WO2015044930A1

WO2015044930A1 - Device specific outlier rejection for stable optical shape sensing

Info

Publication number: WO2015044930A1
Application number: PCT/IB2014/064967
Authority: WO
Inventors: Bharat RAMACHANDRAN; Sander Hans DENISSEN
Original assignee: Koninklijke Philips NV
Current assignee: Koninklijke Philips NV
Priority date: 2013-09-30
Filing date: 2014-09-30
Publication date: 2015-04-02
Anticipated expiration: 2016-03-30

Abstract

A shape sensing system includes a workstation (112) configured to receive optical signals from a shape sensing enabled device (102) having at least one optical fiber (126) and to interpret the optical signals to determine a shape of the shape sensing enabled device. The workstation includes a processor (114); a memory (116) coupled to the processor; and a stability module (140) stored in the memory and configured to determine outlier shape data for data positions that exceed an acceptable threshold, and reject the shape data corresponding to the data positions that exceed the acceptable threshold for rendering as a stable shape sensing data set. A device specific baseline module (145) is configured to store device specific baseline measurements of the shape sensing enabled device such that the baseline measurements are employed to adjust measured shape data to improve determinations of outlier shape data.

Description

Device Specific Outlier Rejection For Stable Optical Shape Sensing

BACKGROUND:

Technical Field

This disclosure relates to shape sensing instruments and more particularly to a system and method for use with shape sensing optical fibers having improved stability by improved outlier rejection. Description of the Related Art

Optical shape sensing (OSS) based on fiber optics exploits the inherent backscatter in a conventional optical fiber. The principle involved makes use of distributed strain measurement in the optical fiber using characteristic Rayleigh backscatter patterns or by employing fiber Bragg gratings (FBGs).

A physical length and index of refraction of a fiber are intrinsically sensitive to environmental parameters, temperature and strain and, to a much lesser extent, pressure, humidity, electromagnetic fields, chemical exposure, etc. The wavelength shift, Δλ or frequency shift, Δν, of the backscatter pattern due to a temperature change, AT, or strain along the fiber axis, ε is: Δλ/ & — -Δν fv — κ_τ ΔΤ + κ_εε, where

The temperature coefficient K_T is a sum of the thermal expansion coefficient

. = (l /A)(dA/ dT ) and the thermo-optic coefficient, ξ = (I /n)(9n/ dT ), with a typical value of 0.55xl0^"6 °C^_1 and a value of 6.1xl0^"6 °C^_1 for germanium-doped silica core fibers. The strain coefficient Κε is a function of group index n, the components of the strain-optic tensor, Pij and Poisson' s ratio, «. Typical values given for n, p₁₂, pn and μ for germanium-doped silica yield a value for Κε of about 0.787. Thus, a shift in temperature or strain is merely a linear scaling (for moderate temperature and strain ranges) of the spectral frequency shift Δν. Naturally, this linear model would not apply if strains approach the elastic limit of the fiber, or temperatures approach the glass transition temperature of the fiber.

OSS, while having the ability to deliver accurate shape reconstructions, can at times become unstable and reconstruct incorrect shapes. For clinical use, it is essential to detect, remove, or correct the inaccurate shapes and in the process, improve the stability and performance of OSS. Furthermore, laser realignment is necessary from time to time, an OSS system needs to be calibrated with the polarization states having a proper phase difference, this being performed for each of optical cores (e.g., four). If not performed properly, the OSS system can become unstable.

The strain and shape measurements reconstructed interferometrically from correlation and phase tracking require a fixed launch region, and a tether proximal to the launch region needs to be held steady. Any motion of these can also result in shape instability. A wobble measurement defines a characteristic of the fiber. Any mechanical compression, tension or pressure at any location on the fiber can change the mechanical properties and position of any of the cores, hence also changing the backscatter pattern from the fiber and resulting in instability. SUMMARY

In accordance with the present principles, a shape sensing system includes a workstation configured to receive optical signals from a shape sensing enabled device having at least one optical fiber and to interpret the optical signals to determine a shape of the shape sensing enabled device. The workstation includes a processor, a memory coupled to the processor; and a stability module stored in the memory and configured to determine outlier shape data for data positions that exceed an acceptable threshold, and reject the shape data corresponding to the data positions that exceed the acceptable threshold for rendering as a stable shape sensing data set. A device specific baseline module is configured to store device specific baseline measurements of the shape sensing enabled device such that the baseline measurements are employed to adjust measured shape data to improve determinations of outlier shape data.

A method for device specific stability control for shape sensing with optical fiber includes integrating an optical fiber shape sensing sensor into a shape sensing enabled device to collect initial shape data from the optical fiber shape sensing device; creating a baseline device specific shape using the initial shape data; collecting shape data from the optical fiber shape sensing device; and adjusting the shape data based on the baseline device specific shape to provide an adjusted shape with improved outlier rejection.

