(Image courtesy of Popular Mechanics Magazine)

1. Introduction
In a collaboration with Children's Hospital of Pittsburgh, Carnegie Mellon University, and the Pittsbugh Life Sciences Greenhouse, Verimetra is developing catheter sensors to be used during pre-natal cardiac interventions.

2. Project Review
This project is one critical component of a multidisciplinary program with the goal of developing and implementing novel technologies that enable successful prenatal cardiac interventions (PCI) within a comprehensive Fetal Diagnosis and Treatment Program. Dramatic advances in fetal cardiovascular (CV) imaging of structure and function over the past 20 years have documented the critical role of mechanical loading conditions on the growth and remodeling of developing CV systems. For example, the obstruction of forward flow through the left ventricle (LV) due to aortic valve stenosis results in progressive growth failure of each of the LV associated structures (mitral valve, left ventricular myocardium, aortic valve) and eventually in the clinical syndrome termed Hypoplastic Left Heart Syndrome (HLHS). Prior to 1980, HLHS was universally fatal. Infants born with HLHS currently require a minimum of 3 palliative surgical procedures to allow the remaining single (right) ventricle to support both the pulmonary and systemic circulations and long-term survival for these patients into adulthood is significantly less than 50%. Congenital CV experts agree that the preservation of biventricular growth and function results in the best long-term survival of patients with congenital heart diseases, which has led to the development of prenatal interventional strategies.

To preserve LV growth and function in the setting of severe aortic valve stenosis, interventional strategies have evolved to include prenatal cardiac interventions (PCI) as a novel approach to avoid HLHS and a univentricular outcome. However, for the limited number of fetal interventions performed to date (approximately 32 world-wide) the major barriers to technical success are the lack of customized and dynamic 3D imaging technology and a limited choice of interventional catheters. Current practice for PCI includes standard 2D fetal echocardiography and the use of standard balloon catheters developed for coronary interventions. These balloon catheters currently lack catheter-tip sensors to provide accurate hemodynamic data on the effectiveness of balloon interventions (pressure or flow). In addition, to accomplish the complex task of positioning the catheters in utero in the correct location for effective valvuloplasty now requires between 1 and 3 hours in the operating room prior to completing a balloon inflation that lasts for less than 3 minutes. Only a small number of clinical centers (less than 5) offer this experimental clinical protocol for selected patients. Successful development of novel imaging and interventional technologies will improve the effectiveness and increase the availability of PCI for the fetus with severe congenital heart diseases.

The overall goal of our work was to develop sensing technologies that enhance catheter-based PCI. In this initial Phase, we developed and characterized a prototype solution for simultaneously measuring blood flow proximal and distal to a flow constriction using a customized cardiac catheter. The solution exploited proprietary Verimetra technology for thin film processing to fabricate sensors on the surface of catheters and ultrasound imaging. The goals of later Phase Work are to extend this initial solution to multiple blood velocity measurements including the estimation of volumetric flow, the addition of blood pressure measurements, and the validation of this novel technology in animal models. The resulting technology will allow post-manufacturing modifications of a wide range of existing interventional catheters that will add value to a growing market for cardiac, vascular and other minimally invasive interventions.

Children’s Hospital provided Verimetra with catheters used in coronary balloon angioplasties. These catheters are similar to the ones used for pre natal cardiac intervention. Verimetra used its semiconductor facilities to embed two structures on the distal end of the catheter: a velocity sensor and a series of bar codes, as shown in the picture below.

The velocity sensor was used to map out blood flow on both sides of the heart valve, before and after the balloon has been dilated to open the valve and restore blood flow. This should be a very good tool to measure the effects of fetal intervention or, e.g. transient balloon occlusion of vena cava, ductus or PFO in the fetus. A schematic of the experiment is shown below:

The “bar code” or “smart code” structures were tested to see if they could be used as echogenic markers to help determine the exact position of the catheter during the procedure. By seeing the ultrasound reflection of a particular subcomponent of the barcode, the doctor will know where that part of the catheter is located.

3. Velocity Measurements
Six standard cardiac balloon catheters were provided to Verimetra for processing. Two catheters were Boston Scientific NC Rangers; two were Cordis (J&J) Raptors, and two were Cordis (J&J) Ninjas. Verimetra embedded thin films thermistor structures on the catheters as described next.

3.1. Thin Film Thermistors on Catheters
Our device involves the use of a simple thermistor thin film device integrated on the outer surface of the catheter, at the distal-end of the catheter, as shown below. When multiple devices are integrated at different positions at the distal end of the catheter, the system will allow for the measurement of distributed velocity along a blood vessel, such as in the case of a partially blocked artery or in the case of a PCI, across the aortic valve.

The blood velocity sensor uses the well-known engineering principles of hot-wire and hot-film anemometry in which the dependence of a material’s electrical resistance on temperature is mapped out. A thermistor is an example of a sensor that exploits that dependence; thermistors can be used in a number of ways to infer the velocity of the fluid in which they are immersed.

