DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1-12, 15-18, 28, and 42 are rejected under 35 U.S.C. 103 as being unpatentable over Fuerst et al (US20210290311A1; hereinafter referred to as Fuerst) in view of Besser (US20220117676A1) and further in view of Sebkhi et al (US 20240082032 A1; hereinafter referred to as Sebkhi) .
Regarding Claim 1, Fuerst discloses a method for tracking a medical tool during a medical procedure, the method comprising (“FIG. 15 is a process flow of a method of sensing position and/or orientation of an object, which can be performed by embodiments of the surgical robotic system of FIG. 12 and variations thereof.” [0021]):
providing a tracking system comprising a medical tool comprising a magnetic or electromagnetic element (“The machine learning model is coupled to receive output data of the magnetic field sensors. The machine learning model is trainable to output a three-dimensional sensed position, a three-dimensional sensed orientation, or both (sensed pose, in six degrees of freedom), of a trocar that is producing a magnetic field.” [0006], “The trocar 63 includes a first magnet 71 and a second magnet 73 producing respective magnetic fields B1, B2 with known properties” [0046]);
and an array of magnetic sensors; utilizing a processor(“an arm-to-trocar docking capability of a surgical robotic system senses position, orientation or both (pose) of the trocar. The surgical robotic system includes a surgical robotic arm, magnetic field sensors on the arm, and a digital processor that implements a machine learning model (e.g., an artificial neural network “neural network”).” [0006]):
receiving one or more signals from the array of magnetic sensors, the one or more signals are indicative of a change and/or disturbances in one or more of: a strength, direction, flux and magnitude of a magnetic vector detected in a magnetic field generated by the magnetic or electromagnetic element within the body of a subject (“The machine learning model is coupled to receive output data of the magnetic field sensors. The machine learning model is trainable to output a three-dimensional sensed position, a three-dimensional sensed orientation, or both (sensed pose, in six degrees of freedom), of a trocar that is producing a magnetic field.” [0006], ”Each sensor, e.g., of magnetic field sensors 55, 57 labeled sensor 1, sensor 2, etc., senses magnetic field in each of three orthogonal directions, and is of a type commonly known as an XYZ or three axis magnetic field sensor. Here, these magnetic field sensors are depicted as each sensing and producing signals for B0, B1 and B2 magnitudes of the magnetic field vector at the location of the sensor.” [0075]);
applying the one or more signals to one or more deep learning algorithms, the algorithms trained on a database comprising signals associated with a plurality of known spatial positions and orientations of the magnetic or electromagnetic element (“The machine learning model is coupled to receive output data of the magnetic field sensors. The machine learning model is trainable to output a three-dimensional sensed position, a three-dimensional sensed orientation, or both (sensed pose, in six degrees of freedom), of a trocar that is producing a magnetic field.” [0006], “a machine learning model is trained to perform regression analysis on the measured sensor readings (information obtained by the sensor readings) to thereby effectively estimate the pose of the trocar. The training process is performed offline, and may require a number of sensor measurements and corresponding known, ground truth pose data. The network is sufficiently trained when the change of the loss converges. To evaluate the performance of the machine learning model, pose is predicted (estimated) on a set of measurement data that has not been used for training and then the predicted pose is compared to the corresponding known ground truth pose to find out the error.” [0063]);
and determining a spatial location and/or orientation of the medical tool within the body, based on the one or more deep learning algorithms, wherein the determination is performed without applying a mathematical model or formula (“The machine learning model is coupled to receive output data of the magnetic field sensors. The machine learning model is trainable to output a three-dimensional sensed position, a three-dimensional sensed orientation, or both (sensed pose, in six degrees of freedom), of a trocar that is producing a magnetic field.” [0006], “A physical/mathematical model to estimate the pose of one or more magnets in a target (such as a trocar) is difficult to determine and may yield incorrect results due to sensor or signal noise, imprecise modelling, or other/unknown magnetic fields. In order to offer an alternative, or improve upon the above-presented surgical robotic system, a further surgical robotic system is described below in which a programmed processor makes the prediction or estimate of the trocar pose relative to the robot arm, by means of a machine learning model, e.g., an artificial neural network, or simply “neural network.” The neural network is trained to predict the pose of the trocar, based on magnetic sensor readings such as in the embodiments described above” [0062]).
