MXPA99010206A

MXPA99010206A - System and method to determine the location and orientation of an indwelling medical device

Info

Publication number: MXPA99010206A
Application number: MXPA/A/1999/010206A
Authority: MX
Inventors: P Somogyi Christopher; R Haynor David; N Golden Robert
Original assignee: Lucent Medical Systems Inc
Priority date: 1997-05-08
Filing date: 1999-11-05
Publication date: 2001-05-17

Abstract

Un dispositivo para detectar la localización de un imán acoplado a un dispositivo médico residente adentro de un paciente, utiliza tres o más conjuntos de sensores magnéticos, teniendo cada uno, elementos sensores configurados en una forma conocida. Cada elemento sensor detecta la fuerza del campo magnético generado por el imán, y proporciona datos que indican la dirección del imán en un espacio tridimensional. El dispositivo utiliza ecuaciones fundamentales para electricidad y magnetismo que relacionan la fuerza del campo magnético medida y el gradiente del campo magnético con la localización y la fuerza de un dipolo magnético. El dispositivo utiliza un proceso iterativo para determinar la localización y orientación reales del imán. Una estimación inicial de la localización y orientación del imán, da como resultado la generación de valores de campo magnético predichos. Los valores de campo magnético predichos se comparan con los valores medidos reales proporcionados por los sensores magnéticos. Basándose en la diferencia entre los valores predichos y los valores medidos, el dispositivo estima una nueva localización del imán, y calcula nuevos valores de fuerza de campo magnético predichos. Este proceso de iteración continúa hasta que los valores predichos concuerdan con los valores medidos dentro de un grado deseado de tolerancia. En ese punto, la localización estimada concuerda con la localización real dentro de un grado de tolerancia previamente determinado. Un despliegue visual bidimensional proporciona una indicación de la localización del imán con respecto al alojamiento del detector. Se puede utilizar una porción indicadora de profundidad del despliegue visual para proporcionar una indicación relativa o absoluta de la profundidad del imán adentro del paciente.

Description

SYSTEM AND METHOD TO DETERMINE THE FOUNDATION AND ORIENTATION OF A RESIDENT MEDICAL DEVICE TECHNICAL FIELD This invention relates in general to a system and method for detecting the location of a medical device resident inside the body of a patient, and more specifically, to a detection device that detects the magnetic field strength generated by a magnet. associated with the resident medical device.

BACKGROUND OF THE INVENTION There are many cases in clinical medicine in. It is important to detect the location of a medical tube inside a patient. For example, when feeding tubes are placed through the mouth or nose of a patient, it is essential that the end of the feeding tube pass into the patient's stomach, and not "curl up" and remain in the esophagus If the end of the feeding tube is not properly placed inside the stomach, aspiration of the feeding solution into the patient's lungs may occur. In addition to the feeding tubes, a variety of other medical tubes require precise placement inside a patient's body, including dilator tubes to widen a structure of the esophagus, tubes to measure pressure waves in the stomach and in the esophagus of a patient suspected of having motor disorders of the esophagus, Sengstaken-Blakemore tubes in the stomach and in the esophagus of a patient to control bleeding from varicose veins in the esophagus, colonic decompression tubes in the colon a patient to assist in the relief of colon distention by gas, urological tubes in the bladder, in the ureter, or in the kidney of a patient, laser tubes inserted into the heart for transmyocardial revascularization, and vascular tubes in the heart or pulmonary arteries of a patient. Currently, the location of a medical tube inside a patient's body is routinely detected by the use of imaging equipment, such as chest x-rays or abdominals. However, this procedure requires transporting the patient to an X-ray facility, or conversely, transporting the X-ray equipment to the patient. It is both inconvenient - and expensive for the patient, and is particularly stressful in those cases in which the patient removes in a repeated and inadvertent manner a medical tube, such as a feeding tube, thus requiring reinsertion and X-rays. repeated. Previous attempts to detect the location - of medical tubes inside a patient, have only had limited success. For example, in US Pat. No. 5,099,845 to Besz et al., A transmitter is located inside a catheter, and an external receiver, tuned to the frequency of the transmitter, is used to detect localization of the catheter inside the catheter. patient. However, this approach requires an external or internal power source to drive the transmitter. • An external power source adds a significant risk associated with shock or electrocution and requires electrical connections to be made before placing the catheter inside the patient. An internal power source, such as a battery, must be relatively small, and can only provide power to the transmitter for a limited time. This precludes the long-term detection of the location of the catheter, and presents additional risks associated with placing a battery internally in a patient, such as the risk of leakage or rupture of the battery. X In addition, the transmitter is relatively complex , and requires an active electronic circuit (either internal or external to the catheter), as well as the different wires and connections necessary for its proper functioning. Finally, the signal produced by the transmitter is attenuated differently by different weaves and bone of the body. This attenuation requires adjustments in the strength of the signal and in the frequency of the transmitter, depending on the location of the catheter inside the patient's body. A further attempt to detect the location of medical tubes inside a patient is disclosed in U.S. Patent No. 4,809,713 to Grayzel. There, an electric heart pacemaker catheter is stopped in place against the inner wall of a patient's heart by attracting it between a small magnet located on the tip of the pacemaker catheter and a large magnet located on (eg, sewn in) the wall of the patient's chest. A three-dimensional compass is used with indications to determine the best location for the large magnet. The operation of the compass is supported by the torque generated by the magnetic forces between the small magnet and the tip of the magnetized compass, in order to point the compass towards the small magnet. However, this compass will simultaneously try to orient itself towards the environmental magnetic field of the Earth. Due to. this, the forces between the small magnet and the tip of the magnetized compass at distances greater than several centimeters, are not strong enough to precisely guide the compass towards the small magnet. Additionally, although the compass will help place the large magnet, the placement of the small magnet, and hence the pacemaker catheter, still requires the use of imaging equipment, such as X-rays or ultrasound.

For the above reasons, there is a need in the art for a medical tube, apparatus, and method to stop-detect the location of the medical tube inside the body of a patient, which eliminates the problems inherent in the existing techniques. The medical tube, apparatus, and method should provide detection of the medical tube at distances ranging from several centimeters to several decimeters, it should not require that the medical tube have an internal or external energy source, and should obviate the need to independently verify the placement of the medical tube with an imaging equipment.

COMPENDIUM OF THE INVENTION The present invention is incorporated in a system and method for detecting a position of a magnet associated with a resident medical device. The system includes a plurality of magnetic sensors that each generate a set of signals as a function of the strength of the magnetic field generated from the magnet, and an address from the sensor to the magnet. A processor calculates a predicate position of the magnet in a three-dimensional space, and calculates a predicted value related to the magnetic field strength of the magnet at the predicted location. The processor calculates a real value related to the strength of the magnetic field of the magnet using signals generated by the magnetic sensors, and determines the location of the magnet in the three-dimensional space based on the difference between the predicted value and the actual value. The system may also include a neural network to generate the estimated position based on the set of signals generated by the magnetic sensors. In one embodiment, the processor performs an iterative process to calculate the predicted position and the predicted value related to the magnetic field, and alters the predicted position based on the difference between the predicted value and the actual value. The iterative process continues until the predicted value and the actual value agree with each other within a previously determined tolerance. The system also includes a visual display to provide a visual display of data related to the position of the magnet in three-dimensional space. With the iterative process, the system must first generate an initial estimate. The neural network can be used to generate the initial estimate, based on the signals generated by the magnetic sensors. In one embodiment, the visual display is a two-dimensional visual display, which indicates the position of the magnet with respect to the housing. A depth indicator portion of the two-dimensional visual display provides an indication of the distance of the magnet from the housing. The visual display may include a visual indicator to assist the attendant to central the accommodation on the magnet. In one embodiment, the visual display is integral with the housing, and includes a transparent portion to allow the patient to be seen under the housing. In an alternative way, the visual display may be an external visual display electrically coupled with the measuring device. With an external visual display, the data related to the position of the magnet can be combined with an image of the internal anatomy of the patient, generated by a conventional image-forming device, such as a fluoroscope, X-ray, MRI, and the like. . The sensors themselves can be selected from a group of magnetic sensors comprising Hall effect sensors, flow gate sensors, inductive wound core sensors, squid sensors, magneto-resistive sensors, and nuclear precession sensors. The system may further include a position detection system, such as a digitizing arm, to determine the location of the measuring device. In this mode, the device can be easily moved by the assistant, the new location of the device being propoxcionada by the position detection system. Based on the position data provided by the position detection system, The calibration processor can recalibrate the system, even in the presence of the magnet. In this mode, the effects of the magnet are subtracted by calculating the contribution to the real magnetic field measured by the magnetic sensors in the new location. The calibration determines the effects of the Earth's magnetic field at the new location, based on the difference between the actual magnetic field measured by the magnetic sensors, and the contribution to the actual magnetic field resulting from the magnet. The position detection system can also be used to provide reference marks to the user. Prior to detection of the magnet, the user may indicate one or more reference mark positions using the position detection system. In the subsequent operation, as the magnet is inserted into the patient, the previously determined reference marks are displayed in the visual display, together with the data 'related to the position of the magnet. This allows the user to monitor the insertion of the catheter along the route marked by the reference marks. The magnet has a magnetic dipole moment that indicates the orientation of the magnet. The sensors can detect the magnetic dipole moment and provide a visual indication in the visual display, to indicate the orientation of the magnet. In one embodiment, each sensor comprises first, second, and third sensor elements configured in an orthogonal manner to detect the strength of the magnetic field in three dimensions corresponding to the first, second, and third sensor elements orthogonally accommodated.

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 illustrates the location of a magnet attached to the end of a medical tube placed inside the body of a human patient, using a known detection apparatus. Figure 2 illustrates the orientation of the magnetic sensors x, y, and z used in a known detection apparatus. Figure 3 is a top plan view of the detector of the present invention, illustrating a possible configuration of the magnetic sensors. Figure 4 illustrates the generation of the magnetic field force vectors using the magnetic sensor configuration of Figure 3 to determine the location of a magnet. Figure 5A is a functional block diagram of an exemplary embodiment of a system constructed in accordance with the present invention to determine the location of a magnet. Figure 5B is a functional block diagram illustrating the operation of the system of Figure 5A to display the location of a magnet in conjunction with a conventional image forming system.