Another method for device specific stability control for shape sensing with optical fiber includes integrating an optical fiber shape sensing sensor into a shape sensing enabled device to collect initial shape data from the optical fiber shape sensing device; creating a baseline shape using the initial shape data; rejecting data positions in the shape data that exceed acceptable thresholds for measured parameters; prior to use, collecting second shape data from the optical fiber shape sensing device; adjusting the second shape data based on the baseline shape to provide an adjusted shape; comparing the initial shape data to the adjusted shape to determine if the shape sensing enabled device with the optical fiber shape sensing sensor is acceptable for use.

These and other objects, features and advantages of the present disclosure will become apparent from the following detailed description of illustrative embodiments thereof, which is to be read in connection with the accompanying drawings. BRIEF DESCRIPTION OF DRAWINGS

This disclosure will present in detail the following description of preferred embodiments with reference to the following figures wherein:

FIG. 1 is a block/flow diagram showing a shape sensing system which employs a stability module in accordance with one embodiment;

FIG. 2 is a block/flow diagram showing the stability module in greater detail in accordance with one embodiment; and

FIG. 3 is a flow diagram showing a method for optical shape sensing with device specific outlier determination capabilities in accordance with illustrative embodiments.

DETAILED DESCRIPTION OF EMBODIMENTS

In accordance with the present principles, systems and methods for device specific outlier rejection are provided for obtaining stable reconstructions for optical shape sensing (OSS). Instability in shape sensing may arise due to a number of factors. A few of these factors may include: loss of laser alignment, phase tracking failure, loss of correlation between sample and reference measurements, motion of a tether patch cord proximal to a launch region, improper calibration, variations in temperature (ambient as well as sharp local variations) relative to the calibrated reference acquisition, local pressure/stress resulting in loss of original geometry of optical cores (any of the multiple cores (e.g., 4)), high twist/roll (e.g., multiple π turns about the fiber's axis), high curvature, vibration, axial tension, etc.

Every device with OSS has a specific pattern of signal to noise, twist and tension accumulation, etc. along the fiber's length. This is sometimes seen near a tip or termination of the device while on other occasions, a high tension is seen close to a bonding at a launch unit. Methods for obtaining a device specific signal patterns and employing the same to optimize outlier (bad shape) rejection for the concerned device are disclosed. In accordance with the present principles, a device specific stability-enhancing method or algorithm for OSS is provided that detects deviant or erroneous shapes, corrects the shapes and, if correction is not possible, eliminates those shapes from a display output. During calibration of an OSS enabled device, either post-integration with an OSS fiber or just prior to clinical use, a pattern of the OSS fiber within the OSS enabled device (such as twist and/or tension at every position) is computed and stored. This process ensures that if a region is under stress, for example, the tip shows a high twist value, but this value does not correspond to bad shapes, then the improved outlier rejection method or algorithm performs a subtraction of measured twist from this initially measured signal and hence permits the OSS enabled device to display a good shape as opposed to throwing away all the shapes as bad. This process can be performed iteratively and on a real-time basis to constantly update the data to improve stability.

Device specificity relies on the differences in optical patterns which are specific to a fiber or fibers, or to the device within which the fiber is embedded. For example, one optical fiber post-calibration may have high twist/optical peaks near the distal end. While comparing twist, the high twist can be accounted for in this distal portion as the baseline/steady state, and deviations from this can be classified as bad shapes. This provides improvements over the alternative where the shape is considered good but has an overall high value in a distal section which renders all shapes above the threshold so that all shapes would be classified as outliers and rejected.

Such a device specific algorithm may be considered as adding a layer of intelligence to outlier rejection, which improves performance and allows detection of some of the causes of failure such as termination error or too much tension at the launch region post-integration (after integrating the sensing fibers in the instrument) of the OSS fiber into the OSS enabled device. Different parameters may be employed for outlier rejection and include but are not limited to: twist/roll (node twist, cumulative twist and/or differential of nodal twist), spatiotemporal continuity of the twist, prior knowledge exploiting spatiotemporal smoothness of instrument position, velocity, and acceleration, temporal filtering (such as averaging), bend single or bend angle, any combination of the above. Combinations of the above parameters are likely to yield better results.

Furthermore, when a portion of the shape (distal) is erroneous while a proximal shape is accurate, the algorithm can detect the fiber index/node where a low confidence

measurement arises, and display a portion of the shape and/or correct a remainder of the shape and display the same and/or display an earlier correct shape to substitute for the incorrect portion.

It should be understood that the present invention will be described in terms of medical instruments; however, the teachings of the present invention are much broader and are applicable to any fiber optic instruments. In some embodiments, the present principles are employed in tracking or analyzing complex biological or mechanical systems. In particular, the present principles are applicable to internal tracking procedures of biological systems, procedures in all areas of the body such as the lungs, gastro-intestinal tract, excretory organs, blood vessels, etc. The elements depicted in the FIGS, may be

implemented in various combinations of hardware and software and provide functions which may be combined in a single element or multiple elements.

The functions of the various elements shown in the FIGS, can be provided through the use of dedicated hardware as well as hardware capable of executing software in association with appropriate software. When provided by a processor, the functions can be provided by a single dedicated processor, by a single shared processor, or by a plurality of individual processors, some of which can be shared. Moreover, explicit use of the term "processor" or "controller" should not be construed to refer exclusively to hardware capable of executing software, and can implicitly include, without limitation, digital signal processor ("DSP") hardware, read-only memory ("ROM") for storing software, random access memory

("RAM"), non-volatile storage, etc.