Fuesrt does not specifically disclose that the spatial location or orientation is registered to a scan of the subject in real time with a precision of less than about 1.0 mm.
However, in a similar field of endeavor, Besser teaches a guidance system for positioning an insertion device.
Besser also teaches the spatial location or orientation is registered to a scan of the subject in real time (“identify the location of the first and the second anatomic landmarks on the loaded X-ray, CT, ultrasound or MRI image of the subject's chest; aligning the subject coordinate system with the loaded X-ray, CT, ultrasound or MRI image by aligning the registered first and second anatomic landmarks with the location of the first and second anatomic landmarks in the X-ray, CT, ultrasound or MRI image, and display, on the image, a path of the insertion device insertion with respect to the first and the second anatomic landmarks; wherein the path is generated according to changes in the strength of the electromagnetic field sensed by the tip sensor's during the insertion of the insertion device.” [0007])
It would have been obvious to an ordinary skilled person in the art before the effective filing
date of the claimed invention to modify the system of Fuerst as outlined above with the spatial location or orientation is registered to a scan of the subject in real time as taught by Besser, because it provides an electromagnetic positioning guidance system reliably operable regardless of the subject's movement or position [0006].
Fuerst in view of Besser does not specifically teach a precision of less than about 1.0 mm.
However, in a similar field of endeavor, Sebkhi teaches a motion tracking and localization system that can provide a non-line-of-sight motion tracking with millimetric accuracy for a tracer moving in close proximity to a magnetic source [Abstract].
Sebkhi also teaches a precision of less than about 1.0 mm (“The magnetic sensor readings are adjusted to remove background magnetic field readings and provided to a trained neural network or trained machine learning operation to output a localization-associated measurement value (e.g., the localized position or relative position of the magnetic sensor). While one magnetic sensor would be sufficient to provide millimetric measurement for most applications, the algorithm can operate using multiple magnetic sensor readings. In such embodiments, the exemplary motion tracking and/or localization system can employ multiple magnetic sensors, which may be suitable, e.g., in applications where higher fidelity measurements are warranted.” [0006], “the errors observed for the training set have median and Q3 values of 0.76 mm and 1.29 mm, respectively. The validation set (FIG. 4B) has a median error of 0.93 mm and a Q3 error of 1.49 mm. It can be observed that overfitting did not occur (the errors for validation and training sets are similar). In FIG. 4C, the results for the testing set have a median error of 0.83 mm and a Q3 of 1.42 mm. Overall, the results show that 75% of the errors are within 1.5 mm, and virtually all errors are within 3 mm for any position of the tracer in a volume of 8×8×5 cm3.” [0117])
It would have been obvious to an ordinary skilled person in the art before the effective filing
date of the claimed invention to modify the system of Fuerst in view of Besser as outlined above with a precision of less than about 1.0 mm as taught by Sebkhi, because it improves the operation of motion tracking technologies, including permanent magnet localization [0005].
Regarding Claim 2, Fuerst discloses that the magnetic or electromagnetic element is asymmetric (“The first magnet 71 and the second magnet 73 each can have a different axis of polarization, e.g., an axis extending between opposite poles of the respective magnets 71, 73.” [0046]).
Regarding Claim 3, Fuerst discloses that the medical tool is manipulatable in 3 to 6 degrees of freedom (“In this regard, the arrangement of the sensors 55, 57 provides the processor in the control tower 3 with electrical signals corresponding to the magnetic fields B1, B2 according to the algorithm such that a real or sensed pose of the attachment portion 69 of the trocar 63 relative to the docking interface 27 can be determined with respect to six degrees of freedom (DOF): X-axis position, Y-axis position, Z-axis position, X-axis rotation, Y-axis rotation, and Z-axis rotation. In one variation, at least six measurements from the sensors 55, 57 can be used to determine the pose of the trocar 63.” [0052]).