Figure 5C illustrates one embodiment of the system of Figure 5A, to monitor the location of the detector system. Figure 6A illustrates the use of the system of Figure 5C to select the locations of reference marks in a patient. Figure 6B illustrates the visual display of the selected locations and the location of a magnet. - Figure 7A is a flow chart used by the system of Figure 5A to determine the location of a magnet. Figure 7B is a flow chart illustrating the automatic calibration function of the system of Figure 5A. Figure 8A illustrates one embodiment of the visual display used by the detector of Figure 3. Figure 8B is an alternative embodiment of the indicator used with the detector of Figure 3. Figure 8C is still another alternative modality of visual display used with the detector of Figure 3. Figure 8D is still another alternative embodiment of the visual display of the detector of Figure 3, with a depth indicator, indicating the distance of the magnet from the detector. -, - Figure 9 is a graph illustrating the results of the clinical test of the system of Figure 5A.

DETAILED DESCRIPTION OF THE INVENTION The present invention provides a medical tube, apparatus, and method for detecting the location of the medical tube inside the body of a patient. As used herein, the term "medical tube" means any type of tube or device that can be inserted into a patient's body, including (but not limited to) catheters, guide wires and medical instruments. For example, catheters include articles such as feeding tubes, urinary catheters, guide wires, and dilator catheters, as well as nasogastric tubes, endotracheal tubes, stomach pump tubes, wound drainage tubes, rectal tubes, vascular tubes, tubes of Sengstaken-Blakemore, colonic decompression tubes, pH catheters, motility catheters, and urological tubes. Guide wires are often used to guide or position dilators and other medical tubes. Medical instruments include laser devices, endoscopes, and colonoscopes. In summary, the location of any foreign object within a patient's body is a device suitable for detection by the present invention, and is encompassed within the term "medical tube". The present invention detects the location of the medical tube by detecting the magnetic field produced by a permanent magnet associated with the medical tube. As used herein, the term "associated with" means permanently fixed, removably attached, or in close proximity to, the medical tube. In one embodiment, such as a feeding tube, the magnet is associated with the end of the medical tube. In another modality, such as in a tube of Sengstaken-Blakemore, the magnet is associated with the medical tube in a location above the gastric balloon. Preferably, the magnet is a rare, cylindrical, rotationally attached rare earth magnet. Suitable magnets include rare earth magnets, such as samarium-cobalt, and neodymium-iron-boron, both of which generate high field strengths per unit volume. Although magnets that generate a high field force for their size are preferred, weaker magnets, such as Alnico or ceramic, can also be used. Because the magnet is permanent, it does not require a power source. In accordance with the above, the magnet maintains its magnetic field indefinitely, which allows for long-term placement and detection-of medical tubes without the drawbacks associated with an internal or external power source. In particular, by eliminating the use of a power source, the undesirable electrical connections necessary for the use of a power source are eliminated. Accordingly, there is no risk of electric shock for (or possible electrocution of) the patient. In addition, the static magnetic field of the magnet passes without being attenuated through the tissue and the bones of the body. This property allows the use of the present invention to detect the medical tube at any location within the patient's body. A known technique for locating a medical tube in the body of a patient is described in U.S. Patent No. 5,425,382. Figure 1 illustrates the techniques described in U.S. Patent Number 5,425,382. A tube 10, with a permanent magnet 12 located at its tip, is inserted into the patient. In the example illustrated in Figure 1, tube 10 is a feeding tube that is inserted into the patient's nose, down the esophagus, and into the stomach. However, the system can be easily used with other types of tubes. A detection apparatus 14 is used to detect the strength of the static magnetic field of the magnet 16 at two different distances 18 and 20 while submerging in the environmental magnetic field of the Earth 22. By measuring the strength of the static magnetic field 16 a two different distances 18 and 20, the detection apparatus 14-determines the gradient of the magnetic field. As the sensing apparatus 14 moves around the patient's body, larger and smaller magnetic field gradients are indicated. The tube 10 is located by moving the detection apparatus 14 until the largest amount in the detection apparatus is indicated. - - The detection apparatus 14 described in United States Patent Number 5,425,382 uses first and second magnetic sensors 24 and 26, respectively. As described in that patent, the magnetic sensors 24 and 26 may each comprise toroidal flow gate sensors for detecting the magnetic field gradient. An alternative magnetic field gradient detector system is described in U.S. Patent Number 5,622,169. Figure 2 illustrates the configuration of the magnetic sensor described in U.S. Patent Number 5,622,169. The magnetic sensors 24 and 26 each comprise three toroidal sensor elements of orthogonally accommodated flux composite. The magnetic sensor 24 comprises the magnetic sensor elements 24x¿ 24y, and 24z which are orthogonally arranged to measure the strength of the magnetic field in three orthogonal directions, illustrated in Figure 2 by the axes x, y, and z, respectively. In a similar manner, the magnetic sensor 26 comprises the magnetic sensor elements 26x, 26y, and 26z, for measuring the strength of the magnetic field in the directions x, y, and z, respectively. Using the sensors 24 and 26, the gradient of the magnetic field in the directions x, y, and z can be determined. With the gradient measurements of X magnetic field in three directions, the location of the magnet 12 can easily be terminated (see Figure 1) using a conventional vector mathematics. The mathematical sign of the magnetic gradient indicates the direction of the dipole of the magnetic field of magnet 12. The magnet, and therefore the medical tube, is detected using a known detection device containing at least two magnetic-static field strength sensors. geometrically configured to override the detection of homogeneous magnetic fields in the environment (for example, the Earth's field), while still detecting the force gradient of the magnetic field produced by the magnet. The magnet detection apparatus illustrated in Figures 1 and 2, detects the location of the magnet based on the difference in magnetic field strength in the two sensors. However, it is possible to construct a magnetic field detection apparatus with different configurations of. sensor, to provide additional data related to the position and orientation of the magnet. The present invention relates to a technique for detecting a magnet using an array of multiple sensors, and a convergence algorithm that can precisely locate the position of the magnet in three dimensions. An example embodiment of the invention is incorporated in a detector system 100, shown in Figure 3. The detector system 100 includes a housing 102, control switches 104, such as an energy switch and an establishment switch, and a display 106. In an exemplary embodiment, the visual display 106 is a two-dimensional liquid crystal display. The visual display 106 may have an opaque background, or it may have a transparent area that allows the assistant to see the skin below the surface of the detector system 100. As will be discussed in more detail below, the ability to see the reference marks The external display of the patient aids significantly in the placement of catheters using the detector system 100. Alternatively, the visual display 106 may be an external visual display, such as a video monitor. Also mounted inside the housing 102, there are first, second, third, and fourth magnetic sensors 108, 110, 112, and 114, respectively. In a preferred embodiment, the static magnetic sensors 108-112 are spaced apart to provide maximum separation within the housing 102. In an exemplary embodiment, the magnetic sensors 108-112 are configured in a substantially planar fashion within the housing 102, and are located next to the corners of the housing. The orientation of the magnetic sensors 108-114 is illustrated in Figure 4, where the magnetic sensors 108-114 are placed in the locations Sl to S4, respectively, -about the corners of the housing 102. Although the system 100 described in FIG. Figures 3 and 4 illustrate a rectangular configuration for magnetic sensors 108-114, the principles of the present invention are easily applicable to any arrangement of multiple sensors. In accordance with the foregoing, the present invention is not limited by the specific physical configuration of the magnetic sensors. In an exemplary embodiment, each of the magnetic sensors 108-114 comprises three independent magnetic detection elements orthogonally accommodated to provide a three-dimensional measurement in the directions x, y, and z, such as that illustrated in Figure 2. The elements of detection of magnetic sensors 108-114 are aligned with respect to a common origin, such that each magnetic sensor detects the static magnetic field in the same directions x, y, and z. This allows detection of the strength of the magnetic field in a three-dimensional space by each of the magnetic sensors 108-114.- The configuration of the magnetic sensors 108-114 allows the detection of a magnet in a three-dimensional space inside the patient. That is, in addition to locating the magnet inside the patient, the detector system 100 provides depth information. The configuration of magnetic sensors 108-114 can be easily changed for a specialized application. For example, a plurality of magnetic sensors can be configured in a spherical configuration around the head of a patient, to detect the location of the magnet 120 in the brain. In addition, the magnetic detection elements do not need to be configured in an orthogonal manner. For example, the magnetic detection elements can be configured in a flat arrangement or in another convenient configuration suitable for the particular application (e.g., the spherical configuration). The only requirement for a satisfactory operation of the detector system 100 is that the detector system must have at least as many detection elements-to provide data as there are unknowns in the equations to be resolved, and that the location and orientation is known. of the magnetic detection elements. In the present case, it is desirable to detect the position and orientation of the magnet 120 in a three-dimensional space.- This results in 5 unknown parameters, which can be conveniently considered as x, y, z,? , and f where x, y, and z represent the coordinates of the magnet 120 in the three-dimensional space in relation to an origin, such as the center of the housing 102,? is the angular orientation of the magnet in the YZ plane, and f is the angular orientation of the magnet in the XY plane. In addition, the contribution of the Earth's magnetic field in the directions x, y, and z is unknown. Accordingly, the model used by the detector system 100 has 8 unknown parameters that require 8 independent measurements. In an exemplary embodiment of the detector system 100 described herein, a set of twelve magnetic detection elements is used to provide an oversampling. This results in greater reliability and accuracy, while maintaining the computing requirements at a reasonable level. The mathematical description provided hereinafter can be more easily understood with respect to a Cartesian co-ordinate system, using magnetic detection elements orthogonally accommodated in the directions x, y, and z. However, it should be clearly understood that the present invention is not limited to this configuration. Any alignment of the magnetic sensing elements with the detector system 100 can be used, provided the location and orientation of the magnetic sensors 108-114 are known. Accordingly, the present invention is not limited by the specific configuration of the magnetic sensing elements. As illustrated in Figure 4, a magnet 120 is placed in a location a. As is known in the art, the magnet 120 has a magnetic dipole which is represented by the vector m. The vector m represents the strength and orientation of the magnetic dipole. Under ideal conditions, the magnetic sensors 108-114 can measure the static magnetic field generated by the magnet 120, and determine the location of the magnet at location a with a single measurement. However, the presence of the magnetic field of the Earth, the deviated magnetic fields that may be present near the vicinity of the magnet 120, the internal noise from the magnet sensors 108-114, the internal noise generated by the electronics associated with the Magnetic sensors, such as amplifiers and the like, make it virtually impossible to perform a measurement under "ideal" conditions. In order to provide precise position information for the magnet 120 in the presence of different forms of noise, the detector system 100 uses known formulas for magnetic field strength, plus actual measurements of the sensor, as inputs for a converging estimation algorithm. to provide an accurate reading of the location and orientation of the magnet 120. The elements used to process data from the magnetic sensors 108-114 are illustrated in a functional block diagram of Figure 5A where the magnetic sensors 108-114 are they couple with the analog circuit 140. The specific shape of the analog circuit 140 depends on the specific shape of the magnetic sensors 108-114. For example, if the magnetic sensors 108-114 are orthogonally placed toroidal flow gate sensors, similar to those illustrated in Figure 2, the analog circuit 140 may include amplifiers and integrators, such as those discussed in the United States Patents. of North America