Moreover, all statements herein reciting principles, aspects, and embodiments of the invention, as well as specific examples thereof, are intended to encompass both structural and functional equivalents thereof. Additionally, it is intended that such equivalents include both currently known equivalents as well as equivalents developed in the future (i.e., any elements developed that perform the same function, regardless of structure). Thus, for example, it will be appreciated by those skilled in the art that the block diagrams presented herein represent conceptual views of illustrative system components and/or circuitry embodying the principles of the invention. Similarly, it will be appreciated that any flow charts, flow diagrams and the like represent various processes which may be substantially represented in computer readable storage media and so executed by a computer or processor, whether or not such computer or processor is explicitly shown.

Furthermore, embodiments of the present invention can take the form of a computer program product accessible from a computer-usable or computer-readable storage medium providing program code for use by or in connection with a computer or any instruction execution system. For the purposes of this description, a computer-usable or computer readable storage medium can be any apparatus that may include, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device. The medium can be an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system (or apparatus or device) or a propagation medium. Examples of a computer-readable medium include a semiconductor or solid state memory, magnetic tape, a removable computer diskette, a random access memory (RAM), a read-only memory (ROM), a rigid magnetic disk and an optical disk. Current examples of optical disks include compact disk - read only memory (CD-ROM), compact disk - read/write (CD-R/W), Blu-Ray™ and DVD.

Referring now to the drawings in which like numerals represent the same or similar elements and initially to FIG. 1, a system 100 for optical shape sensing is illustratively shown in accordance with one embodiment. System 100 may include a workstation or console 112 from which a procedure is supervised and/or managed. Workstation 112 preferably includes one or more processors 114 and memory 116 for storing programs and applications. Memory 116 may store an optical sensing and interpretation module 115 configured to interpret optical feedback signals from a shape sensing device or system 104. Optical sensing module 115 is configured to use the optical signal feedback (and any other feedback) to reconstruct deformations, deflections and other changes associated with a medical device or instrument 102 and/or its surrounding region. The instrument 102 may include a catheter, a guidewire, a probe, an endoscope, a robot, an electrode, a filter device, a balloon device, or other medical component, etc. In one embodiment, instrument 102 is simply an elongated flexible device including a plurality of optical fibers for conducting shape sensing. The instrument 102 is advanced into a volume or subject 131 to take on a shape which will be measured and corrected in accordance with the present principles.

The shape sensing system 104 on instrument 102 includes one or more optical fibers 126 which are coupled to the instrument 102 in a set pattern or patterns. The optical fibers 126 connect to the workstation 112 through cabling 127. The cabling 127 may include fiber optics, electrical connections, other instrumentation, etc., as needed.

Shape sensing system 104 with fiber optics may be based on fiber optic Bragg grating sensors. A fiber optic Bragg grating (FBG) is a short segment of optical fiber that reflects particular wavelengths of light and transmits all others. This is achieved by adding a periodic variation of the refractive index in the fiber core, which generates a wavelength-specific dielectric mirror. A fiber Bragg grating can therefore be used as an inline optical filter to block certain wavelengths, or as a wavelength- specific reflector.

A fundamental principle behind the operation of a fiber Bragg grating is Fresnel reflection at each of the interfaces where the refractive index is changing. For some wavelengths, the reflected light of the various periods is in phase so that constructive interference exists for reflection and, consequently, destructive interference for transmission. The Bragg wavelength is sensitive to strain as well as to temperature. This means that Bragg gratings can be used as sensing elements in fiber optical sensors. In an FBG sensor, the measurand (e.g., strain) causes a shift in the Bragg wavelength.

One advantage of this technique is that various sensor elements can be distributed over the length of a fiber. Incorporating three or more cores with various sensors (gauges) along the length of a fiber that is embedded in a structure permits a three dimensional form of such a structure to be precisely determined, typically with better than 1 mm accuracy. Along the length of the fiber, at various positions, a multitude of FBG sensors can be located (e.g., 3 or more fiber sensing cores). From the strain measurement of each FBG, the curvature of the structure can be inferred at that position. From the multitude of measured positions, the total three-dimensional form is determined.

As an alternative to fiber-optic Bragg gratings, the inherent backscatter in

conventional optical fiber can be exploited. One such approach is to use Rayleigh scatter in standard single-mode communications fiber. Rayleigh scatter occurs as a result of random fluctuations of the index of refraction in the fiber core. These random fluctuations can be modeled as a Bragg grating with a random variation of amplitude and phase along the grating length. By using this effect in three or more cores running within a single length of multi- core fiber, the 3D shape and dynamics of the surface of interest can be followed.