Regarding Claim 4, Fuerst discloses that the magnetic or electromagnetic element is positioned along a middle portion of the medical tool (“ The object, which can be for example the trocar 63—see FIG. 6—could have structural elements such as the flanged upper portion of head 67 and the attachment portion 69 that are made of magnetic material.” [0066], “FIG. 9 is an enlarged schematic of the area 9 identified in FIG. 6.” [0015], “The trocar 63 includes a first magnet 71 and a second magnet 73 producing respective magnetic fields B1, B2 with known properties” [0046], see Fig. 9 for middle positioning of the attachment portion 69.).
PNG
media_image1.png
507
322
media_image1.png
Greyscale
Regarding Claim 5, Fuerst discloses that the magnetic or electromagnetic element is positioned in a distal portion of the medical tool (“ The object, which can be for example the trocar 63—see FIG. 6—could have structural elements such as the flanged upper portion of head 67 and the attachment portion 69 that are made of magnetic material.” [0066], “FIG. 9 is an enlarged schematic of the area 9 identified in FIG. 6.” [0015], “The trocar 63 includes a first magnet 71 and a second magnet 73 producing respective magnetic fields B1, B2 with known properties” [0046], see Fig. 9 for distal positioning of the attachment portion 69.).
PNG
media_image1.png
507
322
media_image1.png
Greyscale
Regarding Claim 6, Fuerst discloses that the magnetic or electromagnetic element is positioned in a proximal portion of the medical tool and wherein the deep learning algorithm is trained to account for the positional effect of the proximal configuration(“a surgical robotic arm having a trocar docking interface at its distal end is manually maneuvered by a user until the interface is adjacent to and aligned with an attachment portion (e.g., a mating interface) on the proximal end of the trocar (outside the patient.)” [0003], “ One or both of the magnets 71, 73 can be embedded in or otherwise coupled to the trocar 63, for example, by being integrally molded therein, by being inserted into a receiving portion thereof, or by being otherwise secured to the trocar 63. “ [0046], “a machine learning model is trained to perform regression analysis on the measured sensor readings (information obtained by the sensor readings) to thereby effectively estimate the pose of the trocar. The training process is performed offline, and may require a number of sensor measurements and corresponding known, ground truth pose data. The network is sufficiently trained when the change of the loss converges. To evaluate the performance of the machine learning model, pose is predicted (estimated) on a set of measurement data that has not been used for training and then the predicted pose is compared to the corresponding known ground truth pose to find out the error.” [0063], Fuerst machine learning model is trained with ground truth poses which would inherently take into account the proximal positional effect since that is where the magnets are located).
Regarding Claim 7, Fuerst discloses comprising a plurality of magnetic and/or electromagnetic elements, each independently positioned in a distal portion, proximal portion, or a middle portion of the medical tool (“ The object, which can be for example the trocar 63—see FIG. 6—could have structural elements such as the flanged upper portion of head 67 and the attachment portion 69 that are made of magnetic material.” [0066], “FIG. 9 is an enlarged schematic of the area 9 identified in FIG. 6.” [0015], “The trocar 63 includes a first magnet 71 and a second magnet 73 producing respective magnetic fields B1, B2 with known properties” [0046], see Fig. 9 for distal positioning of the attachment portion 69.).
PNG
media_image1.png
507
322
media_image1.png
Greyscale
Regarding Claim 8, Fuerst discloses determining the spatial location and/or orientation of the medical tool comprises compensating for a dimension of the medical tool in relation to the location of the magnetic or electromagnetic element in relation to the medical tool (“the processor in the control tower 3 measures a sensed pose of the attachment portion 69 of the trocar 63 with respect to a 3-axis coordinate system, such as a system of X-, Y-, and Z-axes, by measuring and coordinating the electrical signals output by the sensors 55, 57 of the sensor system 47 to determine the relative strength of the magnetic fields B1, B2 of the respective magnets 71, 73 received at different locations, e.g., depths D1, D2, D3, D4, on the sensor boards 49, 51. For example, if the sensors 55, 57 in the column C1 output electrical signals corresponding to the received magnetic fields B1, B2 that is greater than the output electrical signals of the sensors 55, 57 in the column C2, a determination of a depth distance, e.g., an X-axis location, between the attachment portion 69 of the trocar 63 and the docking interface 27 can be calculated.” [0051]).