Numbers 5, 425,382 and 5,622,669. In another example embodiment, magnetic sensors 108-114 comprise magneto-resistive elements, whose resistance varies with the strength of a magnetic field. Each magnetic sensor 108-114 comprises three magneto-resistive detection elements orthogonally configured to detect the static magnetic field in the directions x, y, and z respectively. However, the magnetic sensors 108-114 may be any form of magnetic sensor. Various different types of magnetic sensors can be used in the practice of the present invention, including, but not limited to, Hall effect, flow gate, inductive by wound core, squid, magneto-resistive, nuclear precession sensors. , and similar. Commercial magnetic field gradient sensors in the form of an integrated circuit can also be used with the detector system 100. Additionally, magnetic sensors 108-114 need not be identical sensor types. For example, the magnetic sensors 108-112 may be a type of sensor, while the magnetic sensor 114 may be a different type. The analog circuit 140 is designed to operate with the specific shape of the magnetic sensors 108-114. The specific design of the analog circuit 140 is well within the knowledge of one of ordinary skill in the art, and need not be described in more detail herein. The output of the analog circuit 140 is coupled with an analog-to-digital converter (ADC) 142. The analog-to-digital converter 142 converts the analog output signals from the analog circuit 140 into a digital form. The operation of the analog-to-digital converter 14-2 is well known to those of ordinary skill in the art, and will not be described in detail herein. The detector system 100 also includes a central processing unit (CPU) 146, and a memory 148. In an exemplary embodiment, the central processing unit 146 is a microprocessor, such as a Pentium or the like. The memory 148 may include both read-only memory and random access memory. The different components, such as the analog-to-digital converter 142, the central processing unit 146, the memory 148, and the visual display 106, are coupled together by a busbar system 150. As can be appreciated by the experts common in the art, the busbar system 150 illustrates a typical computer busbar system, and may carry power and control signals in addition to data. In the functional block diagram of Figure 5A, an estimation processor 152 is also illustrated. As will be described in more detail below, the estimation processor 152 performs an iterative comparison between an estimated position of the magnet 120 (see Figure 2), and a measured position of the magnet 120, based on the data derived from the magnetic sensors 108-114. The iterative process continues until the estimated position and the measured position converge, resulting in an accurate measurement of the location a (see Figure 4) of the magnet 120. It should be noted that the estimation processor 152 is preferably implemented by the computer instructions stored in the memory 148, and executed by the central processing unit 146. However, for clarity, the functional block diagram of Figure 5A illustrates the estimation processor 152 as an independent block, because it performs an independent function In an alternative way, the estimation processor 152 may be implemented by other conventional computing components, such as a digital signal processor (not shown). The detector system 100 assumes that the magnetic sensors 108-114 are sufficiently far from the location a of the magnet 120, so that the magnet can be treated as a dipole source of a point. In addition, it is assumed that the spatial variation of any extraneous magnetic fields, such as the magnetic field of the earth, is small compared to the inhomogeneity produced by the. presence of the point dipole source. However, under some circumstances, foreign sources can cause disturbances in the Earth's magnetic field, such as nearby electrical equipment, metal building structural elements, and the like. As will be discussed in detail below, the detector system 100 can be easily calibrated to compensate for these disturbances. The equations used by the estimation processor 152 are easily derived from the fundamental laws of physics related to electricity and magnetism. A static magnetic field B produced by the magnetic dipole of a force m, and located at a location a, and measured at a location s, is given by the following: 3 ((s-a) * m) (s-a) - || s-a || 2 m B (s) = _ (l) s -a where || s-aj | is a well-known module value in matrix mathematics (for example, || s-a || is a square module). It should be noted that the values a, m, s, and B are all vector values. The term "static magnetic field" is intended to describe the magnetic field generated by magnet 120, as opposed to a variable time electromagnetic field or an alternating magnetic field. The magnet 120 generates a fixed (ie, static) constant magnetic field. The strength of the magnetic field detected by the detector system 100 depends on the distance between the magnet 120 and the magnetic sensors 108-114. Those skilled in the art may appreciate that the strength of the detected magnetic field may vary as the magnet 120 moves within the patient, or as the sensing system 100 moves with respect to the magnet. However, the relative movement between the detector system 100 and the magnet 120 is not essential. The detector system 100 can easily determine the location and orientation of the magnet 120 in the three-dimensional space, inclusive when the sensing system and the magnet are not moving with respect to each other. The values from the magnetic sensors 108-114 can be used in equation (1) to determine the strength of the magnetic field B at the locations Sl -34, respectively. Changes in the magnetic field B over distance are defined as a gradient G (s) of B, which is a derivative of B with respect to s. The gradient G (s) can be represented by a 3x3 matrix derived from equation (1) and expressed as follows: (15 ((s-a) m)) (s-a) (s-a) +3 II s -aII? (s -a) m + m (s -a) + ((s -a) mm) I) G (s) \ a-a rr (2) where T is a matrix transposition, and I is a 3x3 identity matrix that has the following form: It should be noted that equation (1) could be solved directly for a value a, given the values B, m, and s. However, this calculation can be difficult to solve, and may require significant computing power. The iterative estimation process described below determines the location and orientation of the magnet 120, by estimating the location a and comparing a predicted or estimated magnetic field that would result from the magnet 120 located at the estimated location with the measured magnetic field real, as measured by the magnetic sensors 108-114. The iterative process varies the estimated location in a controlled manner, until the predicted magnetic field agrees closely with the measured magnetic field. At that point, the estimated location and orientation are in accordance with the actual location and orientation of the magnet 120. This iterative process can be performed very quickly by the detector system 100 without the need for costly computation calculations required to resolve the location to directly using the equation (1). The difference between the predicted magnetic field and the actual measured magnetic field is an error, or an error function - which can be used to quantitatively determine the location a of the magnet 120. The error function is used in the iterative process to refine the Estimated location of the magnet 120. Equation (2), which indicates the gradient G (s), is used by the estimation processor 152 (see Figure 5A), to determine the magnitude and an error direction in the estimated location. Accordingly, equation (1) is used to generate the predicted values, and equation (2) uses the error results to determine how to alter the estimated position of the magnet 120. The strength of the magnetic field B is measured in each one of the locations S ^ - S ^ by the magnetic sensors 108-114, respectively. Although only four magnetic sensors are illustrated in Figure 3 to Figure 5A, the measurement. it can be generalized to n sensors, in such a way that each of the magnetic sensors provides a measurement of B (s¿) at points B ^, where i = 1 a. The estimation processor 152 calculates the quantities? - (measurements) = B (s ^) - B (S). This calculation provides a measure of the gradient from the magnetic sensor i to the magnetic sensor j, and also cancels the effects of the Earth's magnetic field, which is a constant(that is, gradient = 0) on the magnetic sensor i and on the sensor j. The estimation processor 152 also calculates the predicted values? - • (predicted) from equation (1). The estimate for the value a is adjusted until the measured values? (Measured) and the predicted values? -. (predicted) agree as closely as possible. For example, the detector system 100 may initially assume that the location a of the magnet 120 is centered below the housing 102. Based on this estimated location, the estimation processor 152 calculates the predicted values for the strength of the magnetic field in each of the magnetic sensors 108-114, which would result if the magnet 120 were actually in the estimated location. In an exemplary embodiment, the sensing elements of each of the magnetic sensors 108-114 provide a measure of the magnetic field B in three orthogonal directions, resulting in the force values of the magnetic field B? ^, B •, and Bz¿, where i is equal to 1 to 22. In a similar way, the gradient G (s) is also calculated for each of the three orthogonal directions. The estimation processor 152 also uses the strength values of the magnetic field measured from each of the magnetic sensors 108-114, and compares? - • (predicted) with? .. (measured). Based on the difference between? -. (predicted) and? _? • (measured), the estimation processor 152 generates a new estimated location for the magnet 120 (see Figure 4). and iterates the prediction process until? - (predicted) matches closely with? - - (measured). The degree of agreement between? j • (predicted) and? • • (measured), can be measured by a cost function that includes the sum of the squares of the difference between?. • (predicted) and? _ • • (measured), and then using non-linear iterative optimization algorithms to minimize the value of the cost function. The required gradients of the cost function are calculated using equation (2) above. Many well-known different cost functions and / or optimization techniques, such as neural networks, can be used by the estimation processor 152 to achieve the desired degree of agreement between? • • (predicted) and? - - (measured). The iterative measurement process performed by the estimation processor 152 can be done in a short period of time. A typical measurement cycle is performed in fractions of a second. When the tube and the associated magnet 120 are moved inside the patient, the position and orientation of the magnet will change. However, because the measurement cycle is very short, the change in the position and orientation of the magnet will be very small during any given measurement cycle, thus facilitating the real-time tracking of the magnet as the magnet moves. magnet inside the patient, or as the housing 102 moves over the patient's surface. As discussed above, the estimation processor performs an iterative comparison between an estimated position of the magnet and a measured position of the magnet. The initial estimated location can be derived by a number of possible techniques, such as random selection, a location under the sensor element 108-114 having the strongest initial reading, or by way of example, the detector system 100 can initially estimate the location a of the magnet 1210 centered below the housing 102. However, it is possible to provide a more accurate initial estimate of the location a of the magnet 120, using a neural network 154, shown in Figure 5A. It should be noted that the neural network 154 is preferably implemented by computer instructions stored in the memory 148 and executed by the central processing unit 146. However, for clarity, the functional block diagram of Figure 5A illustrates the network neural 154 as an independent block, because it performs an independent function. In an alternative way, the neural network 154 can be implemented by other conventional computing components, such as a digital signal processor (not shown). Neural networks are capable of receiving and processing large amounts of data, and by virtue of a learning process, determine which data is most important. The operation of a neural network is generally known in the art, and therefore, will be described herein only with respect to the specific application. That is, the operation of the neural network 154 will be discussed to generate an estimate of an initial position. The neural network 154 has a learning mode and an operating mode. In the learning mode, the neural network 154 is provided with real measurement data from the magnetic sensors 108-114. Because each of the magnetic sensors 108-114 has three different detection elements, a total of 12 parameters are provided as inputs to the neural network 154. Based on the 12 parameters, the neural network 154 estimates the location and orientation of the magnet 120. Then the neural network 154 is provided with data indicating the real location and orientation of the magnet 120. This process is repeated a great number of times, in such a way that the neural network 160"learns" to precisely estimate the location and orientation of the magnet 120 based on the 12 parameters. In the present case, the learning process described above (for example, providing 12 parameters, estimating the location, and providing the actual location) was repeated 1000 times. The neural network 154 learns the best estimated position for a set of 12 parameters. It should be noted that the user of the detector system 100 does not need to operate the neural network 154 in the learning mode. Rather, the data from the process in the learning mode is provided together with the detector system 100. In normal operation, the neural network 154 is used only in the operational mode. In the operative mode, the 12 parameters from the magnetic sensors 108-114 are given to the neural network 154, which generates an initial estimate