Design of a Tracked Bipolar Stimulator



Design of a Tracked Bipolar Stimulator



Submitted by: Greg Wempe

April 24, 2001

Advisor: Dr. Robert L. Galloway, Ph.D.

Graduate Student: Steven L. Hartmann

Professor: Dr. Paul H. King, Ph.D., P.E.

BME 273: Senior Design

Vanderbilt University

Department of Biomedical Engineering

Abstract

The solution-neutral problem statement is that there does not exist a way to track a bipolar stimulator and use it to register the desired points. The goals of this project was to implement a tracking system that would integrate into the existing ORION® software package. This goal was accomplished through consultation with experienced advisors and machinists, extensive literature research, detailed problem analysis, consideration of safety, selection of final solution approach, integration of previously-developed technology with newly-machined additions, and testing of final prototype. These processes were applied to the three main divisions of this project: tracking the bipolar stimulator, developing a carriage for the stimulator, and reducing the stimulator lead flexion. A previously implemented localizer-probe was connected to the bipolar stimulator with an existing carriage clamp. A Delran sheath was then designed and machined to reinforce the flexible leads of the stimulator. Testing of the newly implemented, sheathed system revealed a significant decrease in the overall error of the system in comparison to the unsheathed system. In conclusion, the prototype developed met all of the specifications established for the project but requires continued work to increase the accuracy of the system to its full potential.

Introduction

Bipolar stimulators (A-1, Fig.1) are common tools implemented for cortical mapping. Stimulation of the surface of the brain is accomplished by the depolarization of underlying nerves using variable amounts of current. The depolarization results in the induction of an action potential, which propagates along the nerve pathway and results in an observable response [1]. It is these responses that allow for a mapping of the various areas of the brain associated with different functions such as speech and voluntary motor movement. There is a desire to establish a relationship between the cortical mapping obtained intraoperatively and preoperative images. The fulfillment of this desire would require the development of a method by which to track the movement of the bipolar stimulator in three-dimensional space and to register that tracking information with respect to the previously obtained images..

Thus, the establishment of the focus of this project: to design a method of tracking a bipolar stimulator in three-dimensional space. The Technology-Guided Therapy (TGT) group at Vanderbilt University has previously solved the registration portion of the problem. They have developed a software package (ORION®) capable of registering tracked points with respect to a variety of types of preoperative images (MRI, CT, PET, and SPECT) [2].

An initial literature review suggested a variety of possible ways for tracking the stimulator. The first conceivable option was the combination of a stereotactic frame and an articulated arm [3]. Another option involved the implementation of a spatial localizer tracked using infrared light-emitting diodes and a position sensor [4],[5]. Subdural grids have also been used to accomplish the task of tracking surface points and could conceivably be developed for use in tracking the bipolar stimulator [6].

A variety of obstacles were realized upon first consultation with Dr. Galloway and Mr. Hartmann. Integration of the tracking system into the ORION® package was a necessity due to its current ability to process preoperative images. Also, no modifications could be made directly to the bipolar stimulator that was being used. The stimulator handle and lead-tip had been provided to the TGT group by the Vanderbilt University Medical Center on a loan basis. Although it was not the stimulator most commonly used during operations in the medical center, it was necessary that it remain intact and fully functional without any modifications in case it was needed for emergency use. This fact led to another realization concerning the needs and wishes of this project. It would be nice, although not totally necessary, if the methodology used for tracking the bipolar stimulator would be easily applicable to other makes and models of bipolar stimulator handles and lead-tips.

Initial examinations of the structure and characteristics of the bipolar stimulator brought another realization concerning the actual implementation of this project. The flexible nature of the bipolar leads, although often necessary when utilizing the bipolar stimulator for other tasks, is very disruptive to accurate localization when employing both the articulated arm and spatial localizer methodologies. In both of these cases, the relationship between the tip of the stimulator and a known point must remain constant throughout the measurement period and any corruption of that relationship would result in destruction of the measured data.

These considerations were taken into account and the project was initially divided into two tasks: developing a tracking methodology and developing a technique to reduce the flexion of the stimulator leads. Possible solutions for both problems were analyzed giving consideration to the guidelines set out in the initial consultation with Dr. Galloway and Mr. Hartmann. Cost-efficiency and safety were also used as analysis criteria. Once the decisions were made concerning the best possible implementations, additional consultation and testing were used to determine the quality and validity of the implementation. The result of this effort is a working prototype that has been proven to provide accurate three-dimensional tracking of the bipolar stimulator and has done so in accordance with all guidelines determined for this project.