Regarding Claim 9, Fuerst discloses that the sensor array is configured in cartesian, radial, or cylindrical coordinate system (“the processor in the control tower 3 measures a sensed pose of the attachment portion 69 of the trocar 63 with respect to a 3-axis coordinate system, such as a system of X-, Y-, and Z-axes, by measuring and coordinating the electrical signals output by the sensors 55, 57 of the sensor system 47 to determine the relative strength of the magnetic fields B1, B2 of the respective magnets 71, 73 received at different locations, e.g., depths D1, D2, D3, D4, on the sensor boards 49, 51. For example, if the sensors 55, 57 in the column C1 output electrical signals corresponding to the received magnetic fields B1, B2 that is greater than the output electrical signals of the sensors 55, 57 in the column C2, a determination of a depth distance, e.g., an X-axis location, between the attachment portion 69 of the trocar 63 and the docking interface 27 can be calculated.” [0051]).
Regarding Claim 10, Fuerst discloses that the sensor array is configured to wirelessly associate with the magnetic or electromagnetic element and/or the medical tool (“it will be understood that any of the components described herein can be in communication via wired and/or wireless links, using any suitable ones of a variety of data communication protocols.” [0044]).
Regarding Claim 11, Fuerst discloses that determining the spatial location and/or orientation of the medical tool comprises compensating for static or dynamic background interference (“via optimization by the processor of the estimated sensor readings produced through the deterministic model and the measured sensor readings received from the sensor system 47, the surgical robotic system 1 is operable to discriminate between the magnetic fields B1, B2 that are representative of the pose of the trocar 63 and other magnetic fields or electromagnetic interference such as those produced by other trocars or other surgical equipment in the operating arena.” [0059])
Regarding Claim 12, Fuerst discloses further comprising associating variations in the magnetic field with one or more of a type of tissue, a type of procedure, and a type of medical tool (“the processor in the control tower 3 measures a sensed pose of the attachment portion 69 of the trocar 63 with respect to a 3-axis coordinate system, such as a system of X-, Y-, and Z-axes, by measuring and coordinating the electrical signals output by the sensors 55, 57 of the sensor system 47 to determine the relative strength of the magnetic fields B1, B2 of the respective magnets 71, 73 received at different locations, e.g., depths D1, D2, D3, D4, on the sensor boards 49, 51. For example, if the sensors 55, 57 in the column C1 output electrical signals corresponding to the received magnetic fields B1, B2 that is greater than the output electrical signals of the sensors 55, 57 in the column C2, a determination of a depth distance, e.g., an X-axis location, between the attachment portion 69 of the trocar 63 and the docking interface 27 can be calculated.” [0051]).
Regarding Claim 15, Fuerst discloses that the processor is configured to train the deep learning algorithm on a training set comprising a database associated with the changes of the magnetic field due to a change of one or more coordinates of the magnetic or electromagnetic element (“To train the machine learning model, training data is required that comprises sensor measurements and ground truth poses. In one embodiment described above with reference to FIG. 12, magnets are attached to a trocar, magnetic field sensors are attached to a robot end-effector, and a large number of samples covering the entire workspace and entire range of rotations of each of many positions are obtained. This dataset is used to train the machine learning model, by optimizing the weights for each node in the network graph.” [0072]).
Regarding Claim 16, Fuerst discloses that the database comprises data obtained by receiving one or more signals associated with a change of the magnetic field generated by a change in the spatial location and/or orientation of the magnetic or electromagnetic element between a plurality of pairs of coordinates (“To train the machine learning model, training data is required that comprises sensor measurements and ground truth poses. In one embodiment described above with reference to FIG. 12, magnets are attached to a trocar, magnetic field sensors are attached to a robot end-effector, and a large number of samples covering the entire workspace and entire range of rotations of each of many positions are obtained. This dataset is used to train the machine learning model, by optimizing the weights for each node in the network graph.” [0072]).
