Inline flow sensor for ventriculoperitoneal shunts: Experimental evaluation in swine
Eskildsen, Dane B.S.
Qin, Chuchu M.S.
Williams, Arthur Jr. B.S.
Dasgupta, Pernandu PhD
Mallet, Robert PhD
Yurvati, DO, Ph.D. ,FACOS, FICS, FAHA, Albert
MetadataShow full item record
Background: Hydrocephalus is a potentially life-threatening disorder in which cerebrospinal fluid (CSF) fails to circulate properly, causing a dangerous buildup of pressure in the cerebral ventricles and the surrounding brain tissue. It is seen most often in infants with congenital abnormalities of the CSF tract, and in patients with traumatic brain injury. The only treatment in most cases is the placement of a tubing to drain the excess CSF from the brain to the abdomen. These shunts are prone to blockage that requires invasive replacement surgery, so a reliable flow sensor is needed to detect shunt failure at its early stages. Currently available flow sensors often fail to detect blockage. Objective: The purpose of this project is to evaluate an advanced in-line electronic flow sensor capable of monitoring CSF flow over time for use in the treatment of hydrocephalus. This project evaluated the performance of this sensor in domestic swine. Methods/Materials: Ventriculo-peritoneal shunts were installed in the third cerebral ventricle of juvenile Yorkshire pigs, and routed to the peritoneal space in the abdomen. The flow sensor was positioned halfway between the cephalic and peritoneal ends of the shunt. Data were acquired on a laptop computer. Shunt flows were obtained at 30 s intervals. The sensor alternately heated the shunt fluid for 5 s and then monitored temperature decline, the rate of which was proportional to flow, for 25 s. The fluid was diverted into pre-weighed vials for 1- or 5-min to determine flow gravimetrically. At regular intervals, 5-20 ml boluses of artificial CSF were injected into the third ventricle. Flows reported by the sensor were compared to concomitant gravimetric flows by linear regression. Results: Over 4300 sensor measurements of flow were obtained in 6 experiments. The flow sensor reliably reports shunt flows up to 35 ml/min, the highest rate produced by 20 ml CSF injections. Four experiments showed strong linear correlations (r2 ³ 0.90) between gravimetric and sensor flows. The slope of the linear regression between the two flows was 1.05 ± 0.14 in the 6 experiments, indicating that the sensor accurately reported flows of up to 35 ml/min. Conclusions: The results of this experiment indicate that the flow sensor can report accurately ventriculo-peritoneal shunt flows over a wide range in a large animal model. Studies are planned to evaluate performance of chronically implanted shunts in ambulatory pigs.