Search

International audience

Remote magnetic navigation offers various possibilities for medical interventions. Magnetic catheters can be wirelessly steered with high precision and accuracy through complex structures, as they are generally more dexterous and flexible than their manually steered counterparts. Position feedback is essential for many tasks. However, most commercially available systems do not integrate well with the magnetic navigation systems. As a result, fluoroscopy is still widely used in many interventions despite the known associated health risks. In this study, we propose a localization method that uses multiple Hall sensors to measure the magnetic fields produced by the magnetic navigation system and estimate the full sensor pose without the need for a separate dedicated mapping system. This makes the magnetic navigation system a 2-in-1 system that can be used for simultaneous navigation and localization of a medical tool. We perform an optimization of the sensors’ array design in simulation and investigate the influence of the magnetic fields and gradients on the localization accuracy to provide information on the minimal requirements for a magnetic navigation system for this task., IEEE Transactions on Magnetics, 59 (2), ISSN:0018-9464, ISSN:1941-0069

This paper presents a tracking system using magnetometers, possibly integrable in a deep brain stimulation (DBS) electrode. DBS is a treatment for movement disorders where the position of the implant is of prime importance. Positioning challenges during the surgery could be addressed thanks to a magnetic tracking. The system proposed in this paper, complementary to existing procedures, has been designed to bridge preoperative clinical imaging with DBS surgery, allowing the surgeon to increase his/her control on the implantation trajectory. Here the magnetic source required for tracking consists of three coils, and is experimentally mapped. This mapping has been performed with an in-house three-dimensional magnetic camera. The system demonstrates how magnetometers integrated directly at the tip of a DBS electrode, might improve treatment by monitoring the position during and after the surgery. The three-dimensional operation without line of sight has been demonstrated using a reference obtained with magnetic resonance imaging (MRI) of a simplified brain model. We observed experimentally a mean absolute error of 1.35 mm and an Euclidean error of 3.07 mm. Several areas of improvement to target errors below 1 mm are also discussed.

Rogowski coils (RC) intended for high-accuracy contactless current measurement ask for precise sensor geometry and specific conditions, such as a centered primary conductor. To release these constraints, we propose in this paper a recently patented continuous calibration system which features a reference conductor added coaxially to the primary conductor and a proportional-integral controller integrated in CMOS technology. The chip is mounted on a board with a precise shunt and a AC reference signal generator. The magnitude of this reference signal may exhibit slow variations, only its frequency needs to be accurate, which is a great advantage. The architecture of the chip is described as well as the experimental setup. Experimental results obtained with commercial flexible RC show that the sensor immunity to the conductor position, as well as the dispersion between several RC, is reduced to ?0.1 %, i.e. a 0.1 class measurement. In addition, the system may ensure an absolute calibration of the RC current transducer, avoiding any costly calibration procedure at the end of the sensor manufacturing.

This paper presents an approach to model the effect of magnetic field on the electrical behavior of MOS transistor. It is shown that an out-of-plane magnetic field ${B}$ induces a reduction of the channel conductivity, and that this reduction is proportional to the square of ${B}$ and the square of the carrier mobility. Providing an accurate model of the mobility versus the gate to source and drain to source voltages in all working conditions of the transistor, the effect of ${B}$ on the MOS transistor can be accurately modeled. Experimental data in strong inversion and linear regime are in very good agreement with the proposed model.

We present a three-dimensional magnetic camera consisting in an array of 64 CMOS integrated monolithic three-dimensional Hall-Effect sensors. The camera covers a 40x40x40 mm3 volume and performs a synchronous sampling at 100 samples per second of each of the 64 three-dimensional magnetic flux density signals. Recent results in the design of magnetic manipulation instrumentation systems for e.g. electrophysiology procedures or magnetic particle imaging have shown that an extensive use of calculations and finite element method simulations is done during the development. In practice, there is a lack of suitable configurable magnetic camera to perform experimental magnetic field mapping and verify simulation results. Magnetic field and field gradients determine respectively the force and torques applied to the manipulated object and shall therefore be precisely known for an accurate steering of the object. The new magnetic camera described in this paper has to our knowledge no commercial equivalent. It answers to the need of experimental characterization of magnetic manipulation systems and offers the designers some new experimental performance assessment capabilities. The camera is both suited for small object manipulation where the steering field reaches a magnitude of about 1 mT and for electrophysiology applications where the magnetic steering of catheters requires a magnetic flux density that typically can reach up to 150 mT. Based on Hall-Effect, the magnetic camera could be easily adapted to reach even larger dynamic ranges and could fulfill the requirements of magnetic manipulation of objects using a MRI magnetic field.