Regarding Claim 17, Fuerst discloses that the database comprises data sets obtained using the array of magnetic sensors, and wherein each of the data sets comprises three signals from each individual sensor of the array of magnetic sensors for each change in spatial location or orientation of the magnetic or electromagnetic element (“To train the machine learning model, training data is required that comprises sensor measurements and ground truth poses. In one embodiment described above with reference to FIG. 12, magnets are attached to a trocar, magnetic field sensors are attached to a robot end-effector, and a large number of samples covering the entire workspace and entire range of rotations of each of many positions are obtained. This dataset is used to train the machine learning model, by optimizing the weights for each node in the network graph.” [0072], “ the processor in the control tower 3 measures a sensed pose of the attachment portion 69 of the trocar 63 with respect to a 3-axis coordinate system, such as a system of X-, Y-, and Z-axes, by measuring and coordinating the electrical signals output by the sensors 55, 57 of the sensor system 47 to determine the relative strength of the magnetic fields B1, B2 of the respective magnets 71, 73 received at different locations, e.g., depths D1, D2, D3, D4, on the sensor boards 49, 51. For example, if the sensors 55, 57 in the column C1 output electrical signals corresponding to the received magnetic fields B1, B2 that is greater than the output electrical signals of the sensors 55, 57 in the column C2, a determination of a depth distance, e.g., an X-axis location, between the attachment portion 69 of the trocar 63 and the docking interface 27 can be calculated. Similarly, if the sensors 55, 57 in the row R1 output electrical signals corresponding to the received magnetic fields B1, B2 that is greater than the output electrical signals of the sensors 55, 57 in the columns R2, a determination of a vertical distance, e.g., a Z-axis location, between the attachment portion 69 of the trocar 63 and the docking interface 27 can be calculated.” [0051]).
Regarding Claim 18, Fuerst discloses that the three signals comprise changes in the magnetic field generated in an x-axis, y-axis, and z-axis generated by the magnetic or electromagnetic element, wherein the change in the spatial location and/or orientation of the magnetic or electromagnetic element comprises a translation of the magnetic or electromagnetic element in one or more axes at a specified distance (“To train the machine learning model, training data is required that comprises sensor measurements and ground truth poses. In one embodiment described above with reference to FIG. 12, magnets are attached to a trocar, magnetic field sensors are attached to a robot end-effector, and a large number of samples covering the entire workspace and entire range of rotations of each of many positions are obtained. This dataset is used to train the machine learning model, by optimizing the weights for each node in the network graph.” [0072], “ the processor in the control tower 3 measures a sensed pose of the attachment portion 69 of the trocar 63 with respect to a 3-axis coordinate system, such as a system of X-, Y-, and Z-axes, by measuring and coordinating the electrical signals output by the sensors 55, 57 of the sensor system 47 to determine the relative strength of the magnetic fields B1, B2 of the respective magnets 71, 73 received at different locations, e.g., depths D1, D2, D3, D4, on the sensor boards 49, 51. For example, if the sensors 55, 57 in the column C1 output electrical signals corresponding to the received magnetic fields B1, B2 that is greater than the output electrical signals of the sensors 55, 57 in the column C2, a determination of a depth distance, e.g., an X-axis location, between the attachment portion 69 of the trocar 63 and the docking interface 27 can be calculated. Similarly, if the sensors 55, 57 in the row R1 output electrical signals corresponding to the received magnetic fields B1, B2 that is greater than the output electrical signals of the sensors 55, 57 in the columns R2, a determination of a vertical distance, e.g., a Z-axis location, between the attachment portion 69 of the trocar 63 and the docking interface 27 can be calculated.” [0051]).