Workstation 112 may include a display 118 for viewing internal images of a subject (patient, mechanical system, model or test platform, etc.), if an imaging system 110 is employed. The display 118 may be employed to view a rendering of shape data from the shape sensing system 104. Imaging system 110 may include a magnetic resonance imaging (MRI) system, a fluoroscopy system, a computed tomography (CT) system, etc. Display 118 may also permit a user to interact with the workstation 112 and its components and functions, or any other element within the system 100. This is further facilitated by an interface 120 which may include a keyboard, mouse, a joystick, a haptic device, or any other peripheral or control to permit user feedback from and interaction with the workstation 112. The shape sensing data and the imaging data may be registered as a method of verification for the shape sensing data. An image processing program 130 may be employed to permit manipulation of images and data to provide verification of the positional shape sensing data.

Workstation 112 includes an optical source 106 to provide optical fibers with light (e.g., laser light). An optical interrogation unit or module 108 is employed to detect light returning from all fibers. This permits the determination of strains or other parameters, which will be used to interpret the shape, orientation, or other characteristics, sensed by the system 104. The light signals will be employed as feedback to make adjustments to access errors and to calibrate the shape sensing system 104 or system 100.

In accordance with one embodiment, a stability module 140 is included in memory 116. The stability module 140 detects erroneous shapes, corrects the erroneous shapes, or, if correction is not possible, eliminates the shapes from displayed output. The stability module 140 detects errors and corrects shapes by employing parameters. The parameters described below may be employed together or separately. Each parameter set may be more adaptable to a given contortion or shape of the fiber, and different portions of the fiber may be described by different parameters to express the same shapes. Some of the parameters may include but are not limited to, e.g.: - Cartesian parameters, e.g., nodal position, nodal velocity, and/or nodal

acceleration. Nodal position is a position at a node i and at time t=j, and can be given by xy along the length of the fiber. As the node increases, i increases from i = 1 until a last index (e.g., N). Nodal position may be expressed as an equation, a string of segments or elements in a numerical method expression. Nodal velocity, e.g., dx/dt of i,j, includes a difference (numerical method) or differential (analytical method) with respect to time t=j. Nodal acceleration is the second derivative of the nodal position performed to get nodal

acceleration. It should be understood that the expressions with x extend to y and z directions as well. For example the position has x, y and z components, e.g., Xy, Yy, ¾_.

- Bend angle space, e.g., node angular position, angular velocity and/or angular acceleration. Bend angle is illustrated where Pi and P₁₊₁ can denote planes drawn through node indexes i and i+1 parallel to a cross section of the optical fiber. Angles may be in multiple axes. Here, the computation can be analytical or numerical as described above. The angular position may be expressed as a formula (analytical) or nodal elements (numerical). The velocity is a first derivative or difference, and the acceleration is a second derivative or second difference of the position relationship.

- Twist and curvature or derived parameters (these include twist, roll, pitch yaw, torsion, etc.). Twist will be referred to herein as a catch-all descriptor for curving in three dimensions (Pitch, Roll, Yaw). Twist can be defined as any rotation (roll, pitch and/or yaw) of the local coordinate systems between node locations. Twist is the effect descriptive of the wobble/pitch/frequency of rotation of the fiber cores of the shape sensing device. Roll around an axis may be considered different from bend angles. Physically, it would be far more difficult to have high rates of twist changes than to have high rates of bend changes, but both could be used for filtering purposes.

Twist and bend angle should be differentiated. For example, in the event of flopping motion, say at the distal end due to flow of fluid/blood, a rapid change in the bend angle is introduced but may not change the twist. If this results in an outlier, bend angle will pick this up and can reject that shape, or vice versa; the bend angle may be an accurate measurement, so the shape should not be rejected. Another example is in a section where bend angle is restricted, like in an introducer (portion through which device enters the body without allowing blood to flow out). Here, we can constrain the bend angle and if a change is noticed, the shape can be classified as an outlier and be rejected. This case may also not be picked up by twist.

- Temporal filtering such as temporal averaging, with or without filter thresholds. - Forward temporal filtering. This filtering passes a suspected incorrect value to the visualization unit (display) only after processing a next sample or samples. If the last good sample and the latest sample are very similar in characteristics and an intermediate sample or samples are not, it can be determined that the intermediate samples are incorrect. This is a useful feature because the likelihood of a rapid change is higher than the likelihood of a rapid change with a subsequent return to the previous state. This filtering needs a frame rate where a total allowed latency of the system is more than a processing latency plus N times the sample time.

In one example, forward temporal filtering may be thought of as a buffer that has bad shapes. Suppose a system has a frame rate of 500Hz, and hence a frame time of 2 msec. If we know that a maximum allowed latency that an observer can notice ranges from 100 to 200 msec (for X-ray it is about 180msec), a buffer can be built up that ranges from 50 to 100 shapes (calculated as 100msec/2msec to 200msec/2msec). This range is non-limiting and for explanation only.

- Processing based on additional prior knowledge or other computational models 142 which capture spatio-temporal constraints on instrument position and dynamics, e.g., velocity, acceleration, or shape evolution. These models 142 may include possible shape configurations that are possible for the optical fibers. Shapes that exceed these models (plus a tolerance) may be deemed unacceptable or erroneous. For example, a model may include a unit length of the shape sensing system 104. Due to its geometry, (e.g., its thickness) a radius of less than 1.0 mm may not be physically possible. If the shape sensing data results in a kink that exceeds this radius, high confidence would exist that the data is erroneous and can be eliminated. Bending or curvature (e.g., in the x and y direction) may be employed in such measurements used.