This paper presents a study on the influence of strong magnetic field on NMOS transistors' electrical characteristics. Experiments have been carried out in a small animal 7T MRI scanner, and have shown that up to 7T the influence exists but remains manageable. It is demonstrated that it depends on the transistor size, on the orientation of the chip inside the field, and on the gate-to-source voltage. A theoretical analysis in good agreement with experiments has been developed. Extrapolation to ultra-high field, i.e. above 10T, shows that at such a field the influence may be challenging, and that it can be lowered by using high overdrive voltage V gs –V th .

Resume L’imagerie par resonance magnetique (IRM) cardiaque devient un outil essentiel pour le diagnostic des pathologies cardiaques. Cette technique est tres dependante de la qualite de l’electrocardiogramme (ECG) et de la synchronisation des sequences d’acquisition IRM. Notre objectif a ete de developper un capteur ECG « intelligent » totalement compatible avec l’environnement electromagnetique de l’IRM afin de : (1) permettre une synchronisation fiable des sequences IRM en incluant le traitement des troubles du rythme et les patients ayant des ECG de faible amplitude ; et (2) permettre un monitorage continu des patients pour tous types de sequences IRM. Pour atteindre ces objectifs, des capteurs ECG et de mesure de champ magnetique en 3D ont ete developpes en technologie microelectronique, auxquels nous avons associe un capteur pour la mesure des mouvements respiratoires. Ces derniers representent correctement la respiration du patient permettant un meilleur controle de l’acquisition et de la reconstruction en IRM. Les circuits concus sont totalement amagnetiques et peuvent donc etre places directement dans le champ de vue de l’IRM. La creation d’une base de donnees ECG a ete necessaire pour mettre en œuvre et evaluer de nouvelles methodes de traitement de signal comme des techniques de separation de sources, de decomposition en ondelettes et des approches bayesiennes. Grâce a ces methodes, il a ete possible de supprimer la majorite des artefacts sur l’ECG pour un meilleur diagnostic et une tres bonne synchronisation des sequences. En revanche, les methodes de classification des troubles du rythme en IRM n’ont pas encore ete completement developpees.

This paper reports on the optimization of an Electro-Optic Voltage Transducer (EOVT) using a VHDL-AMS model. Specifically, two different voltage reconstruction schemes have been assessed: reconstruction using arctangent and using arccosine. Both schemes have been simulated for different applied voltages and phase shift errors between the two electrical signals produced by the EOVT. Despite the high complexity of the device (optical and electronic parts, digital and analogue signal processing), modeling has allowed estimating the performance of the system accurately. This would have been practically impossible through theoretical considerations alone. Results show the arctangent method to outperform arccosine. However, both methods reach the desired accuracy of 0.1% when the phase shift error remains below 6%.

Current researches on CMOS integrated Hall devices focus on the vertical Hall device (VHD) since this structure limits the performances of monolithic three-dimensional Hall probes. The well known conventional horizontal Hall device (HHD) has probably been brought closed to its optimal performances while integrated in CMOS technologies. On the contrary, the VHD which is more complex to study theoretically still asks for a better understanding in order to improve its performances that are lower than those of a HHD as it was demonstrated experimentally in many previous works. The aim of this paper is to discuss this experimentally observed difference between VHD and HHD through an analysis of 2D FEM calculations. Especially, we explain why the conventional VHD with two sensing contacts exhibits an intrinsic lower sensitivity. This lower sensitivity is responsible for the lower signal to noise ratio and the higher offset. We also introduce coefficients which allow to determine the sensitivity of a VHD depending on its geometry. Finally we explain why a measurement of the Hall voltage through four sensing contacts instead of two is mandatory to reach the maximum sensitivity.

In order to lower the short-circuit effect due to the measurement contacts, Vertical Hall Devices (VHDs) are generally designed either in bulky N-type silicon or in the deep N-well of high-voltage CMOS technologies. In this last case, VHD can benefit from on chip circuitry for offset and 1/f noise reduction, but HVCMOS remains a costly technology. Using spinning-current, HVCMOS compatible VHDs with a resolution of 76 μT rms over a 1.6-kHz bandwidth have been demonstrated. The VHD presented here is designed in the shallow N-well of a low-cost 0.35 μm standard CMOS technology. Unlike conventional VHD, its measurement contacts are located outside the sensor active area. FEM simulations and experimental results show that the new geometry suppresses the short-circuit effect and strongly reduces the intrinsic offset and noise. Thus, without any noise and offset reduction method, this new small VHD (63 μm 2 ) reaches a resolution of 79 μT rms over a (5 Hz–1.6 kHz) bandwidth, and opens the way to the integration of 3D Hall sensors in low-cost standard CMOS technologies.