Regarding Claim 28, Fuerst discloses that the three signals comprise changes in the magnetic field generated in an x-axis, y-axis, and z-axis generated by the magnetic or electromagnetic element, wherein the change in the spatial location and/or orientation of the magnetic or electromagnetic element comprises a translation of the magnetic or electromagnetic element in one or more axes at a specified distance (“To train the machine learning model, training data is required that comprises sensor measurements and ground truth poses. In one embodiment described above with reference to FIG. 12, magnets are attached to a trocar, magnetic field sensors are attached to a robot end-effector, and a large number of samples covering the entire workspace and entire range of rotations of each of many positions are obtained. This dataset is used to train the machine learning model, by optimizing the weights for each node in the network graph.” [0072], “ the processor in the control tower 3 measures a sensed pose of the attachment portion 69 of the trocar 63 with respect to a 3-axis coordinate system, such as a system of X-, Y-, and Z-axes, by measuring and coordinating the electrical signals output by the sensors 55, 57 of the sensor system 47 to determine the relative strength of the magnetic fields B1, B2 of the respective magnets 71, 73 received at different locations, e.g., depths D1, D2, D3, D4, on the sensor boards 49, 51. For example, if the sensors 55, 57 in the column C1 output electrical signals corresponding to the received magnetic fields B1, B2 that is greater than the output electrical signals of the sensors 55, 57 in the column C2, a determination of a depth distance, e.g., an X-axis location, between the attachment portion 69 of the trocar 63 and the docking interface 27 can be calculated. Similarly, if the sensors 55, 57 in the row R1 output electrical signals corresponding to the received magnetic fields B1, B2 that is greater than the output electrical signals of the sensors 55, 57 in the columns R2, a determination of a vertical distance, e.g., a Z-axis location, between the attachment portion 69 of the trocar 63 and the docking interface 27 can be calculated.” [0051]).
Regarding Claim 42, Fuerst discloses a system for tracking a medical tool during a medical procedure, the system comprising (“FIG. 15 is a process flow of a method of sensing position and/or orientation of an object, which can be performed by embodiments of the surgical robotic system of FIG. 12 and variations thereof.” [0021]):
a medical tool comprising a magnetic or electromagnetic element; an array of magnetic sensors; and a processor configured to: (“an arm-to-trocar docking capability of a surgical robotic system senses position, orientation or both (pose) of the trocar. The surgical robotic system includes a surgical robotic arm, magnetic field sensors on the arm, and a digital processor that implements a machine learning model (e.g., an artificial neural network “neural network”).” [0006], “The system 1 can incorporate any number of devices, tools, or accessories used to perform surgery on a patient 6. For example, the system 1 may include one or more surgical tools 7 used to perform surgery. A surgical tool 7 may be an end effector that is attached to a distal end of a surgical arm 4, for executing a surgical procedure.” [0027], “The trocar 63 includes a first magnet 71 and a second magnet 73 producing respective magnetic fields B1, B2 with known properties” [0046]):
receive one or more signals from the array of magnetic sensors, the one or more signals are indicative of a change and/or disturbances in one or more of: a strength, direction, flux and magnitude of a magnetic vector detected in a magnetic field generated by the magnetic or electromagnetic element within the body of a subject (“The machine learning model is coupled to receive output data of the magnetic field sensors. The machine learning model is trainable to output a three-dimensional sensed position, a three-dimensional sensed orientation, or both (sensed pose, in six degrees of freedom), of a trocar that is producing a magnetic field.” [0006], ”Each sensor, e.g., of magnetic field sensors 55, 57 labeled sensor 1, sensor 2, etc., senses magnetic field in each of three orthogonal directions, and is of a type commonly known as an XYZ or three axis magnetic field sensor. Here, these magnetic field sensors are depicted as each sensing and producing signals for B0, B1 and B2 magnitudes of the magnetic field vector at the location of the sensor.” [0075]);
applying the one or more signals to one or more deep learning algorithms, the algorithms trained on a database comprising signals associated with a plurality of known spatial positions and orientations of the magnetic or electromagnetic element (“The machine learning model is coupled to receive output data of the magnetic field sensors. The machine learning model is trainable to output a three-dimensional sensed position, a three-dimensional sensed orientation, or both (sensed pose, in six degrees of freedom), of a trocar that is producing a magnetic field.” [0006], “a machine learning model is trained to perform regression analysis on the measured sensor readings (information obtained by the sensor readings) to thereby effectively estimate the pose of the trocar. The training process is performed offline, and may require a number of sensor measurements and corresponding known, ground truth pose data. The network is sufficiently trained when the change of the loss converges. To evaluate the performance of the machine learning model, pose is predicted (estimated) on a set of measurement data that has not been used for training and then the predicted pose is compared to the corresponding known ground truth pose to find out the error.” [0063]);