In one embodiment, information about the physical constraints of the device or environment may be employed. For example, if a lumen in which the OSS is placed is sufficiently small, a bend at a position that does not match the allowed curvature of the device (e.g., a reinforced straight section or an introducer), the data is likely erroneous. Similarly if the roadmap of a device (i.e., the anatomy of the patient) is known, the likelihood that the device cuts through several layers of tissue between frames is very low and can be used for error determination.

- Any combination of the above parameters may be employed as well as others. For example, the bend angle parameters may be employed for a gradual bending section of the fiber while the twist parameters may be employed for a spiralling section of the same fiber.

- Other parameters including statistical or historic parameters may be employed (such as, e.g., a positive prediction based on previously observed shapes).

Using a combination of the above parameters provides the best results. It should be understood that the optical fiber system is modeled based upon a series of nodes placed along a length of the optical fiber. These nodes are monitored to obtain relative distances and positions with respect to other nodes on the same optical fiber or between optical fibers or to a reference system.

The stability module 140 can be run in real-time, this may be important for some applications such as cardiac interventions (where the heart is beating). In one example, when a portion of a shape (distal) is erroneous while a proximal shape is accurate, the stability module 140 can: 1) detect the fiber index/node where the instability arises, and display a portion of the shape; and/or 2) correct the reminder of the shape and display the correct shape; and/or 3) display an earlier correct shape for the incorrect portion, etc. The stability module 140 may run in real-time during a procedure or may be switched off.

The stability module 140 includes a device specific baseline module 145. The device specific baseline module 145 includes baseline shapes or images which can be employed in the determination and reconstruction of data for OSS. Every device 102, after being integrated with an OSS sensor (104) (post-integration), has a specific pattern including, e.g., signal to noise, twist, tension accumulation, etc. along the fiber's length. This is sometimes seen near the tip/termination while on other occasions, a high tension is seen close to the bonding at the launch unit. A device specific signal pattern is collected and stored in the device specific baseline module 145 and employed to optimize outlier (bad shape) rejection for the concerned device. The device specific baseline can be employed a signature to differentiate different devices or to differentiate different signals, e.g., if multiple OSS devices are employed.

Either immediately after integration into an OSS instrument or just prior to clinical use, a signal may be acquired and used to determine the optimal signal characteristics of the OSS enabled device 102 and stored in the device specific baseline module 145. For example, a high twist value near a termination while displaying stable shapes may be calibrated out by the stability module 140 using the device specific baseline module 145, by performing a temporal difference using the initial measurement. Similarly, a high tension near the launch region can be detected, and if this results in outliers, then a flag may be raised post- calibration that the integration in the launch region is suspect, and the device should not be used in its current configuration. The OSS sensor 104 may be re-bonded/re-integrated into the OSS enabled device 102 to solve this problem.

Furthermore, when a portion of the shape (distal) is erroneous while the proximal shape is accurate, the algorithm/stability module 140 can detect the fiber index/node where the instability arises, and display a portion of the shape and/or correct the remainder of the shape and display the same and/or display an earlier correct shape (from the baseline 145 or previous data) of the incorrect portion. The device specific baseline module 145 may also store thresholds and stability criteria for the specific device. By way of non-limiting example, two illustrative devices, a catheter and an ultrasound probe, may be employed during a procedure. Both have different features. One (the catheter) has a known (and changing) curvature at the distal end, while the other (the probe) is relatively rigid (if OSS is inside both devices). Hence, the baseline stored in module 145 for the OSS fiber within these two devices and the constraints for each would be different.

An implementation of the stability module 140 with the device specific baseline module 145 is shown and described with respect to FIG. 2.

Referring to FIG. 2, one implementation of an algorithm for improving stability of a shape sensing system is described for detecting whether a shape is "good" or "bad" and then modifying the output shape measurement stream accordingly. FIG. 2 is presented as a plurality of blocks, which may be implemented in any order and may be implemented in software or hardware, as needed. In one embodiment, the method of FIG. 2 is carried out by the stability module 140 and/or the device specific baseline module 145. In block 202, sample shape and OSS parameters are read into an interpretation module (e.g., module 115 in FIG. 1) from the OSS enabled device (post-integration of the OSS sensor in the OSS enabled device) as reflected optical signals (feedback). This initial data collection may be employed in a determination of an initial device specific baseline for the OSS enabled device. The OSS sensor may be removed for shipment or storage. In block 204, data outliers in parameters (e.g., differential twist, tension, pivoting, accuracy, termination and/or SNR) are checked to determine if a parameter or parameters are within a defined or desired threshold. For example, a threshold may be exceeded for a twist of greater than +/- 360 degrees.

If the device checks out, then, the OSS sensor can be re-integrated into the OSS enabled device and shipped or stored. In block 206, the sample shape and OSS parameters are read in from the OSS enabled device prior to clinical use. In block 208, the pre-clinical values of OSS parameters are compared with post-integration values. If outliers are present based on the pre-defined criteria, damage or sub-optimal performance may be due to changes during transport, abuse or lack of use. The OSS enabled device may need to be repaired or discarded.

