Quantitative contrast-enhanced ultrasound measurement of cerebrospinal fluid flow for the diagnosis of ventricular shunt malfunction

Abstract

Object: Cerebral shunt malfunction is common but often difficult to effectively diagnose. Current methods are invasive, involve ionizing radiation, and can be costly. The authors of this study investigated the feasibility of quantitatively measuring CSF flow in a shunt catheter using contrast-enhanced ultrasound.

Methods: A syringe pump was used to push a solution of gas-filled microbubbles at specific flow rates through a shunt catheter while a high-frequency ultrasound imaging system was used to collect ultrasound images for offline processing. Displacement maps and velocity profiles were generated using a speckle-tracking method based on a cross-correlation algorithm. An additional correction factor, to account for a predictable underestimation and to adjust the measured flow rates, was calculated based on the geometry of the ultrasound imaging plane and assuming a simple model of laminar flow.

Results: The developed method was able to differentiate between physiologically relevant flow rates, including no flow and 0.006 to 0.09 ml/min, with reasonable certainty. The quantitative measurement of flow rates through the catheter using this method was determined to be in good agreement with the expected flow rate.

Conclusions: This study demonstrated that contrast-enhanced ultrasound has the potential to be used as a minimally invasive and cost-effective alternative method for outpatient shunt malfunction diagnosis.

Keywords: CSF flow; hydrocephalus; microbubbles; shunt malfunction; speckle tracking; ultrasound.

Fig. 1

Upper: Experimental setup consisting of a syringe pump to simulate flow in a shunt catheter and an ultrasound transducer for image acquisition. Lower: Representative B-mode ultrasound images of the ventricular shunt catheter with and without microbubbles (ultrasound contrast agent).

Fig. 2

A schematic representation of the ultrasound beam in the axial (z), lateral (x) and elevational (y) directions. The width of the ultrasound beam in the elevational direction is on the order of the inner diameter (d_i) of the catheter (left). The variations in the elevational width of the ultrasound beam results in an underestimation of velocity. The effect of the elevational beam width on the measured flow velocity can be accounted for by taking the average of the velocities across the elevational width of the beam in a simple laminar flow model (right). d_o = out diameter.

Fig. 3

Velocity maps for flow rates of 0.090 ml/min (A), 0.043 ml/min (B), 0.014 ml/min (C), 0.006 ml/min (D), and 0.000 ml/min (E) in a shunt catheter in water. Prior to display, a region of interest was defined based on the B-mode images such that errant velocity measurements outside the catheter are not shown.

Fig. 4

Upper: Expected velocity profiles based on modeled data where the transducer beam width was taken into account. Lower: Mean velocity profiles obtained from 5 independent measurements. Error bars indicate the standard deviation, at that particular position, of 5 independent velocity profiles.

Fig. 5

Measured flow rate versus true flow rate (as determined by flow calibration experiments). Gray region represents standard deviation of flow calibration results. Error bars represent standard deviations of measured data.

Fig. 6

Ex vivo B-mode images of a skin-on pork-belly sample with shunt catheter subcutaneously implanted. Velocity maps within catheter at input flow rates of 0.09 ml/min (A) and 0.043 ml/min (B) are shown overlaid on the B-mode images. Measured flow rates, calculated from the velocity maps, were 0.098 ml/min and 0.040 ml/min, respectively.