and determine a spatial location and/or orientation of the medical tool within the body, based on the one or more deep learning algorithms, wherein the determination is performed without applying a mathematical model or formula (“The machine learning model is coupled to receive output data of the magnetic field sensors. The machine learning model is trainable to output a three-dimensional sensed position, a three-dimensional sensed orientation, or both (sensed pose, in six degrees of freedom), of a trocar that is producing a magnetic field.” [0006], “A physical/mathematical model to estimate the pose of one or more magnets in a target (such as a trocar) is difficult to determine and may yield incorrect results due to sensor or signal noise, imprecise modelling, or other/unknown magnetic fields. In order to offer an alternative, or improve upon the above-presented surgical robotic system, a further surgical robotic system is described below in which a programmed processor makes the prediction or estimate of the trocar pose relative to the robot arm, by means of a machine learning model, e.g., an artificial neural network, or simply “neural network.” The neural network is trained to predict the pose of the trocar, based on magnetic sensor readings such as in the embodiments described above” [0062]).
Fuesrt does not specifically disclose that the spatial location or orientation is registered to a scan of the subject in real time with a precision of less than about 1.0 mm.
However, in a similar field of endeavor, Besser teaches a guidance system for positioning an insertion device.
Besser also teaches the spatial location or orientation is registered to a scan of the subject in real time (“identify the location of the first and the second anatomic landmarks on the loaded X-ray, CT, ultrasound or MRI image of the subject's chest; aligning the subject coordinate system with the loaded X-ray, CT, ultrasound or MRI image by aligning the registered first and second anatomic landmarks with the location of the first and second anatomic landmarks in the X-ray, CT, ultrasound or MRI image, and display, on the image, a path of the insertion device insertion with respect to the first and the second anatomic landmarks; wherein the path is generated according to changes in the strength of the electromagnetic field sensed by the tip sensor's during the insertion of the insertion device.” [0007])
It would have been obvious to an ordinary skilled person in the art before the effective filing
date of the claimed invention to modify the system of Fuerst as outlined above with the spatial location or orientation is registered to a scan of the subject in real time as taught by Besser, because it provides an electromagnetic positioning guidance system reliably operable regardless of the subject's movement or position [0006].
Fuerst in view of Besser does not specifically teach a precision of less than about 1.0 mm.
However, in a similar field of endeavor, Sebkhi teaches a motion tracking and localization system that can provide a non-line-of-sight motion tracking with millimetric accuracy for a tracer moving in close proximity to a magnetic source [Abstract].
Sebkhi also teaches a precision of less than about 1.0 mm (“The magnetic sensor readings are adjusted to remove background magnetic field readings and provided to a trained neural network or trained machine learning operation to output a localization-associated measurement value (e.g., the localized position or relative position of the magnetic sensor). While one magnetic sensor would be sufficient to provide millimetric measurement for most applications, the algorithm can operate using multiple magnetic sensor readings. In such embodiments, the exemplary motion tracking and/or localization system can employ multiple magnetic sensors, which may be suitable, e.g., in applications where higher fidelity measurements are warranted.” [0006], “the errors observed for the training set have median and Q3 values of 0.76 mm and 1.29 mm, respectively. The validation set (FIG. 4B) has a median error of 0.93 mm and a Q3 error of 1.49 mm. It can be observed that overfitting did not occur (the errors for validation and training sets are similar). In FIG. 4C, the results for the testing set have a median error of 0.83 mm and a Q3 of 1.42 mm. Overall, the results show that 75% of the errors are within 1.5 mm, and virtually all errors are within 3 mm for any position of the tracer in a volume of 8×8×5 cm3.” [0117])
It would have been obvious to an ordinary skilled person in the art before the effective filing
date of the claimed invention to modify the system of Fuerst in view of Besser as outlined above with a precision of less than about 1.0 mm as taught by Sebkhi, because it improves the operation of motion tracking technologies, including permanent magnet localization [0005].