This paper describes an integrated analogue instrumental chain dedicated to measuring magnetic pulses with an amplitude of a few tens or hundreds of microteslas, superimposed to a strong static magnetic field, i.e. 1.5 T. The system features a silicon Hall sensor with nested-chopping-like biasing circuitry, DC Hall voltage compensation, amplifying and multistage filtering. Low-noise performances are achieved by applying 100 kHz spinning current to the Hall device and chopping at 200 kHz to the signal conditioning blocs downstream. The microsystem achieves 20 μT resolution over a 1.6 kHz bandwidth extending from 5 Hz to 1.6 kHz under 1.5 T static field.

This paper presents an on-chip analog Proportional Integral controller for continuous calibration of contactless Rogowski coil current transducers. Experimental results show that the proposed system ensures a sensor sensitivity deviation lower than ±0.1% whatever the position of the coil around the primary conductor in which the current flows. In addition, it has the ability to ensure an absolute calibration of the sensor. Although this feature is not filled by the system prototype yet, a simple architecture improvement is proposed to solve the issue.

Rogowski coils (RC) used for high-accuracy contact-less current measurement ask for precise sensor geometry and specific conditions, such as a centered primary conductor. To relax these constraints, we propose here a recently patented continuous calibration system which features a reference conductor added coaxially to the primary conductor and a proportionalintegral controller integrated in CMOS technology. Experimental results obtained with commercial flexible and rigid RC show that the sensor immunity to the conductor position, as well as the dispersion between RC, is reduced to +/−0.1%, i.e. a 0.1 class measurement. In addition, the system may ensure an absolute calibration of the RC current transducer, avoiding any costly calibration procedure at the end of the sensor manufacturing.

The Rogowski coil is a traditional current-to-voltage transducer. To use it for high-accuracy measurement of low frequency AC current, it is necessary to have a very precise sensor geometry and very specific conditions, such as a stable temperature and a centered primary conductor in order to minimize errors. Conversely, it is also possible to compensate for these errors by using a sensor calibration. This paper presents a new continuous calibration system for Rogowski coil current transducer (RCCT). It features a reference conductor added close to the primary conductor and a calibration circuit which has been integrated in CMOS technology. First experimental results obtained with a homemade RCCT show that we drastically improve the immunity of the sensor to the primary conductor position with an error below ±2 % when the calibration is active against −8% to +4% without calibration. Limitations of the method are also discussed.

We present an electro-optic sensor for accurate AC voltage measurement in high voltage substations. This paper focuses on the novel signal processing method developed to retrieve the applied voltage from the sensor's periodic optical signals. The sensor head was presented in an earlier publication. The method is capable of extracting the sensor temperature directly from the optical signals and uses this information for temperature compensation. This enables the sensor to operate with an accuracy of ±0.2% between -40 °C and +80 °C. Compared with earlier solutions, robustness has been improved as the method can now determine the intrinsic birefringence of the optical sensing element and thereby remove a fundamental source adding to the measurement error. This allows relaxing the requirements for high quality optics.

The IEC 60044-7 and 8 standards apply to Electronic Current and Voltage Transformers (ECT and EVT) for metering and protection measurements in AC electrical networks. This paper describes a new architecture for Sensor interface to be used with dedicated primary sensors, Rogowski coil and resistive divider, for ECT and EVT designs respectively. The sensor interface features original self calibration technique. It exhibits less than 0.1% amplitude error and less than 1.5 mrad phase error for both current and voltage measurement.

A crucial part in the development of embedded systems in the industrial automation domain is their verification using simulation-based techniques. A comprehensive set of domain-specific modeling and simulation tools are available for this purpose, but they can only focus on certain aspects of a design. This paper presents a co-simulation framework for bringing them together in a structured and standardized way. The results are multi-domain virtual platforms where overall dynamic verification is possible. The advantage of multi-domain simulation models becomes clear when one recalls that embedded systems are used to interface with a physical environment. Often, they are coupled and cannot be tested in separation. Having a framework where physical and digital domains can interact in a correct and reproducible manner opens up new possibilities for embedded system developers. We demonstrate the capabilities of our framework by adding a co-simulation scheme between SystemC and the VHDL-AMS simulator SMASH, and present its application in an industrial case study for the verification of a new Rogowski Current Coil Transducer (RCCT) electronic front end architecture and its embedded controller software algorithm.

We present an electro-optic high voltage sensor with novel oil- and gas-free insulation based on capacitive electric field steering. The sensor’s accuracy is within ±0.2% in a temperature range from −40 °C to +80 °C.

