A.P. Rijpma

Fetal magnetocardiography: clinical relevance and feasibility

We investigated the feasibility of a high-Tc SQUID system for fetal magnetocardiography (fetal MCG) aiming at a system without a magnetically shielded room and cooled by a cryocooler. The targeted SQUID resolution was 50 fT/√Hz (1–100 Hz). The research was performed along three lines: environmental noise suppression, cooling and low-Tc experiments. Environmental noise can be suppressed by forming second-order gradiometers from individual magnetometers. Concerning cooling, we investigated the applicability of commercially available coolers. In the low-Tc experiments, the medical relevance of fetal MCG was clearly shown. However, they also indicated that, in order to fully exploit the medical potential, the targeted resolution has to be 10 fT/√Hz. This increased resolution, in combination with the required high reliability of the sensors, will be hard to realize in high-Tc technology. This paper describes the results of the project and discusses the feasibility of a clinical system.

Interference characterisation of a commercial Joule–Thomson cooler to be used in a SQUID-based foetal heart monitor

At the University of Twente, a foetal heart monitor based on a high-TC SQUID magnetometer system is under development. The purpose of this system is to measure a foetal heart signal in a clinical environment. For cooling a first demonstrator version, a closed-cycle Joule–Thomson cooler from APD Cryogenics – the Cryotiger® – was selected. In this work, the Cryotiger is characterised with respect to three noise generating mechanisms: electromagnetic interference, mechanical vibrations and temperature fluctuations. The electromagnetic interference of the cold tip was below the resolution of our 3-axis fluxgate magnetometer (7 pT/VHz). The interference from the compressor, however, requires it to be placed 2 m or more from the sensor head in order to remain below the environmental power line disturbances. As mechanical vibrations of a magnetometer in a background field – the earth magnetic field – will result in an apparent field at the frequency of vibration, we require the rotation of the cold tip to remain below 200 μrad in the frequency band of interest (0.5–100 Hz). This was checked by applying a magnetic field to a SQUID magnetometer mounted on the coolers cold tip. In this way, rotations of 3 μrad and translations of 40 nm of the Cryotiger were measured; both at a frequency of 49 Hz. A further issue with respect to our application is the fluctuation in operating temperature. Under no-load operation, the temperature occasionally increased from around 70 K to about 82 K. With a load of roughly 2 W, a temperature of about 74 K was obtained, which increased about 2 K over a 20 day period.

Optimization of a third-order gradiometer for operation in unshielded environments

The optimum geometry of a third-order gradiometer for operation in unshielded environments is discussed. The optimization result depends on the specific signal and noise conditions. The fetal heart is considered as an example of the signal source. We optimized the gradiometer such that the signal-to-noise ratio is maximized in an averaged sense for all relevant environmental noise conditions and distances to the signal source. The resulting design consists of two second-order gradiometers that can be combined to form a third-order gradiometer in noisy environments, whereas a single second-order gradiometer can be used in low-noise environments. The gradiometer can provide the signal-to-noise ratio that allows detection of fetal heart signals in all relevant environmental noise conditions.

Highly balanced gradiometer systems based on HTS-SQUIDs for the use in magnetically unshielded environment

Two different concepts for gradiometer formation were tested applying high-temperature rf SQUIDs operated at 77 K in liquid nitrogen. All gradiometer systems are fully based on magnetometers. The first concept applies a compensating magnetometer at different positions to actively cancel the magnetic field at the location of other magnetometers. These magnetometers were arranged in an axial direction. In parts, a system of superconducting plates was used to align the relative magnetic orientation of the magnetometers. The outputs of these sensors were used to form a highly balanced electronic gradiometer. The second concept is based on electronic noise cancellation. A set of three magnetometers arranged in an axial direction was used to form an electronic second-order gradiometer. Different types of reference systems based on HTS-SQUID magnetometers and fluxgate sensors were applied to the gradiometer signal for achieving a high common mode rejection of the environmental disturbances. The performance of the different systems is demonstrated in a magnetically unshielded environment as well as in a shielded environment and the common mode rejection of homogeneous magnetic fields is measured. To demonstrate the performance of the systems, biomagnetic measurements have been performed in shielded and unshielded environments.