Methodology

I. Bipolar Stimulator Tracking

This issue was the central focus of the entire project. Without its successful completion, further work would be futile. The necessity of developing an accurate method for tracking the bipolar stimulator forced a careful examination of several solution possibilities. After a careful literature search, as described above, three possible methodologies were selected and considered for implementation.

The first solution possibility was the use of subdural grids (A-1, Fig. 2). This solution proved exciting as it completely eliminated the error associated with the flexion of the stimulator leads by detecting only the location of the tips of the stimulator as opposed to calculating the location of the tips with respect to a rigid body. It also proved a promising solution as it could likely be easily integrated into the ORION® software package. However, the actual implementation of the grid was not feasible for a senior design project due to the necessity of operating room (OR) testing and the regulatory approvals that would have had to come with attempting to test an undergraduate research project on a subject undergoing open-cranial surgery.

The second solution was the combination of a stereotactic frame and an articulated arm (A-1, Fig. 3). Utilizing a fixed-position frame, it would be possible to attach the stimulator to the Mark II arm previously developed by the TGT group [7]. However, this would have involved tedious recalculation of the location of the new endpoint of the arm with respect to the frame base and would not allow for the utilization of the ORION® package as it is currently incompatible with the articulated arm.

The approach that was eventually pursued involves the integration of a previously developed spatial localizer. This type of implementation has already been shown effective for use with a simple probe tip [4],[5] and an ultrasound hand unit [8]. One advantage of this implementation is that it is completely compatible with the ORION® package. Similarly, the recalculations that would have been associated with the implementation of the articulated arm solution are performed nearly instantaneously by a rigid-body program available from the TGT group, which also serves to make this approach far more feasible. Unfortunately, this technique cannot compensate for the stimulator lead flexion. It was decided that the flexion problem could be addressed by other, simpler means.

Once this method was chosen, it was necessary to develop a new localizer probe that would eventually be attached to the bipolar stimulator. An evaluation of the spatial localizer used in tracking the ultrasound hand unit found it highly specialized for its use and relatively inapplicable due to size relative to the stimulator [8]. However, examination of the probes discussed in [4] and [5] found them easily adaptable for use in this particular application.

The localizer is composed of a cylindrical probe containing 24 infrared light-emitting diodes (IREDs) positioned around the handle (A-1, Fig. 4). These IREDs are then tracked by an Optotrack 3220, a position sensor that, due to the known geometrical arrangement of the IREDs, is capable of determining the location of the tip of the stimulator if three or more of the IREDs can be localized (A-2, Fig. 5) [4]. The position sensor then feeds the tracking information back to a computer workstation operating the ORION® software package that calculates and tracks the location of the tip of the bipolar stimulator in three-dimensional space every 300 microseconds.

Utilization of this previously developed localizer probe presented another problem. It provided no method by which to attach the probe to the bipolar stimulator. This problem would have to be addressed later in order to make this a functional implementation. However, the spatial localizer methodology presented no other immediate problems.

A designsafe analysis of the hazards associated with the implementation of the localizer probe included stabbing/puncture by the probe and mechanical failure of the materials. There is also some concern about the handling of the probe, possible risk of electrocution, and the weight associated with the probe. The probe is constructed primarily of Delran, a nonconductive plastic that serves to insulate the operator from the electronics underneath, and is extremely light. The IREDs exposed to the surface of the probe are also nonconductive to the touch and thereby safe. Thus, the primary concerns about the safety associated with the design of the project are satiated.

The time spent to implement the probe handle was about three hours. This included time spent rummaging through the TGT laboratory in search of ideas for possible implementation, but was primarily spent testing the actual probe handle once it was found. The first handle was found to have several nonfunctioning IREDs. However, all 24 IREDs on the second probe found functioned perfectly. Correspondingly, there were no purchases made for this particular aspect of the project since the entire probe handle was found previously assembled in the TGT laboratory.

II. Bipolar Stimulator Handle Carriage

The need for a manner by which to connect the localizer probe to the bipolar stimulator was the next issue to be addressed. Without a feasible means to connect the stimulator to the localizer, the preferred method for tracking the stimulator would be rendered useless. Fortunately, a temporary solution to this problem was found almost immediately.

