In physical examinations, hemodialysis access stenosis leading to dysfunction occurs at the venous anastomosis site or the outflow vein. Information from the inflow stenosis, such as blood pressure, pressure drop, and flow resistance increases, allows dysfunction screening from the stage of early clots and thrombosis to the progression of outflow stenosis. Therefore, this study proposes dysfunction screening model in experimental arteriovenous grafts (AVGs) using the fractional-order extractor (FOE) and the color relation analysis (CRA). A Sprott system was designed using an FOE to quantify the differences in transverse vibration pressures between the inflow and outflow sites of an AVG. Experimental analysis revealed that the degree of stenosis (DOS) correlated with an increase in fractional-order dynamic errors (FODEs). Exponential regression was used to fit a non-linear curve and can be used to quantify the relationship between the FODEs and DOS (R (2) = 0.8064). The specific ranges were used to evaluate the stenosis degree, such as DOS: <50, 50-80, and >80%. A CRA-based screening method was derived from the hue angle-saturation-value color model, which describes perceptual color relationships for the DOS. It has a flexibility inference manner with color visualization to represent the different stenosis degrees, which has average accuracy >90% superior to the traditional methods. This in vitro experimental study demonstrated that the proposed model can be used for dysfunction screening in stenotic AVGs.
Keywords: Arteriovenous graft (AVG); Color relation analysis (CRA); Fractional-order extractor (FOE); Transverse vibration pressure (TVP).