of the location and orientation of the magnet 120. Based on the experiments carried out by the inventors, the neural network 154 can provide an initial estimate of the location of the magnet 120 within about ± 2 centimeters. This precise initial estimation reduces the number of iterations required by the estimation processor 152 to precisely determine the location a of the magnet 120. It should be noted that, if the location a of the magnet 120 is sufficiently far from the detector system 100, the magnetic sensors 108 -114 will provide very low signal levels. Accordingly, the neural network 154 will not generate an initial estimate until the parameters (i.e., the 12 input signals from the magnetic sensors 108-114) are above a minimum threshold, and therefore, can provide a reliable signal. Given an accurate initial estimate ^ the estimation processor 152 can perform the iteration process described above, and determine the location a of the magnet 120 within +, 1 millimeter. Clinical studies performed using the detector system 100 have demonstrated the satisfactory operation of the detector system 100. These clinical studies are described below. The detector system 100 also includes a visual display interface 156, shown in Figure 5A, to allow the image of the magnet to be displayed in an external visual display (not shown). As will be appreciated by those skilled in the art, many of the components of the detector system 100, such as the central processing unit 146 and the memory 148, are conventional computing components. In a similar manner, the visual display interface 156 may be a conventional interface that allows the image of the detector system to be displayed in a visual PC display or another monitor, such as a live image monitor 168 (see Figure 5B). An advantage of an external visual display is that the housing 102 can remain in a fixed position with respect to the patient. In this embodiment, the four magnetic sensors 108-114 can be replaced with a large number of sensors (eg, 16 sensors) uniformly distributed throughout the housing 102, to form an array of magnetic sensors. As the magnet 120 moves relative to the housing 102, the movement is detected by three or more of the magnetic sensors, and the position of the magnet is calculated is displayed in the external visual display. In this embodiment, the user does not need to reposition the housing, but simply sees the external visual display, where the array of magnetic sensors can track the position of the magnet 120. Another advantage of the external video display is the ability to combine the image generated by the detector system 100 with the image data generated by conventional techniques. For example, Figure 5B illustrates the operation of the detector system 100 in conjunction with a fluoroscope system 160. The fluoroscope system 160 is a conventional system that includes a fluoroscopic head 162, a fluoroscopic image processor 164, and a storage system. of images including a stored image monitor 166, and the live image monitor 168. In addition, a conventional video cartridge recorder 170 can record the images generated by the fluoroscope system 160, and the images generated by the detector system 100. The operation of the fluoroscope system 160 is known in the art, and will not be described in more detail herein. The detector system 100 is fixedly attached to the fluoroscopic head 162 in a known spatial relationship. A single "snapshot" image of the patient can be obtained using the fluoroscopic system 160, and can be displayed, by way of example, on the live image monitor 168. As a catheter containing the magnet 120 is inserted ( see Figure 4) in the patient, the detector system 100 detects the location a of the magnet 120 in the manner described above, and can project the image of the magnet into the live image monitor 168, together with the patient's instant shot image. . In this way, the user can conveniently use the instant trigger fluoroscope image provided by the fluoroscope system 160, combined with the live image data provided by the detector system 100. For a satisfactory operation of this aspect of the invention, it is necessary to have an appropriate alignment between the fluoroscope system 160 and the detector system 100. This alignment, or "registration" can be performed by placing a radio-opaque marker on the patient's chest, wherein the radio-opaque marker is aligned with the corners of the detector system 100. When the fluoroscope system 160 generates the instantaneous trip image, the corners of the detector system 100 are indicated on the live image monitor 168 by virtue of the radio-opaque markers. The advantage of the image overlay using the detector system 100 is that the patient is only momentarily exposed to radiation from the fluoroscopy system 160-. Subsequently, the instant shot image is displayed with the data from the detector system 100 superimposed on top of the instant shot image. Although this process has been described with respect to the fluoroscopy system 160, those skilled in the art can appreciate that the present invention is applicable to any surgical process guided by imaging using X-rays, magnetic resonance imaging (MRI), tomography of positron emission (PET), and the like. The magnetic field of the Earth is also detected by magnetic sensors 108-114. Nevertheless, assuming that the magnetic field of the Earth is constant through the housing 102, the contribution of the Earth's magnetic field to the readings from the magnetic sensors 108-114 will be the same. By generating a differential signal between any two of the magnetic sensors 108-114, the effects of the Earth's magnetic field can be effectively canceled. However, as discussed above, there may be disturbances or inhomogeneity in the Earth's magnetic field, caused by metal elements, such as equipment, hospital layer rails, metal building structural elements, and the like. Due to the unpredictable nature of these interference elements, proper orientation of the detector system 100 requires calibration. The detector system 100 can be easily calibrated to compensate for perturbations located in the Earth's magnetic field, using a calibration processor 158, shown in FIG. 5A. It should be noted that the calibration processor 158 is preferably implemented by computation instructions stored in the memory 148, and executed by the central processing unit 146. However, for clarity, the functional block diagram of Figure 5A illustrates the calibration processor 158 as an independent block, because it performs an independent function. In an alternative way, the calibration processor 158 can be implemented by other conventional computing components, such as a digital signal processor (not shown). An initial calibration is performed before the magnet 120 is introduced into the patient. Accordingly, the initial calibration occurs outside the presence of the magnetic field generated by the magnet 120. A measurement is made using the detector system 100. Under ideal conditions, without localized perturbations in the Earth's magnetic field, the signals generated by magnetic sensors 108-114 will be the same. That is, each of the detection elements oriented in the direction x will have identical readings, while each of the detection elements oriented in the direction and will have identical readings, and each of the elements oriented in the z direction will have readings identical However, under normal operating conditions, there will be localized perturbations in the Earth's magnetic field. Under these circumstances, the signals generated by each sensor element of the magnetic sensors 108-114 will all have some different value, based on the detection of the Earth's magnetic field. The readings of any two of the magnetic sensors 108-114 can be combined differentially, which, theoretically, will cancel the Earth's magnetic field. However, due to disturbances located in the Earth's magnetic field, there may be a phase shift associated with the reading. The calibration processor 158 determines the phase shift values associated with each of the magnetic sensors, and compensates for the phase shift values during the measurement cycle. That is, the phase shift value for each of the magnetic sensors 108-114 is subtracted from the reading generated by the analog-to-digital converter 142 (see Figure 5A). Accordingly, the differential reading between any two of the magnetic sensors 108-114 will be zero before the magnet 120 is introduced. Subsequently, as the magnet 120 is introduced, the differential readings from the magnetic sensors 108-114 will have values which are not zero, due to the static magnetic field generated by the magnet 120. If the detector system 100 is stationary, as illustrated in Figure 5B, a single calibration process is sufficient to cancel out the effects of the Earth's magnetic field, including localized disturbances caused by external objects, such as metallic equipment, structural building elements, and the like. However, in certain embodiments, it is desirable to move the detector system 100 on the patient's surface. As the detector system 100 moves to a new position on the patient, perturbations located in the Earth's magnetic field can cause a degradation in the accuracy of the detector system 100, because the effects of the localized disturbances can already not be canceled completely. However, the calibration processor 158 allows continuous automatic recalibration of the detector system 100, even in the presence of the magnet 120. This is illustrated in Figure 5C, where the detector system 100 is fixedly attached to a digitizer arm 180. The digitizing arm 180 is a conventional component that allows three-dimensional movement. The digitizing arm 180 can be conveniently attached to the side of the patient's bed. In a preferred embodiment, the detector system 100 is attached to the digitizing arm, and is oriented in such a way that the three movement dimensions of the digitizing arm correspond to the x axis, the y axis, and the z axis, respectively, of the detector system 100. As the user moves the detector system 100, the digitizer arm precisely tracks the position of the detector system, and generates data indicating the position. The detector system 100 uses this position data to calculate the change in the measured magnetic field caused by the magnet 120 as the detector system 100 moves. In this way, the localized effects of the magnet 120 can be removed, the measurement being resulting indicator of the localized perturbations of the Earth's magnetic field at the new position of the detector system 100. The automatic recalibration process is particularly useful in a situation, such as a peripherally inserted central catheter (PICC), which can be inserted normally into the patient's arm, and threaded through the venous system to the heart. Using conventional technology, the surgeon would normally place marks on the patient's chest to mark the expected route over which the catheter will be inserted. Without the present invention, the surgeon must blindly insert the catheter, and verify its location using, by way of example, fluoroscopy. However, the detector system 100 allows the surgeon to track the placement of the peripherally inserted central catheter. In the previous example, the detector system 100 can be located on the patient's arm, where the peripherally inserted central catheter will initially be inserted. Following the initial calibration (in the absence of the magnet 120), the detector system 100 is calibrated and will compensate for the effects of the Earth's magnetic field, including any localized disturbances. When the magnet 120 is introduced, the detector system 100 detects and displays the location a of the magnet in the manner described above. As the surgeon inserts the centrally inserted peripheral catheter (with the attached magnet 120), it may be desirable to re-locate the detector system to thereby track the progress of the peripherally inserted central catheter. Using the digitizing arm 180, the surgeon returns to locate the detector system 100 in a new location. For example, assume that the detector system 100 - moves 15.24 centimeters in the direction y, 7.62 centimeters in the x direction, and has not moved in the z direction. Based on the new location of the detector system 100, and using the technology described above, the estimation processor 152 (see Figure 5A) can calculate the magnetic field in the new location, due to the magnet 120. Given the contribution to the magnetic field in the new location that results from the magnet 120, it is possible to subtract the effects of the magnet 120. In the absence of the magnetic field from the magnet 120, it is assumed that any remaining or "residual" magnetic field is the result of the Earth's magnetic field. The residual reading is processed in the manner described above for an initial calibration, for resetting or recalibrating the detector system 100 in this way, in order to compensate for the Earth's magnetic field, including disturbances located in the new one. location. Following this recalibration process, a measurement cycle can be started, the resulting measurement of the magnetic field being due exclusively to the presence of the magnet 120. The user can manually recalibrate the detector system 100 at any point of time. However, the advantage of the technique described above is that the detector system 100 can be recalibrated automatically on a continuous basis as the detector system 100 is used. The digitizer arm 180 provides a continuous reading of the position of the system. detector 100, and therefore, makes it possible to accurately track the location of the detector system. As the detector system 100 moves, it is constantly recalibrated to compensate for the Earth's magnetic field. In the previous example, the detector system 100 can be moved to taste to follow the movement of the peripherally inserted central catheter, as it is inserted into the heart, without worrying about external influences, such as a rail of a hospital bed. , cause a degradation in the accuracy of the measurement. Although the recalibration system has been described above with respect to digitizer arm 180, it can be appreciated that other position detection systems can also be easily used.

