows in an artificial heart, such measurements should enable the time course of clot deposition on artificial valves to be determined, with a view to identifying initial deposition sites and the dynamics of clot growth. Observations of the growth of clot on the test chamber wall in this study suggest that both the rate of deposition and nature of deposit formed are strongly influenced by fluid mechanical properties such as shear and mass transfer rates. In particular, our results appear to suggest a different structure of deposit, whose rate of deposition is relatively slow, under conditions of high shear.