On Gradiometer Imbalance

We present methods to compute the imbalance in a gradiometer of arbitrary shape due to imperfections in its geometry, eddy currents induced in the radio-frequency interference shield, and screening currents induced in the modules of the superconducting quantum interference devices (SQUIDs). As an example, the methods are applied to evaluate the maximum expected initial imbalance of second- and third-order axial gradiometers in a measuring setup designed for fetal magnetocardiography. Mechanical imperfections in this specific setup appear to have the largest effect: the field imbalance is 2middot10 -2; the first-order gradient imbalance is 10-3 m; the second-order gradient imbalance is 10-4 m2. In the example, the imbalance caused by the other effects is one order smaller

Magnetic flux fluctuations due to eddy currents and thermal noise in metallic disks

We derive expressions for the magnetic flux in a circular loop due to eddy currents and thermal noise in coaxial metallic disks. The eddy currents are induced by an applied field that changes sinusoidally in time. We give expressions for the eddy current noise when the frequency of the applied field is very low as well as when it is very high. We combine these expressions to obtain one that is valid over the whole frequency range. The theoretical results agree well with experimental ones obtained by means of a superconducting quantum interference device (SQUID) magnetometer system. We also studied the flux due to thermal noise; again, the theoretical results show fair agreement with the experimental ones.

A high-Tc SQUID-based sensor head cooled by a Joule-Thomson cryocooler

The goal of the so-called FHARMON project is to develop a high-Tc SQUID-based magnetometer system for the measurement of fetal heart activity in standard clinical environments. To lower the threshold for the application of this fetal heart monitor, it should be simple to operate. It is, therefore, advantageous to replace the liquid cryogen bath by a closed-cycle refrigerator. For this purpose, we selected a mixed-gas Joule–Thomson cooler; the APD Cryotiger©. Because of its magnetic interference, the compressor of this closed-cycle cooler will be placed at a distance of ≈2 m from the actual sensor, which is an axial second order gradiometer. The gradiometer is formed by three magnetometers placed on an alumina cylinder, which is connected to the cold head of the cooler. This paper describes the sensor head in detail and reports on test experiments.

Vibration Reduction in a Set-Up of Two Split Type Stirling Cryocoolers

At the University of Twente research is in progress on performing magnetocardiography in clinical conditions. For the magnetic measurements High-Tc SQUIDs will be used, which will be cooled by commercial cryocoolers. These coolers consist of separate compressor and displacer modules. The compressors are of the dual opposed pistons type. Since the magnetic measurements are sensitive for movements of magnetic materials as well as for movements of the sensors, it is advantageous to minimise the vibrations caused by the cryocoolers. This is achieved by: Using coolers with dual opposed pistons in the compressor modules. Using two coolers and combining the two displacer modules into one rigid unit and operating them in a dual opposed manner.

Interference Characterization of Cryocoolers for a High-Tc SQUID-Based Fetal Heart Monitor

The FHARMON-project at the University of Twente aims at a high-Tc SQUID based fetal heart monitor for use in standard clinical environments. Besides the suppression of environmental magnetic noise, the cooling of this fetal heart monitor is an important issue. For maximum flexibility, we intend to apply a closed-cycle cryocooler instead of a liquid nitrogen cryostat. Because of the extreme sensitivity of SQUID magnetometers, the interference caused by the cryocooler is of major importance. This concerns electromagnetic interference (EMI), mechanical vibrations and temperature fluctuations. We have developed measuring techniques to characterize coolers in this respect. The characterization procedures were tested on a Signaal USFA 7058 Stirling cooler and a Leybold RGD210 GM-cooler. In the paper the measuring techniques are described along with interference characteristics of an APD Cryotiger and a Ricor/AirLiquide K535 Stirling cooler. Also, the impact on the design of the fetal heart monitor is considered.

Construction and tests of a heart scanner based on superconducting sensors cooled by small stirling cryocoolers

At the University of Twente, a heart scanner has been designed and constructed that uses superconducting devices (superconducting quantum interference devices (SQUIDs)) to measure the magnetic field of the heart. A key feature is the elimination of liquid cryogens by incorporating cryocoolers. In the design, two coolers are operated in counter-phase to reduce the mechanical interference. In addition to the application of ferromagnetic shields around the compressors, the magnetic cooler interference is reduced by placing the SQUID magnetometers coplanar with respect to the coolers. In this way, the cooler noise was reduced to a level below the intrinsic sensor noise: 0.16 pT/√Hz. A temperature of 60 K was realised with a cool-down time of about 2 h. The corresponding heat load to the coolers is roughly 0.9 W. Magnetocardiograms were recorded inside a magnetically shielded room.