For example, commercial tracking systems are manufactured by Ascensión Technology and Polhemus. The system manufactured by Ascensión Technology, known as the "Bird Tracker", comprises an array of sensors that measure 6 degrees of freedom, and provide accurate measurements within 1.27 centimeters, at a distance of 1,524 meters, and provide rotating information within grade at a distance of 1,524 meters. The detection elements used in the Bird Tracker can be attached to the housing 102, and the position of the housing can be tracked using the commercial system. In a similar way, the Polhemus device, known as the "3-D Tracker", provides similar location measurements without the need for the digitizer arm 180. Another application of position tracking, which uses, for example, the digitizer arm 180, allows the surgeon to provide digitized reference marks that will be displayed in the visual display. A common surgical technique to assist in the insertion of a catheter is to place reference marks on the patient's surface, which approximate the route that will be taken by the catheter. For example, with conventional technology, the surgeon can place a series of x on the patient's chest with a marker as reference marks, to assist in the insertion of electric pacemaker leads. With the principles of the present invention, the digitizing arm 180 can be used to electronically register reference marks specified by the surgeon. This aspect of the invention is illustrated in Figure 6A, when a computer input pen 182 or other electronic input device is mounted on the digitizing arm 180. The computer pen 182 can be attached to the detector system 100, or it can be joining digitizer arm 180 in a position corresponding to, by way of example, the center of the detector system. Before inserting the catheter with the magnet 120, the surgeon can use the digitizing arm 180 and the computer pen 182, to electronically generate reference marks, illustrated in Figure 6A by a series of x. It should be noted that the computer pen 182 electronically "marks" the patient, but does not need to place real marks on the patient. In the previous example, where heart pacemaker drivers will be inserted, the surgeon can place a series of electronic reference marks from the neck to the heart, along the route where the pacemaker leads will be inserted. At each reference mark, the digitizing arm 180 registers the position marked by the surgeon. In the subsequent operation, when the catheter with the magnet 120 is inserted into the patient, the digitizing arm 180 notes the location of the magnet 120 with respect to the reference marks previously marked by the surgeon. The reference marks are shown in an external visual display 184, shown in Figure 6B, together with the position of the magnet 120, which is indicated by an arrow. When the surgeon inserts the magnet 120, the progress in the external visual display 184 is displayed, such that the magnet 120 passes along from the reference mark 1 to the reference mark 2 to the reference mark 3, and so on With this technique, the surgeon can easily detect the divergence from the expected route. For example, if the catheter and magnet 120 inadvertently deviate to a different vein, the surgeon will easily notice the divergence from the marked path, and will quickly identify the problem. The catheter and the magnet 120 can be removed and re-inserted to follow the path of the reference marks. The general operation of the detector system 100 is illustrated in the flow diagram of Figure 6A. At start 200, magnet 120 (see Figure 4) has been inserted into the patient. In step 201, the system goes through an initial calibration. In an exemplary embodiment, the initial calibration is performed before the magnet 120 is introduced. Accordingly, the system 100 compensates for the effects of the Earth's magnetic field, including localized disturbances, in the absence of any contribution from the magnet 120. Alternatively, the magnet 120 can be placed in a known location with respect to the housing 102, so that the effects of the magnetic field caused by the magnet 120 are known, and can be canceled in the manner described above with respect to to the automatic recalibration process. That is, the contribution to the measured magnetic field caused by the magnet 120 at the known location can be subtracted from the measured readings, with the resulting residual value being caused only by the Earth's magnetic field. Following the initial calibration, in step 202, the detector system 100 measures the detection values from the magnetic sensors 108-114. In step 204, the estimation processor 152 (see Figure 5A) calculates an initial estimate of the location a, and of the orientation of the magnet 120. The initial estimate includes the position data of the sensor from step 208, and the magnet calibration data from step 209. The sensor position data calculated in step 208 provide data related to the position of each of the magnetic sensors 108-114 in relation to a selected origin. For example, a magnetic sensor (e.g., magnetic sensor 108) may be arbitrarily selected as the mathematical origin for purposes of determining the relative positions of the other magnetic sensors (e.g., magnetic sensors 110-114). The common origin provides a frame of reference for the purposes of mathematical calculations. As discussed above, the magnetic sensors 108-114 are aligned with respect to the common origin, such that each magnetic sensor measures the magnetic field in the same directions x, y, and z. As those of ordinary experience in this field can appreciate, any source selected successfully with the detector system 100 can be used. The magnetic calibration data derived in step 209, are usually provided by the magnet manufacturer, and include related data. with the force of the magnetic dipole m (see Figure 4), as well as the size and shape of the magnet 120. The measured values of the sensor, the sensor position data, and the magnet calibration data are provided as inputs to the estimation processor 152 (see Figure 5A) in step 204. In an exemplary embodiment, the initial estimate of location a is provided via neural network 154 (see Figure 5A) based on the measured sensor values derived in step 202. As discussed above, neural network 154 may require minimum values from magnetic sensors 108-114, to guarantee an estimate reliable initial The neural network 154 provides the initial estimate of the location-and orientation of the magnet. - In step 210 the estimation processor 152 (see Figure 5A) calculates the predicted sensor values. As described above, does this require a measurement? .. (predicted) for each combination of the magnetic sensors 108-114 in each of the three orthogonal directions x, y, and z. In step 212, the estimation processor 152 compares the predicted sensor values (i.e.? • • (predicted)) with the sensor measured values (i.e.? • - (measured)). In decision 216, the estimation processor 152 determines whether the predicted and measured values of the sensor agree within a desired degree of tolerance. If the predicted values of the sensor and the measured values of the sensor do not match, the result of decision 216 is NO. Then, the estimation processor 152 calculates a new estimate of the location of the magnet a and of the orientation in step 218. Following the calculation of a new estimated location a of the magnet 120, the estimation processor 152 returns to step 210 to calculate a new set of predicted values of the sensor, using the new estimation of location and orientation of the magnet. The estimation processor 152 continues this iterative process of adjusting the estimated location of the magnet 120 and the orientation, and comparing the predicted values of the sensor with the measured values of the sensor, until a "concordance" is achieved. the predicted values of the sensor and the measured values of the sensor, the result of the decision 216 is SI.In that case, in step 220, the detector system 100 exhibits the location of the magnet and the orientation in the visual display 106 (see In addition, the detector system 100 may exhibit a confidence value indicating a degree of confidence with which the location and orientation of the magnet 120 have been determined. The calculation of a confidence value - based on statistical data is well known in the art, and does not need to be described in detail here, following the visual display of location and orientation data in step 220, the detector system 100 returns to step 202, and repeats the process on a new set of sensor measured values. If the cost function is too high, a match can not be reached in decision 216. These conditions can arise, for example, in the presence of extraneous magnetic fields. In practice, it has been determined that matches have a cost function in the range of 1 to 2, while the minimum cost function for an inaccurate local minimum is orders of magnitude higher. If a match can not be reached (ie, the cost function is too large), the detector system 100 can start the measurement process again with a new estimated location, or generate an error message, indicating a cost function unacceptably high. The flow chart of Figure 7B illustrated the steps performed by the calibration processor 158 if the automatic calibration is implemented within the detector system 100. In this implementation, following completion of step 220, the system 100 can optionally be moved to the step, illustrated in Figure 7B, wherein the calibration processor 158 obtains the position data from the digitizer arm 180 (see Figure 5C), indicating the current location of the detector system 100. Given the new location of the detector system 100 , and the known location of the magnet 120, the calibration processor 158 calculates the magnetic field resulting from the magnet, and subtracts the effects of the magnet from the current measurements in step 226. As a result of this process, the remaining residual values measured by the magnetic sensors 108-114 (see Figure 5A) are due to the effects of the Earth's magnetic field, including perturbations l ocated. In step 228, this residual value is used to reset the detector system 100 to zero, in order to compensate for the effects of the Earth's magnetic field at the new location. Following the recalibration process, the detector system 100 returns to step 202, shown in Figure 7A, to perform additional measurement cycles with the detector system 100 at the new location, and recalibrate for operation at the new location. It should be noted that the automatic recalibration process illustrated in the flow diagram of Figure 7A, automatically and continuously recalibrates the detector system 100. However, in an alternative embodiment, the calibration processor 158 will perform the recalibration process only if the detector system 100 has been moved by a predetermined amount. This prevents unnecessary recalibration when the detector system 100 has not been moved. The iterative estimation process is described above using the difference in magnetic force B provided by different pairs of magnetic sensors 108-114. In an alternative way, the detector system 100 can use the measured field gradient values G. In this embodiment, equation (2) can be adjusted to the measured values, in a manner as described above with respect to the iterative process, to adjust the measurements of B. With respect to the flow chart of Figure 7A, step 202 provides gradient values with respect to the pairs of magnetic sensors 108-114. For example, a magnetic gradient measurement can be calculated using the magnetic field B measured by the magnetic sensor 114 with respect to the magnetic field measured by each of the remaining magnetic sensors 108-112, respectively. In step 204, the estimation processor 152 determines an initial estimate of the location and orientation of the magnet, and in step 210, calculates the predicted values of the sensor using equation (2). In step 212, the measured values of the sensor are compared with the predicted sensor values, using conventional techniques, such as the cost functions described above. The iterative process continues until the measured values of the sensor and the predicted values of the sensor match within the previously determined degree of tolerance. In still another alternative technique, the detector system 100 uses the measurement data, and solves the equation (2) for a directly. The direct solution approach uses the fact that G is a symmetric matrix with positive eigenvalues. The eigenvalues and the eigenvectors of the G matrix can be calculated and used algebraically to solve for location a and m directly. This assumes that the magnitude, but not the direction of m is known. In practice, the magnitude m is known because the manufacturer provides the calibration data of the magnet. It should be noted that this technique requires an additional magnetic sensor to determine the orientation of the magnetic dipole. Mathematically, the orientation of the magnetic dipole is indicated by a + or - sign. The additional magnetic sensor, which only needs to measure the strength of the magnetic field B, is used to determine the sign of the mathematical function. In addition, combinations of these different techniques can be used by the detector system 100, to determine the location of the magnet 120. In still another alternative, a Kalman filter can be used with the equations (1) and (2) above, to track the position of the magnetic dipole m with respect to the array of multiple detectors formed by the magnetic sensors 108-114. As is known to ordinary experts in the field, Kalman filters are statistical predictive filters that utilize statistical signal processing and optimal estimation. Numerous textbooks, such as "Tracking And Data Association", by Y. Bar-Shalom and R.E. Fortmann, Academic Press, Bosto, 1988, provide details on the theory and operation of Kalman filters. In addition to the individual techniques described above, it is possible to use any or all of these techniques in combination, such as a sum of cost functions for each type of sensor. For example, it may be required that the differences between? - - (predicted) and? (Measured) agree within a certain tolerance. If multiple mathematical techniques are not able to identify a solution for which all difference values satisfy that tolerance, then an error can be signaled to the operator using the visual display 106 (see Figure 5A). Assuming that the errors of each sensor measurement are independent and small, the uncertainty in the estimation of the location can be calculated using, for example, Cramer-Rao links. Accordingly, a degree of redundancy can conveniently be implemented between the measurement techniques by the detector system 100. This redundancy is highly desirable for biomedical applications. Figure 3 illustrates the operation of the detector system 100 for a specific configuration of the magnetic sensors 108-114. Nevertheless, the techniques described above can be generalized to virtually any fixed sensor consideration. A minimum of one gradient sensor or eight magnetic field sensors is required to measure G (s) and B (s), respectively, assuming the strength of the magnetic dipole m is known. The magnetic sensors can be configured in a relatively arbitrary manner, and therefore, can be easily placed in locations within the housing 102 (see Figure 3A and 3B), based on the design of the instrument and / or other signal or noise considerations . Magnetic sensors 108-114 can be calibrated using the known strength of the Earth's magnetic field. In the absence of any inhomogeneous fields (ie, away from any strong magnetic dipoles), the sensor element X of all sensors 108-114 can be read at the same time. In a similar manner, the sensor elements Y and the sensor elements Z can be read at the same time. In any consideration, the sum of the squares of the average readings of the strength of the magnetic field for each orthogonal direction (ie, Bx, By, and Bz) must be constant. The constant value of the Earth's magnetic field can be used to determine the appropriate calibration factors for each magnetic sensor, using conventional algebraic and least-squares fitting methods. An alternative calibration technique uses a small magnet of a known force placed in one or more locations relative to the magnetic sensors 108-114. Measurements are made in each of the one or more locations to determine the appropriate calibration factors for each magnetic sensor. Other techniques, such as the use of an electromagnetic cage, Helmholtz cage, or the like, can also be used to calibrate the magnetic sensors 108-114. The visual display 106 (see Figure 3) provides a visual visual display of the position of the magnet 120 with respect to the housing 102. Figures 8A through 8D illustrate some of the different techniques used by the detector system 100 to indicate the location of the magnet 120 (see Figure 4). In the embodiment illustrated in Figure 8A, the visual display 106 uses a circle 250, and a pair of orthogonal lines 252a and 252b, to indicate the location a of the magnet 120 in relation to the housing 102. The orthogonal lines 252a and 252b provide a visual indicator to the assistant to help determine when the magnet 120 is centered under the detector system 100. In an alternative embodiment, illustrated in Figure 8B, a fixed indicator 254, such as orthogonal lines 254a and 254b, forms beautiful cross-over the center of the screen 106. The circle 250, or other indicator, is used to provide a visual indication of the location a of the magnet 120 in relation to the housing 102. The circle 250 centers on the beautiful crusaders in the center of the display 106, when the magnet 120 is centered directly below the detector system 100. In yet another embodiment, shown in Figure SC, the visual display 106 provides an in different indicator, such as a row 260, to provide a visual indication of the location a of the magnet 120. The row 260 can also be used to indicate the orientation of the magnet 120. The depth of the magnet 120 below the patient's surface can be indicate in the visual display 106 in a variety of ways. For example, a portion 106a of the visual display 106 may provide a visual indication of the depth of the magnet 120 using a bar graph, as illustrated in Figure 8D. However, the depth indicator portion 106a of the visual display 106 may also provide a numerical reading of the depth of the magnet 106 in absolute units, such as centimeters, or in relative units.