In block 209, a reference OSS measurement is obtained in the clinical environment. This may be performed under certain conditions, e.g., a normal temperature or other environmental conditions of a clinical environment. In block 210, the reference measurement may be rejected (e.g., classified as an outlier) under one or more of the following conditions, which are illustrative and can be changed/configured as needed (e.g., during a procedure):

1) if cumulative twist is greater than "+" two pi (360 degrees) or less than "-" two pi; or 2) if SNR is low in one or more of the OSS fiber cores; or 3) if an absolute of maximum of differential twist is greater than a set threshold; or 4) if spatiotemporal continuity of a series of reference measurements is poor; or 5) if the reference measurement can be classified as bad by any other means including but not limited to data from imaging or tracking such as magnetic resonance (MR), X-ray, compute tomography (CT), ultrasound (US), positron emission spectroscopy (PET)/ spectral positron emission computed tomography (SPECT), optical, electromagnetic (EM), infrared (IR), etc. or if a user/clinician defines the reference measurement as bad by observing the same; or 6) if the reference measurement is suspect due to improper calibration such as loading an incorrect tether file, reference file, wobble file or configuration file, etc.

In block 212, OSS data is read to be displayed to a user. Shapes may be rejected by the user if, e.g., the cumulative twist or other parameters are above a threshold. In block 214, shapes that are observably incorrect (e.g., by a user) based on anatomical and device constraints are rejected (e.g., some shapes are impossible for the OSS enabled device). In block 216, shapes are rejected based on low signal to noise ratio or different (high peaks) in optical signal such as those indicating a break in termination. In block 218, shapes are rejected based on absolute differential twist exceeding a given threshold. In block 220, a comparison between the differential twist (and/or cumulative and nodal twist) of the current OSS shape is made against the reference (e.g., the initial measurement), and if the absolute difference between the two is greater than a defined threshold or amount, then the shape is rejected. In block 221, shapes are rejected based on bend angle exceeding a given threshold. In block 222, a current measured OSS shape (with e.g., the differential and/or cumulative twist, nodal twist, etc.) may be compared with a previous (classified as "good") shape to dynamically update the reference (initial measurement or other reference). If the absolute of the difference between the two is greater than a defined threshold, then the shape is rejected.

In block 224, while all blocks in FIG. 2 may be iteratively executed, blocks 210, 220 and/or 222 may be intermittently repeated on a pre-defined basis with a different

decimation/scale or power. The received signals will be changed accordingly (with the different decimation/scale or power) making bad and/or good data easier to distinguish since parameters, e.g., cumulative, nodal and differential twist signals will be changed.

In block 226, a user may choose to ignore certain nodes in the OSS measurement (such as a high reflection point just prior to a termination) and adjust the outlier rejection algorithm accordingly. For example, the user through an interface may enable/disable nodes in the shape data. These nodes may be identified to be ignored based upon the device specific baseline for the given device. In block 228, a predictive filtering method may be added for outlier rejection. When OSS shapes are rejectable in a repetitive or predictable fashion, the filter predicts these and displays correct shapes. For example, if the device is close to the heart and fails in a specific phase of the cardiac cycle, a forward predictive filter, such as, a kalman filter may be added to determine and correct for these outliers. In block 230, an image based correction scheme may be added wherein the OSS shape in the device is compared to the shape from an image (e.g., X-ray), and if an outlier is observed, then the OSS displays a correct shape instead of the outlier based on the shape from the X-ray image (or from a previous OSS shape). In block 232, in the case of multiple tether tracking (when more than one OSS enabled device is used in conjunction, such as while using an OSS enabled guidewire within an OSS enabled catheter), one OSS device may be used to correct for outliers in the other OSS device.

The following blocks (234-242) may be implemented independently or provide independent features when or as needed or selected. In block 234, the algorithm and display unit may have an enable/disable feature allowing the user/clinician to decide whether or not to use the device specific outlier rejection algorithm. In block 236, the algorithm has a spatial and/or temporal interpolation/extrapolation feature, wherein when a good shape appears at a distance from a previous good shape due to rejected shapes in between, OSS reconstructions are interpolated in space (with or without adding a delay in time) such that a jump or discontinuity in position is not visible to the user and a smooth continuous motion appears instead of outliers.

In block 240, additionally, the outlier correction mechanism can be based on the previous instance of shape, wherein when an outlier is detected, the bad node(s) causing the bad shapes is/are overwritten with good data, and the remaining good nodes are employed by recalculating the position (for example by multiplying a step size with a cross product of orientation vectors in the two directions, say, x and y). In block 242, similar to block 240, but where the correction mechanism involving overwriting the bad nodes with good data based on recalculation is based on another tether or OSS device from the same instant or a previous instant in time. The mechanism is provided to perform outlier rejection based on optical signals from the shape sensing console or any other method used in combination with the steps described above.