3.2. Fluid velocity measurement with a single sensor
A single thermistor probe can be employed as a constant temperature anemometer as one element in a Wheatstone bridge circuit. The bridge balances the servo amplifier that drives current through the thermistor, heating it until its temperature, hence resistance, matches that of the setpoint (variable) resistor Rset. If the fluid in which the thermistor is immersed is moving, it convects heat away from the thermistor and the servo amplifier compensates by increasing the current through the thermistor. A similar approach is the constant current method in which, as the name implies, a constant current drives the thermistor. Fluid flow cools the thermistor lowering its resistance and unbalancing the Wheatstone bridge resulting in a measurable voltage. In both cases the measured electrical parameter gives an indirect measurement of fluid velocity – though not the direction in which the fluid is flowing.

A photograph of the control electronics which were designed and manufactured for these tests is shown below.

3.3. Catheter Structures
The following photographs show our sensors embedded on the proximal side and the distal side of the balloon at the tip of the catheter. Also visible is the wiring we used to bring the signal out of the experimental system.

The performance of the sensors was verified using peristaltic pumping systems at Verimetra. No absolute flow calibration was performed on each sensor because we could not anticipate the exact geometry of the animal setting during animal trials. (Absolute flow values are not critical to this work since we are interested in showing that we can detect and track changes in flow during the pre natal cardiac intervention). A typical set of calibration curves is shown below:

3.4. Animal Experiments
Experiments were performed at the Rangos Research Center of Children’s Hospital of Pittsburgh. Rats were prepared and anesthetized. Following a detailed animal protocol (not presented here), a surgeon exposed the rat’s heart by opening the animal’s thorax.

After insertion, (and verification by ultrasound), of the guide wire, the catheter was advanced sequentially into the heart, into the left ventricle, into the aortic valve, and out the myocardium at the other side of the heart.

3.5. Visualization of the Catheter
The following image shows the catheter within the rat heart’s left ventricle, (this is a Boston Scientific catheter):

The following image show the flow measurement and ultrasound images of the catheter now inside the beating rat left ventricle:

3.6. Experimental Flow Measurement Results
The real time output from the Verimetra electronics was captured by a data acquisition system running a custom program written by researchers at Children’s Hospital. All the data was taken as a function of time and we noted what the surgeon was doing as large changes in the output were observed. The following figure shows the complete data set (over 30,000 points) for one of the longer experiments:

Smaller changes were due to slight movements of the surgeon’s hands as he was attempting to keep the catheter steady in a particular position, and/or to variations in the rat’s cardiac output due to the surgery. (This movement resulted in a change of measurement due to the changing relative flow value). The figure clearly shows that the Verimetra sensor was able to track flow changes as the catheter was progressively advanced through the vasculature. This is shown in greater detail below, where the data is expanded to the time when the catheter tip advanced between the left ventricle and the ascending aorta:

4. Ultrasound Measurements using the Barcode
The following image shows one example of the barcodes which were embedded on all the catheters:

We embedded as many as five different barcodes on each catheter and their response to ultrasound signals was analyzed in saline baths, as discussed next. Along the length of a particular catheter, each barcode differed from its adjoining one by the thickness of the device.

The following image shows the echogenic response of a standard Boston Scientific Catheter:

As stated by one of the doctors at Children’s the image above shows that “we can identify the catheter itself, however, we could not identify the marker position”. The ultrasonic unit used in this test has a similar or better resolution compared to the ones that may be used for an actual prenatal valvuloplasty, which means what we have observed is a reasonable representation of image quality that we may get during valvuloplasty.

In summary the initial ultrasonic testing could not create any distinguishable images among catheter, sensor, platinum marker and wires. It most likely meant that all the metal (whether it is sensor, marker or wire) and/or plastic part of the catheter are seen as the same reflection by the ultrasonic system with a given resolution. Looking at the system’s operating frequency (13MHz), its wave length is about 26 microns traveling in water, and is even larger in plastics and metals (>150microns). Practically then, any feature size smaller than 150 microns will not be easily resolved up by the ultrasonic system. This implies, which is evident from the initial image shown above, that the ultrasonic system can distinguish between the catheter and water but could not distinguish among platinum marker, wires, and sensors since their typical thicknesses are less than 25microns.

We thus manufactured the barcodes discussed above. Results of further ultrasound tests are presented below. The following two images show the visualization of the catheter tip (left) and balloon (right) within the surgical setting. These images were generated to confirm the ability of the system to discriminate the catheter’s location.

Next, we took ultrasound images of the Verimetra barcodes. These structures were thinnest at the distal end and became progressively thicker as you moved away from the balloon and towards the proximal end of the catheter. Consequently, in the images below, Marker 1 was the thinnest and Marker 5 the thickest.

The results of these tests show conclusively that the use of thin film barcodes can be implemented on catheters. Based on the excellent echogenic response of the marker wires (top photo, left), and the increasingly strong response of our markers, one can conclude that the optimal barcode thickness, using the current thin film material, is between 20 and 80 microns, (The latter being the diameter of the sensor wires).