Accordingly, the detector system 100 determines the location a of the magnet 120 in a three-dimensional space, and provides an easy-to-read visual indication of the location of the magnet, including a depth indication, as well as the orientation of the magnetic dipole. Although the housing 102 is illustrated as a rectangular housing, with the magnetic sensors 108-114 distributed in an equidistant manner within the housing 102, the rectangular shape was selected for its ease of being held by the assistant. However, the housing 102 may have any shape or size. In addition, the visual display 106, although illustrated as a liquid crystal display, can be any convenient two-dimensional visual display, such as a visual display of dot matrix or the like. Accordingly, the present invention is not limited by the specific size or shape of the housing 102, or by the specific type of visual display 102. In addition, the detector system 100 can operate in a satisfactory manner with a variety of different magnetic sensors. Accordingly, the present invention is not limited by the specific number or type of magnetic sensors employed in the detector system 100.

Clinical Studies The operation of the detector system 100 for the detection of a static magnetic field associated with the magnet 120 inserted inside a patient has been described. The reliability of the 100 detector system has been tested in clinical studies, the results of which are described below. As will be described in detail below, the location of the magnet was determined using the detector system 100, and subsequently verified using conventional fluoroscopic measurements. Although the initial results of the clinical studies indicate a relatively high error in the location detected by the measurement system 100, it is believed that these errors are caused by the imprecise alignment of the detector system and the fluoroscopic measurement system. Accordingly, the errors are due to misalignment rather than inherent inaccuracies in the detector system 100. In addition, revisions in the signal processing software resulted in greater reliability in subsequent measurements in the clinical study, as discuss later. An application of the detector system 100 -_- is for the insertion of a catheter into the heart. The placement of a peripherally inserted central catheter (PICC) in the lower half of the superior vena cava, just above the right atrium, is a critical application for the 100 detector system. Currently, physicians perform this task "blindly", measuring External anatomical reference marks, and inserting the catheter to the measured depth. The success or failure of the insertion is unknown until a chest x-ray is obtained, which may not occur for several days. The detector system 100 was evaluated in an animal model as a possible solution to the "blind" placement. Forty-four locations were made using the detector system to test its accuracy, comparing with the conventional fluoroscope. The detector system 100 located central peripherally inserted magnetic catheters labeled to within an average of 0.4 centimeters, and a scale of 0.2 centimeters to 1.25 centimeters. The detector system 100 also provided valuable real-time information about the trajectory and orientation of the peripheral catheter tip peripherally inserted during difficult insertions. The detector system 100 has demonstrated its ability to precisely locate a centrally inserted peripherally labeled magnetic catheter in relation to an external reference mark, and consequently, aid in the insertion of the catheter. The measurement capabilities provided by the detector system 100 has the potential to improve clinical outcomes, and consequently, reduce the cost of health care by decreasing the problems associated with the catheter in infusion therapy, and by decreasing or eliminate the need for radiographic verification of peripherally inserted central catheter placement, and other placements of medical devices.

Introduction The peripherally inserted central catheter is inserted into a peripheral vein in the patient's arm, and is threaded into the superior vena cava to a point approximately 2 centimeters above the right atrium. The current method for inserting peripherally inserted central catheters is to measure the distance from the insertion point to the third right intercostal space on the patient's chest, and insert the peripherally inserted central catheter to a depth equal to this measurement. Catheters are used for patients who require long-term intravenous access (2 weeks to 6 months) for infusions, blood sampling, or blood transfusion. Currently, peripherally inserted central catheters can be placed in outpatients or at home by nurses, but the catheter can not be used for infusions or sampling, until their locations have been verified by radiography, which is inconvenient, relatively expensive, and can delay therapy for days.

Animal Model Cross-breeding domestic pigs were used as the animal model for this study. Thistles are an accepted model of the human cardiovascular system, and have a cephalic vein in their thoracic limb that offers an acceptable route to the cranial vena cava, which is analogous to the superior vena cava in humans. A necropsy study done before the present study has indicated that the second right external intercostal space is a good external reference mark to locate a point 2 centimeters above the right atrium in the cranial vena cava. -1 study also showed that the distance from the chest wall to the dorsal cranial vena cava is 8.5 centimeters to 10 centimeters in animals weighing approximately 30 kilograms. This distance is analogous to the distance in humans for the analogous procedure. This last factor is significant, because the detector system 100 has a distance limit of approximately 10 centimeters to locate the smallest magnetically labeled catheter used in the study.

Centrally Peripherally Inserted Catheters Magnetically Labeled The peripherally inserted central catheters commercially available and the introducers were modified. placing one or more small cylindrical magnets (NdFeB) on the tips of the catheters, and sealing the ends of the catheters with a non-sterile medical-grade silicone adhesive. Two catheter sizes were used. The smaller size catheters (4 Fr, 65 cm length) contained three N-40 magnets, nickel-plated, 0.8 mm x 2.5 mm, and the large catheters (5 Fr, 65 cm length) contained two N magnets. -40, plated with nickel, of 1.0 mm by 2.5 mm. The strength of the magnetic field of each magnetic tip catheter was 3.129 milliGauss per cubic centimeter.