Referring to FIG. 3, a method for device specific stability control for shape sensing with optical fiber is depicted in accordance with the present principles. This method may also be employed using the stability module 140 and device specific baseline module 145 of FIG. 1. The method of FIG. 3 describes several scenarios and may be employed with some or all of the blocks. In addition, the method may be employed to subtract out device specific baseline data from a real-time OSS measurement, test an OSS device before clinical use, simply stabilize the collection of OSS data for any use, etc.

In block 302, an optical fiber shape sensing sensor is integrated into a shape sensing enabled device to collect initial shape data from the optical fiber shape sensing device. The data may be collected from a plurality of nodes or nodal points defined along the length of the device or an individual fiber of the device. In block 304, a baseline shape is created using the initial shape data. The device specific measurements are employed to improve the outlier filtering process. This may lead to either a reduction of outliers, but can also result in an increase in outliers. For example, if the device specific baseline is strong (say a low signal), and is employed as a reference. The number of outlier shapes being rejected may increase.

In block 306, data positions are rejected in the shape data that exceed acceptable thresholds for measured parameters. The acceptable threshold is preferably determined based upon a shape sensing model or criteria, or based upon previously collected shape data. The acceptable thresholds may be based on geometrical expectations of the shape data. Such tests may include hypothesis tests which compare measured data to training data or models.

Thresholds may account for geometric impossibilities, safe bending radii, etc. In addition, information about a structure in which the shape sensing device is contained may be employed as a model or to provide thresholds or constraints on geometrically possible configurations. In one embodiment, the thresholds represent a confidence score either set by default, by a constraint or by a user to account for a confidence that the shape data is valid. The acceptable thresholds may be determined based upon a shape sensing model (e.g., a statistical model, a physiological model, etc.). The acceptable thresholds may be determined based upon previous collected shape data. This may include data collected at a same location but at a different time.

In block 308, in one embodiment, prior to use, second shape data is collected from the optical fiber shape sensing device. This may include pre-clinical usage. In block 310, the second shape data is adjusted based on the baseline shape to provide an adjusted shape. In another embodiment, the baseline (device specific) is employed directly to adjust real-time measured data. In block 311, shape data may be adjusted by ignoring user-selected (or computer- selected) nodes in the OSS measurement to adjust outlier rejection.

In block 312, in one embodiment, the initial shape data is compared to the adjusted shape to determine if the shape sensing enabled device with the optical fiber shape sensing sensor is acceptable for use.

In block 314, outlier data is determined during operational measurements with the shape sensing enabled device. In one embodiment, outlier data is determined in real-time during measurements with the shape sensing enabled device by subtracting out the baseline device specific shape.

Outlier data is removed from the shape data. The outlier data may be defined in a plurality of ways. In a general sense, outliers may include individual data points that appear in the data and are not logically connected to other points. This may include sets of data or entire segments of the collected data set. The outlier data may be replaced by computing spatio-temporal continuity along measured shape data in block 320, computing a maximum cumulative twist, nodal twist, differential twist, bend angle, etc. along measured shape data in block 322, and then reconstructing the shape data by replacing rejected shape data with valid data in block 324.

Spatio-temporal continuity may be computed along the shape data. This may include a node by node check in each coordinate axis to determine if the nodes are continuous. If a discontinuity exists, the discontinuous nodes will be identified. Twist, e.g., a maximum cumulative twist along the shape data, etc. may also be computed. Certain twists or rolls are not possible due to the material and dimensions of the shape sensing device, and due to a volume containing the shape sensing device. Also, since the shape sensing device includes optical fibers, exceeding certain bending radii will cause the shape sensing device to fail or not provide reasonable results. Node twisting is computed to determine if an impossible or unlikely twist is being measured. If the twist can be so determined, the sensor data can be identified. An adaptive search of nodal twist or roll between nodes in the shape data to compare to expected limits may be performed. This compares a shape of a twist to a data base of shapes to determine in the twist is possible (within a confidence limit). If the twist exceeds a threshold the data is identified. A differential twist along the shape data may be determined. This looks at peaks or spikes in the data by differentiating the nodal twist data. This result can be compared to the accumulated twist as well. Jitter or other effects may be filtered out. Other tests may also be performed in addition to or instead of those described herein.

In block 326, rejected shape data may be replaced by employing previously collected acceptable data. In block 332, bad or missing data may be reconstructed by interpolating or extrapolating acceptable data to determine data between discontinuities in the shape data. In block 334, the shape data may be verified by comparing the shape data to data collected by an imaging system or by an additional shapes sensing system. Other verification techniques are also contemplated.

The shape data corresponding to the data positions that exceed an acceptable threshold (erroneous data) are rejected. These include data that has been tagged or identified as falling outside of the acceptable thresholds. Some of this data can be corrected or replaced. Erroneous shape data is corrected. The shape data may be reconstructed by replacing rejected shape data with valid data. The valid data may be from a model or from previously collected acceptable data. The rejected shape data may be replaced by

extrapolating or interpolating other acceptable data. This permits the reconstruction or patching of rejectable data in a dataset without rejecting a larger amount of data.