Response to Arguments
Applicant's arguments filed 12/30/2025 have been fully considered but they are not persuasive.
Regarding the U.S.C. 102(a)(1) and 102(a)(2) rejection of Claim 1 the applicant argues the following:
Fuerst merely discloses a machine learning model and does not describe or suggest: a signal database as currently claimed (including a vast number of hundreds of millions of entries); Prediction or extrapolation of spatial location from such data, in particular, without the use of mathematical models; and Real-time mapping to an imaging scan at such high resolution and accuracy.
More specifically, the currently amended claims recite that spatial location and/or orientation is determined "without applying a mathematical model or formula." In contrast, Fuerst relies on traditional field models, distance-based triangulation, or similar physics-based mathematical approaches. Fuerst teaches a hybrid approach that may include machine learning components, but does not teach or suggest determining spatial location purely through data- driven deep learning methods without mathematical formulas.
Furthermore, the currently amended claims recite that the deep learning algorithm is trained on a database of signals associated with known spatial locations and orientations. In contrast, Fuerst does not disclose a large-scale, coordinate-tagged database of magnetic field signal measurements for deep learning training. Instead, it discusses calibration-based prediction
In response to applicant's argument that the references fail to show certain features of the invention, it is noted that the features upon which applicant relies (i.e., a signal database as currently claimed (including a vast number of hundreds of millions of entries)) are not recited in the rejected claim(s). Although the claims are interpreted in light of the specification, limitations from the specification are not read into the claims. See In re Van Geuns, 988 F.2d 1181, 26 USPQ2d 1057 (Fed. Cir. 1993).
It is also noted that Feurst does disclose prediction or extrapolation of spatial location from such electromagnetic data, and specifically mentions the limitations of using mathematical models, instead using machine learning to extrapolate spatial locations (“The surgical robotic system includes a surgical robotic arm, magnetic field sensors on the arm, and a digital processor that implements a machine learning model (e.g., an artificial neural network “neural network”). The machine learning model is coupled to receive output data of the magnetic field sensors. The machine learning model is trainable to output a three-dimensional sensed position, a three-dimensional sensed orientation, or both (sensed pose, in six degrees of freedom), of a trocar that is producing a magnetic field.” [0006], “A physical/mathematical model to estimate the pose of one or more magnets in a target (such as a trocar) is difficult to determine and may yield incorrect results due to sensor or signal noise, imprecise modelling, or other/unknown magnetic fields. In order to offer an alternative, or improve upon the above-presented surgical robotic system, a further surgical robotic system is described below in which a programmed processor makes the prediction or estimate of the trocar pose relative to the robot arm, by means of a machine learning model, e.g., an artificial neural network, or simply “neural network.” The neural network is trained to predict the pose of the trocar, based on magnetic sensor readings such as in the embodiments described above” [0062]).
Feurst also teaches training the machine learning model by providing ground truth spatial postions of electromagnetic elements as well as the magnetic signal they would produce (a machine learning model is trained to perform regression analysis on the measured sensor readings (“information obtained by the sensor readings) to thereby effectively estimate the pose of the trocar. The training process is performed offline, and may require a number of sensor measurements and corresponding known, ground truth pose data. The network is sufficiently trained when the change of the loss converges. To evaluate the performance of the machine learning model, pose is predicted (estimated) on a set of measurement data that has not been used for training and then the predicted pose is compared to the corresponding known ground truth pose to find out the error.” [0063])
Finally Applicant’s arguments with respect to “Real-time mapping to an imaging scan at such high resolution and accuracy.” have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument.
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to STEVEN MALDONADO whose telephone number is 703-756-1421. The examiner can normally be reached 8:00 am-4:00 pm PST M-Th Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at
http://www.uspto.gov/interviewpractice. If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Christopher Koharski can be reached on (571) 272-7230. Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/Steven Maldonado/
Patent Examiner, Art Unit 3797
/CHRISTOPHER KOHARSKI/Supervisory Patent Examiner, Art Unit 3797
Prosecution Insights