Magnetic Field Detector Two different versions of the 100 detector system were used in the study. A bank feasibility system was used for 44 locations, and a manual prototype was used for 28 locations. The manual unit included four magnetic field sensors (for example, magnetic sensors 108-114) mounted in a plastic box with control buttons and their associated electronics. A peripheral unit was also used, which contained the processing hardware, the software, and a power supply, with the manual version of the detector system 100. A single manual unit was used with three different software systems. Eight locations were made with software revision 5.0, 16 locations were made with software revision 5.1, and four locations were made with software revision 5.2. As will be discussed later, the first revisions of the software for the manual prototype required a significant debugging and calibration of the software. More reliable measurements were obtained with software revision 5.2. The bank version of the detector system 100 comprises four magnetic field sensors (e.g., magnetic sensors 108-114) mounted on a Plexiglas platform with its associated electronics. The bank version of the detector system 100 was coupled to a personal computer (PC), where software was used to calculate the position and orientation of the magnet in three dimensions, and to display the information on a conventional personal computer monitor, in the shape of an image indicated by the magnet tip catheter. Coupling grids were placed on the Plexiglas platform and on the personal computer monitor, to correlate the position of the monitor with the position of the external anatomy of the subject.

Clinical Procedure The study was conducted in nine healthy crossbred domestic pigs, approximately 25 kilograms. Each subject was completely anesthetized before the procedure, and euthanized immediately following the procedure. After the initiation of anesthesia, each subject was subjected to four catherations in sequence. The subjects were measured externally from the insertion point to the desired external reference mark after establishing venous access by a cut down the axilla. A centrally inserted peripheral 4-Fr magnet catheter was inserted twice via an introducer into the right cephalic vein, and a peripherally inserted central catheter labeled with 5 Fr magnet was inserted twice via an introducer into the left cephalic vein. Each catheter was placed in the locations of the cranial and medial-clavicular vena cava, and the position of the tip with magnetic label was determined by means of a model of the detector system 100 in each location, resulting in the total of eight locations per animal. The location of the catheters was confirmed with a fluoroscope, and an approximate precision of the location correlation system of the detector / fluoroscopic system was determined by aligning the fluoroscope with the detector system 100, using a rig attached to the fluoroscope. Both versions of the detector system 100 were placed on the subject before insertion of the catheter with a cross-linked arm and leveled within a degree in relation to the fluoroscope rigging, using a conventional digital level. In the bank version of the detector system 100, an alignment rod was placed through the center of the rig, and aligned on the grid on the Plexiglas platform corresponding to the grid on the personal computer monitor. In the manual prototype version of the detector system 100, a paper marker was placed on the supply on the screen, and the alignment rod was aligned with the paper marker. Electronically captured fluoroscopic images were analyzed with a commercial drawing program to estimate the error measured from the position determined by the detector system 100 and the center of the mass of the magnet determined by fluoroscopy. This measurement is considered a conservative estimate. The image of the tip of the magnet was used as a reference to determine the scale of the image, and the tip may have been angulated during the studies. An angled tip scales the image to an amplification greater than the real one, thus increasing the measured error. It is believed that this effect is minor, because the magnet tips appeared to be relatively flat in each fluoroscopic image. The surgeon who made the measurements made subjective estimates of the position of the catheter.

Results The bank version of the 100 detector system worked well during all locations, but the first twelve insertions were complicated by the difficulty in aligning the fluoroscope rig with the detector system 100. In the first twelve inserts, it was assumed that the rod The alignment used to align the rig was straight, but it was shown that the alignment rod could be stopped at an angle, which affected the measured error of the location. -After the tenth second insertion, the configuration of the alignment rod was altered to allow it to hang straight, and after the sixteenth insertion, a hollow Plexiglas cylinder was added to the rig platform to keep the alignment rod straight. Following these modifications, the detector system 100 provided more consistent and accurate results. Peripherally inserted central catheters magnetically inscribed with ease were inserted into the cranial venae cavae of the pigs through the introducers provided in the insertion kits that accompanied the centrally inserted peripheral catheters. The detector system 100 helped the researchers twice during the difficult insertions of peripherally inserted central catheters. In a case, the detector system 100 had indicated that the catheter had been bent back on itself in the cranial vena cava, and the catheter was removed until the image indicated that the orientation was correct, and the catheter was properly inserted. In a second case, it was difficult to pass the central catheter peripherally inserted from the left cephalic vein to the external jugular vein, due to an acute angle in this junction, which was subsequently verified using fluoroscopy. The surgeon used the real-time feedback from the detector system 100 to twist, insert, and remove the catheter, until it became clear that the catheter was oriented in an anatomically appropriate direction. When the tip of the catheter passed the acute angle, it was easily inserted. It should be noted that the results of the three versions of the manual prototype of the detector system 100 are not included in this report, because the software and the calibration procedures varied, and the localization results varied according to the same. The precision of placement of the peripherally inserted central catheter tip was determined by measuring the distance from the actual location of the tip labeled with magnet, determined by the detector system 100, to the actual location of the catheter tip determined by fluoroscopy. The 44 locations were made in the positions of the cranial and medial-clavicular vena cava, and there was no significant difference between the error measured in these locations (p = 0.90). The average measured error for the 44 locations in six animals, using the bank version of the detector system 100, was 0.40 centimeters with a standard deviation of. + 0.29 centimeters. The results of the bank version of the detector system are illustrated in Figure 9. The measured error was 0.02 centimeters to 1.25 centimeters, but 5 of 6 locations with errors greater than 0.6 centimeters were made in the first 12 placements. As discussed above, the first placements were complicated by the difficulties in aligning the fluoroscope rig with the detector system 100. As can be easily seen from Figure 9, the alignment difficulties were solved after location number twelve, with the resulting decrease in the measurement error.

Conclusions After the first eight locations, the surgeon was asked to determine the anatomical position of the catheter tip by fluoroscopy without input from other observers. After the peripherally inserted magnet-tipped central catheter was placed using the bench version of the detector system 100, the surgeon confirmed that the peripherally inserted central catheter was in the desired position at each location. - - - The use of an external anatomical reference mark in the placement of peripherally inserted central catheters allows health care providers to insert the catheters in many different facilities, from home to outpatient clinics. The detector system successfully demonstrated that it could locate the tip of the labeled catheter within an average of four millimeters relative to an external reference mark. The external reference marks used in this study do not correlate precisely with human reference marks, due to differences in interspecies anatomy, but the concept of placing a catheter in a prescribed reference mark has been established. the detector system 100. The detector system 100 also allowed users to overcome difficulties in catheter insertions. In several cases during the study, the operator felt resistance at some point during insertion, and used the position and orientation data in real time to place the catheter correctly. This ability proved very useful when the catheter was bent back on itself, which is easily shown using the detector system 100, because the catheter tip stopped its forward progress, and rolled toward a new orientation. At this point in time, the operator withdrew the catheter until the tip of the image resumed its proper orientation, and the insertion was completed. Another valuable application is the ability to observe the image of the tip of the catheter as the catheter passes through acute angles and curves in the venous system. The investigator used this aspect of the detector system 100 while passing the catheter from the left cephalic vein to the left external jugular vein. The user felt considerable resistance that correlated with the supply that looked like a "collapse" in a wall. By twisting and replacing the catheter, it eventually passed into the jugular vein, and the investigator felt comfortable that it was correctly positioned. Without immediate real-time feedback, the user does not know if the catheter takes an erroneous turn or doubles until the procedure is completed, and the patient has undergone radiographic verification. Accordingly, the present study illustrates the ability of the detector system 100 to precisely locate the tip of the catheter relative to an external reference mark in an animal model, and displays the basic work to test its clinical efficacy in centrally positioned peripheral catheters. inserted and other medical devices. From the foregoing, it will be appreciated that, although specific embodiments of this invention have been described herein for purposes of illustration, various modifications may be made without departing from the spirit and scope of the invention. In accordance with the foregoing, the invention is not limited except by the appended claims.

Claims

1. A system for detecting a position of a magnet associated with a medical device resident from a measurement location on the surface of a patient, the system comprising: a plurality of magnetic sensors held in a fixed position with respect to each other, each being of the magnetic sensors oriented in a known direction, and generating signals as a function of the strength of the static magnetic field and the direction, due to the magnet; a processor for calculating: an initial estimated position of the magnet in a three-dimensional space; a magnetic field force predicted for at least a portion of the plurality of sensors, based on the estimated position; a real magnetic field strength for the portion of sensors that use the signals; an error function based on a difference between the strength of the predicted magnetic field and the strength of the real magnetic field; and a visual display that provides a visual display of the data related to the position of the magnet in three-dimensional space.

The system of claim 1, wherein the processor calculates the initial estimated position based on the signal from at least one sensor.

The system of any one of the preceding claims 1 and 2, wherein the processor calculates the initial estimated position based on the signal of at least one sensor above a predetermined threshold.

The system of any one of the preceding claims 1 to 3, which further includes a neural network for generating the initial estimated position, the neural network receiving the signals, and generating the initial estimated position based thereon.

The system of claim 4, wherein the neural network includes a learning mode and an operating mode, the neural network operating in the learning mode to receive a plurality of sets of the signals, and to generate the estimated positions for each of the plurality of sets of the first, second, and third sets of signals, the neural network also receiving dice related to a real position of the magnet after generating each of the estimated positions, and using the plurality of sets of signals, the estimated position data, and the actual position data, to create rules for generating the estimated position data, while operating in the operating mode, the neural network operating in the operating mode to receive the signals and to generate the Initial estimated position of the magnet based on the signals and rules created while operating in the learning mode.

6. The apparatus of any of the claims 1 to 5 above, wherein the plurality of magnetic sensors are orthogonally configured to provide a three-dimensional measurement.

The system of any one of claims 1 to 6 above, wherein the magnet has a magnetic dipole moment, indicating the orientation of the magnet, and the detected magnetic dipole moment is displayed in the visual display to indicate the orientation of the magnet. magnet.

The system of any one of claims 1 to 7 above, which further comprises a housing, wherein the visual display is supported by the housing.

9. The system of claim 8, wherein the visual display is a two-dimensional visual display, with at least a portion of the visual display being transparent, to enable the user to see the patient's surface below the transparent portion.

The system of any one of claims 1 to 9 above, which further comprises a housing for supporting the plurality of sensors, the visual display being an external deployment separate from the housing and electrically coupled to the processor.