In block 340, acceptable shape data is rendered to provide a stable shape sensing data set. This may include displaying acceptable shape data on a display, storing stable shape sensing data or employing stable shape sensing data for a given application.

In interpreting the appended claims, it should be understood that:

a) the word "comprising" does not exclude the presence of other elements or acts than those listed in a given claim;

b) the word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements;

c) any reference signs in the claims do not limit their scope;

d) several "means" may be represented by the same item or hardware or software implemented structure or function; and

e) no specific sequence of acts is intended to be required unless specifically indicated. Having described preferred embodiments for device specific outlier rejection for stable optical shape sensing (which are intended to be illustrative and not limiting), it is noted that modifications and variations can be made by persons skilled in the art in light of the above teachings. It is therefore to be understood that changes may be made in the particular embodiments of the disclosure disclosed which are within the scope of the embodiments disclosed herein as outlined by the appended claims. Having thus described the details and particularity required by the patent laws, what is claimed and desired protected by Letters Patent is set forth in the appended claims.

Claims

CLAIMS:

1. A shape sensing system, comprising:

a workstation (112) configured to receive optical signals from a shape sensing enabled device (102) having at least one optical fiber (126) and to interpret the optical signals to determine a shape of the shape sensing enabled device, the workstation including:

a processor (114);

a memory (116) coupled to the processor;

a stability module (140) stored in the memory and configured to determine outlier shape data for data positions that exceed an acceptable threshold, and reject the shape data corresponding to the data positions that exceed the acceptable threshold for rendering as a stable shape sensing data set; and

a device specific baseline module (145) configured to store device specific baseline measurements of the shape sensing enabled device such that the baseline

measurements are employed to adjust measured shape data to improve determinations of outlier shape data.

2. The system as recited in claim 1, wherein the acceptable threshold is provided from at least one (142) of a shape sensing model or previously collected shape data.

3. The system as recited in claim 1, wherein the stability module (140) is configured to compute spatio-temporal continuity along the shape data.

4. The system as recited in claim 1, wherein the stability module (140) is configured to compute at least one of cumulative twist, nodal twist, differential twist and bend angle along the shape data.

5. The system as recited in claim 1, wherein the stability module (140) is configured to reconstruct the shape data by one or more of: replacing rejected shape data with previously collected valid data and/or interpolating or extrapolating a value using acceptable data.

6. The system as recited in claim 1, further comprising an imaging system (110) for verifying the shape data by comparing the shape data to data collected by the imaging system.

7. The system as recited in claim 1, wherein the device specific baseline module (145) includes a specific signal pattern received from the at least one optical fiber of the shape sensing enabled device.

8. The system as recited in claim 7, wherein the measured shape data is iteratively adjusted by the specific signal pattern in real-time.

9. The system as recited in claim 1, wherein the device specific baseline module (145) includes information about physical constraints of the shape sensing enabled device to be employed in identifying outlier data.

10. A method for device specific stability control for shape sensing with optical fiber, comprising:

integrating (302) an optical fiber shape sensing sensor into a shape sensing enabled device to collect initial shape data from the optical fiber shape sensing device; creating (304) a baseline device specific shape using the initial shape data;

collecting (308) shape data from the optical fiber shape sensing device; and adjusting (310) the shape data based on the baseline device specific shape to provide an adjusted shape with improved outlier rejection.

11. The method as recited in claim 10, further comprising determining (314) outlier data in real-time during measurements with the shape sensing enabled device by subtracting out the baseline device specific shape.

12. The method as recited in claim 11, wherein determining outlier data includes computing spatio-temporal continuity along measured shape data.

13. The method as recited in claim 11, wherein determining outlier data includes computing at least one of a maximum cumulative twist, nodal twist, differential twist and bend angle along measured shape data.

14. The method as recited in claim 11, further comprising reconstructing the shape data by replacing rejected shape data with valid data.

15. The method as recited in claim 14, wherein replacing rejected shape data includes employing previously collected acceptable data.

16. The method as recited in claim 14, wherein reconstructing includes interpolating or extrapolating acceptable data to determine data between discontinuities in the shape data.

17. The method as recited in claim 10, further comprising verifying the shape data by comparing the shape data to data collected by an imaging system.

18. The method as recited in claim 10, further comprising adjusting shape data by ignoring user-selected nodes in the OSS measurement to adjust outlier rejection.

19. A method for device specific stability control for shape sensing with optical fiber, comprising:

integrating (302) an optical fiber shape sensing sensor into a shape sensing enabled device to collect initial shape data from the optical fiber shape sensing device;

creating (304) a baseline shape using the initial shape data;

rejecting (306) data positions in the shape data that exceed acceptable thresholds for measured parameters;

prior to use, collecting (308) second shape data from the optical fiber shape sensing device;

adjusting (310) the second shape data based on the baseline shape to provide an adjusted shape; and

comparing (312) the initial shape data to the adjusted shape to determine if the shape sensing enabled device with the optical fiber shape sensing sensor is acceptable for use.

20. The method as recited in claim 19, further comprising determining (314) outlier data in real-time during measurements with the shape sensing enabled device by subtracting out the baseline device specific shape.