International audience; Due to heart motion, cardiac MRI is made difficult and image acquisitions have to be synchronized with heart activity to suppress cardiac motion artifacts. Electrocardiogram (ECG) is the most accurate tool for this purpose, and Triggering consists in synchronizing MR sequences on the R-waves. The complex MR environment worsens ECG acquisition conditions because of the static magnetic field (Hall Effect), Radio Frequency (RF) and fast switching magnetic field gradients. Many hardware developments have been achieved so as to limit these undesirable effects [1], but gradient artifacts are still a problem and signal processing is required. Two ways of research have been followed; (a) first way consists in building a MR specific QRS detector [2], which is based on the vectocardiogram (VCG) the 3D representation of heart activity. This method does unfortunately not provide a clear ECG and is unable to process patient with low VCG amplitude. (b) Secondly a real-time gradient artifact suppression method has been designed. This method is based on adaptive signal processing [3] and achieves real-time accurate denoising which enables triggering. This method requires gradient signals information, which is a major drawback as the connection to the MR system is rarely available. In this paper a new real-time gradient artifact correction, which does not require any connection to the MR system, is presented. The magnetic field pulse signals will be provided by a new specifically MR designed 3D Hall probe, integrated on a low cost 0.35\μm CMOS technology [4, 5].

This paper presents a 3 dimensional magnetometer based on Hall effect sensors integrated without any post processing in a standard low cost 0.35 mum CMOS technology. The system is dedicated to magnetic pulses measurements under a strong static field. Two vertical Hall devices (VHD) are sensitive to the components of the magnetic field oriented in the plane of the chip, while a horizontal Hall device (HHD) is sensitive to the component of the magnetic field orthogonally oriented to the plane of the chip. 3 identical instrumental chains are integrated to perform amplification of the 3 Hall voltages. The system implements a compensation of the static magnetic field and allows to measure magnetic fields pulses with a resolution of 79 muT over a [5 Hz - 1.6 kHz] bandwidth. Pulses are in the range from hundreds of muT to tens of mT in the frequency range from 1 Hz to 10 kHz. The static field is compensated up to 1.5 T. The spatial resolution is 44 mum. The system power consumption has been optimized to 15 mW.

This paper reports on a standard CMOS integrated system for monitoring the magnetic fields in MRI environments. The sub-micron technology circuit features three horizontal hall devices and their associated electronics that form instrumental chains. Two of them are dedicated to millitesla range magnetic pulse and gradient measurement whereas the third one is for monitoring the strong static field of the MRI setup. The 0.35 mum technology prototype performs 130 muT gradient measurement with 20 muT resolution and can also map static fields as high as 1.5T.

In order to lower the short circuit effect due to the measurement contacts, vertical Hall devices (VHD) are generally designed either in bulky N-type silicon or in the deep N-well of high-voltage CMOS technologies. In this last case, VHD can benefit from on chip circuitry for offset and 1/f noise reduction, but HVCMOS remains a costly technology. Recently, using spinning-current, a HVCMOS compatible VHD with a resolution of 76 muT over a 1.6 kHz bandwidth has been demonstrated. The VHD presented here is designed in the shallow N-well of a low cost 0.35 mum standard CMOS technology. Unlike conventional VHD, its measurement contacts are located outside the sensor active area. FEM simulations and experimental results show that the new geometry suppresses the short circuit effect and strongly reduces the intrinsic offset and noise. Thus, without any noise and offset reduction method, this new small VHD (63 mum2) reaches a resolution of 79 muT over a [5 Hz - 1.6 kHz] bandwidth.

This paper presents a system for correcting ECG signals perturbed by MRI environments. The proposed system is based on a dedicated ASIC for measuring with high resolution the MRI magnetic gradients, which are the cause of the ECG artifacts. The measured data are used to calculate the coefficients of an integration compliant adaptive LMS filter. The filter removes the artifacts from the perturbed ECG. The filter has been implemented together with a QRS complex detection algorithm on an FPGA target. The promising experimental results validate the principle and open new perspectives in terms of ASIC integration of the whole system.

This paper presents an analogue front-end dedicated to magnetic field measurement that is intended to be used in a mixed signal submicron CMOS microsystem. The front-end features a Hall device and its biasing circuitry both associated to a nested chopper-like structure. This system efficiently removes 1/f noise and reduces offset. The biasing amplifier is chopped at a frequency fc in order to ensure steady ground referencing to the Hall device. This latter is operated using the spinning current technique at a frequency fs = fc/2. Sensitivities of 90 V/AT and resolutions of about 15 muT over a 1.6 kHz bandwidth extending from 5 to 1.6 kHz have been measured

An electro-thermal model of a Rogowski Coil sensor system is here described. A co-design methodology between VHDL-AMS and Finite Element Analysis (FEA) has been used for modeling the entire system. The proposed modeling strategy uses geometrical FEA to complete a time-dependent parametrical heat transfer model, which can be implemented in VHDL-AMS or in any other similar hardware description language. This is especially useful for performing simulations with the embedded signal processing electronics of the sensor. Important geometrical, environmental and inner material properties of the Rogowski Coil sensor system, which are difficult, or even impossible to simulate dynamically in a classical lumped-element model, are taken into account indirectly in the proposed model. This allows to study the cross-domain effects in the complete system.