The system of claim 10, for use with an imaging device capable of generating an image of the patient's internal anatomy, where the visual display exhibits the image of the internal anatomy of the patient combined with the data related to the position of the magnet.

12. The system of any one of claims 1 to 11 above, wherein the processor calculates in an iterative manner the estimated position and the strength of the predicted magnetic field, until the error function indicates that the magnetic field strength predicted matches the strength of the actual magnetic field within a previously determined tolerance.

The system of claims 1 to 11 above, wherein the visual display provides the visual display of the data based on a single generation of the error function.

The system of any one of claims 1 to 13 above, for use in the presence of an Earth magnetic field, wherein the processor subtracts a first selected signal generated by a first sensor selected from the plurality of magnetic sensors from of a second selected signal generated by a second sensor selected from the plurality of magnetic sensors, to cancel the effects of the Earth's magnetic field

15. The system of claim 14, which further comprises a housing for supporting the plurality of sensors, and a position detector to detect the position of the housing, and generate position data related thereto, and a calibration processor to compensate for variations in the Earth's magnetic field resulting from the movement of the housing to a new location , calculating the calibration processor the change in the position of the housing based in the position data, calculating the strength of the real magnetic field in the new location, and calculating the contribution to the real magnetic field resulting from the magnet, using in addition the calibration processor a difference between the strength of the real magnetic field in the new location and the contribution-to the real magnetic field resulting from the magnet, to compensate for the effects of the Earth's magnetic field. -

16. The system of any one of claims 1 to 16 above, which further comprises a housing for supporting the plurality of magnetic sensors, and a position detector for detecting the position of the housing and generating position data related to it, being able to operate the system for recording the position of the housing in a plurality of locations selected by the user, the visual display providing a visual display of the selected locations combined with the data related to the position of the magnet.

17. A method for detecting a position of a magnet associated with a medical device resident from a measurement location on the surface of a patient, the method comprising: placing a plurality of magnetic sensors held in a fixed position with respect to each other, each of the magnetic sensors being oriented in a known direction, and generating signals as a function of the strength of the static magnetic field and the direction due to the magnet; calculate an initial estimated position of the magnet in a three-dimensional space; calculating a predicted magnetic field strength for at least a portion of the plurality of sensors, based on the estimated position; calculate a real magnetic field strength for the portion of the sensors that use the signals; - calculating an error function based on a difference between the strength of the predicted magnetic field and the strength of the actual magnetic field; and visually displaying data related to the position of the magnet in three-dimensional space.

The method of claim 17, wherein the calculation of the initial estimated position is based on the signal from at least one sensor.

The system of any of claims 17 and 18 above, wherein the calculation of the initial estimated position is based on the signal of at least one sensor above a previously determined threshold.

The method of any of claims 17 to 19 above, wherein the calculation of the initial estimated position is performed by a neural network, the signals receiving the neural network, and generating the initial estimated position based thereon.

The method of claim 20, wherein the neural network includes a learning mode and an operating mode, the method further comprising the neural network operating in the learning mode to receive a plurality of signal sets, and to generate the positions estimated for each of the plurality of sets of the first, second, and third sets of signals, the neural network also receiving data related to a real position of the magnet after generating each of the estimated positions, and ut-i -lising the plurality of signal sets, the estimated position data, and the actual position data, to create rules for generating the estimated position data, while operating in the operating mode, the neural network operating in the operating mode for receive the signals and to generate the initial estimated position of the magnet, based on the signals, and on the rules created while operating in the learning mode.

22. The method of any one of claims 17 to 21 above, wherein the magnet has a magnetic dipole moment that indicates the orientation of the magnet, and the display of the data comprises displaying the detected magnetic dipole moment to indicate the orientation of the magnet. magnet.

23. The method of any one of claims 17 to 22 above, for use with an imaging device capable of generating an image of the internal anatomy of the patient, the method further comprising displaying the image of the patient's internal anatomy, combined with the data related to the position of the magnet.

24. The method of any of the claims 17 to 23 above, which also includes altering the initial estimated position based on the error function, and iteratively calculate the estimated position, the strength of the predicted magnetic field, and the error function, until the error function indicates that the strength of the predicted magnetic field matches the strength of the actual magnetic field within a previously determined tolerance.

25. The method of claims 17 to 23 above, wherein the. Visual display provides the visual display of the data based on a single generation of the error function.

26. The method of any of claims 17 to 25 above, for use in the presence of an Earth magnetic field, which further comprises subtracting a first selected signal generated by a first sensor selected from the plurality of magnetic sensors from of a second selected signal generated by a second sensor selected from the plurality of magnetic sensors, to cancel the effects of the Earth's magnetic field.

The method of claim 26, which further comprises: detecting the position of the plurality of magnetic sensors resulting from the movement of the plurality of magnetic sensors to a new location, and generating the position data related thereto.; calculating the change in position of the plurality of magnetic sensors based on the position data; calculate the values related to the strength of the real magnetic field in the new location; calculate the contribution to the values related to the real magnetic field resulting from the magnet; and using a difference between the values related to the strength of the real magnetic field at the new location, and the contribution to the actual magnetic field resulting from the magnet, to compensate for the effects of the Earth's magnetic field resulting from the movement of the plurality of magnetic sensors until the new location.

28. The method of any of the claims 17 to 27 above, which further comprises: detecting the position of the magnetic sensors, and generating the position data related thereto; recording the positions of the magnetic sensors in a plurality of locations selected by the user; and displaying a visual display of the selected locations combined with the data related to the magnet's position.

29. A system for detecting a position of a magnet associated with a medical device resident from a measurement location on the surface of a patient, the system comprising: a plurality of magnetic sensors, each oriented in a known direction, and generating a set of signals as a function of the strength of the static magnetic field and the direction due to the magnet; a processor for calculating an estimated position of the magnet in a three-dimensional space, and for calculating values related to a magnetic field strength predicted for at least a portion of the plurality of sensors, based on the estimated position, further calculating the processor's values related to the strength of the real magnetic field using the set of signals, and determining the values related to the location of the magnet, based on a difference between the values related to the strength of the predicted magnetic field and the values related to the strength of the magnetic field real; a position detector to determine the location of the magnetic sensors, and to generate the position data related thereto; a calibration processor for receiving the position data and the values related to the location of the magnet, and for compensating the effects of the Earth's magnetic field, as the location of the magnetic sensors changes with respect to the patient; and a visual display that provides a visual display of the values related to the position of the magnet in three-dimensional space.

The system of claim 29, wherein the calibration processor calculates the change in the position of the magnetic sensors based on the position data, calculates the strength of the actual magnetic field at a new location, and calculates the contribution to the values related to the real magnetic field resulting from the magnet, using in addition the calibration processor a difference between the values related to the strength of the real magnetic field in the new location, and the contribution to the values related to the real magnetic field resulting from the magnet , to compensate for the effects of the Earth's magnetic field.

The system of any one of claims 29 to 30 above, wherein the processor calculates in an iterative manner the estimated position and the values related to the strength of the predicted magnetic field, until the values related to the strength of the predicted magnetic field they agree with the values related to the strength of the real magnetic field within a predetermined tolerance.

The system of claim 31, wherein the processor performs a first iteration based on an initial estimated position, the system also including a neural network to generate the initial estimated position, the neural network receiving the first, second, and third sets of signals, and generating the initial estimated position based on them.

The system of any one of claims 29 to 30 above, wherein the processor calculates the position of the magnet based on a single calculation of the values related to the strength of the predicted magnetic field, and the values related to the strength of the magnetic field. real.

34. The system of any of claims 29 to 33 above, wherein the processor calculates the estimated position using a mathematical equation representative of the strength of the magnetic field.

35. The system of any one of claims 29 to 33 above, wherein the processor calculates the estimated position using a mathematical equation representative of a gradient of the force of the magnetic field.

36. The system of any of the claims 29 to 35 above, wherein the processor calculates a cost function based on the difference between the values related to the strength of the predicted magnetic field, and the values related to the measured value related to the strength of the magnetic field, also generating the processor a minimum value for the cost function, and analyzing the minimum value for the cost function with a previously determined minimum value, and generating a signal for the user, to indicate if the minimum value is above or below the previously determined minimum value.

37. A system for detecting a position of a magnet associated with a medical device resident from a measurement location on the surface of a patient, the system comprising: a plurality of magnetic sensors, each generating a set of electrical signals as a function of a force of the magnetic field and the direction due to the magnet; a processor to calculate the measurements of magnetic field gradients for the magnetic sensors from the sets of electrical signals, and calculate the position of the magnet in a three-dimensional space, using a mathematical equation representative of a gradient of magnetic field strength and of the calculated magnetic field gradient measurements; and a visual display that provides a visual display of the data related to the position of the magnet.

38. A method for detecting a position of a magnet associated with a medical device resident from a measurement location on the surface of a patient, the method comprising the steps of: placing a plurality of magnetic sensors in predetermined locations with respect to the magnet; generate a set of signals related to the strength of the magnetic field and the direction from the sensors to the magnet; calculate the gradient measurements of the magnetic field for the magnetic sensors from the signal set, and calculate the position of the magnet in a three-dimensional space, using a magnetic equation representative of a gradient of magnetic field strength and field gradient measurements magnetic calculated; and display the data related to the position of the magnet. SUMMARY A device for detecting the location of a magnet coupled to a medical device resident within a patient, uses three or more sets of magnetic sensors, each having sensor elements configured in a known manner. Each sensor element detects the strength of the magnetic field generated by the magnet, and provides data indicating the direction of the magnet in a three-dimensional space. The device uses fundamental equations for electricity and magnetism that relate the strength of the measured magnetic field and the gradient of the magnetic field to the location and strength of a magnetic dipole. The device uses an iterative process to determine the actual location and orientation of the magnet. An initial estimate of the location and orientation of the magnet results in the generation of predicted magnetic field values. The predicted magnetic field values are compared to the actual measured values provided by the magnetic sensors. Based on the difference between the predicted values and the measured values, the device estimates a new location of the magnet, and calculates new predicted magnetic field strength values. This iteration process continues until the predicted values agree with the measured values within a desired degree of tolerance. At that point, the estimated location matches the actual location within a predetermined degree of tolerance. A two-dimensional visual display provides an indication of the location of the magnet with respect to the detector housing. A depth indicator portion of the visual display can be used to provide a relative or absolute indication of the depth of the magnet inside the patient.