Two limiting factors determined the design of the connecting device (carriage): it could not permanently alter the bipolar stimulator and it must be light enough to be easily manipulated in the OR. It was quickly conceived that a type of clamp be used in order to hold the stimulator firmly without altering it, recognizing that the material of which the clamp is made could be changed in order to make it light and easy to use. While consulting with Mr. Hartmann concerning the dilemma, he quickly remembered that such a clamp had been used for a previous project and was likely still available for use. A short time later, it was found.

The clamp is composed of two pieces (A-2, Fig. 6). The primary piece screws into the localizer probe handle and serves as the base of the clamp in which the bipolar stimulator handle rests. The secondary piece rests atop the bipolar stimulator handle and clamps to the primary piece via two screws. The design of this carriage was small enough and light enough, due to its aluminum composition, to remain easy to handle and did not alter the stimulator handle in any way, thereby meeting the two specifications placed on this aspect of the project. There was not a designsafe analysis performed on this aspect of the project due to the fact that it was not a planned problem, but was encountered during the course of the project. However, no apparent hazards exist in association with this aspect of the project.

The time spent implementing this aspect of the project was minimal. It only took Mr. Hartmann and I about twenty minutes to find the carriage once we had discusses what it was that I thought we would need. Similar to the localizer probe, there were no costs incurred with the implementation of the off-set carriage since it was found in the TGT laboratory.

III. Flexion Reduction

The goal for the flexion reduction portion of the project was to minimize the ability of the bipolar stimulator leads to flex out of shape while being used. There were several possible solutions that could be implemented in order to solve this problem. Initial brainstorming brought forth ideas like machining new leads for the current stimulator handle or attempting to find stiffer leads that were already manufactured for use with the stimulator handle that we were employing. Concerns with approval for use of new stimulator leads developed by an undergraduate senior deterred me from the first possibility and web-research of the availability of stiffened leads suggested that the second possibility was not likely.

My research then led to me to a solution to a similar problem encountered by a summer intern for the TGT group. Based on a similar design for the reinforcement of a flexible radiofrequency ablator, I decided to design a sheath that would strengthen the leads and prevent them from flexing out of shape.

Several specifications would have to be met in order for such an implementation to work. The sheath would have to light and easily manipulated and must allow the operator to still see the tips of the stimulator. It could not permanently alter the stimulator leads in any way. The material would also have to be nonconductive due to its proximity to the conductive tips of the stimulator leads. A designsafe analysis of this portion of the project revealed hazards associated with the dropping of the sheath away from the stimulator, thereby dictating that the sheath would have to be designed to fit well enough to the stimulator tip to prevent it from slipping off the end of the stimulator. The analysis also suggested a contamination hazard due to the proximity of this part to the patient and possible contact. The material from which the sheath was to be made would have to be able to withstand the sterilization processes that operating room equipment must undergo.

After consulting with Mr. Hartmann concerning the material properties necessary for the sheath, he suggested that Delran, the material used in the localization probe, be used for the sheath as well. According to Mr. Hartmann, the plastic can withstand the cold-sterilization processes used at Vanderbilt Medical Center and can therefore be used in the OR. Also, as discussed before, it is nonconductive and light. It can be easily shaped in a machine shop to the desired dimensions and is readily available, making it ideal for this application.

After taking the necessary measurements, a preliminary design was developed which fit snugly to the top of the stimulator leads and tapered gently as it continued down the leads so as to compensate for the line of vision of the operator. This plan was then taken to the Jacobs Hall machine shop where Phil Davis was consulted for further ideas. No modifications to the design were made at that time and Phil undertook the job of machining the piece.

Once the machining process was complete, it became clear that the design for the sheath met each one of the specifications. Of particular importance was the closeness of fit between the sheath and the top of the stimulator lead. The fit was ideal and actually proved somewhat difficult to remove, suggesting that it would be unlikely for the sheath, once properly in place, to become dislodged and fall into the patient. The remaining specifications were met by the tapering design and the use of Delran in making the sheath (A-2, Fig. 7).

The design and implementation of the sheath was the most time-consuming portion of the project. The time spent research and developing the sheath totaled about 12 hours. It is not known exactly how much time was spent by Mr. Davis on the machining of the sheath. Periodic updates from the machinist suggested that the final prototype was actually his third attempt because the previous two attempts had cracked during the machining process. The sheath remained in the machine shop for nearly two months. Testing and evaluation, once the machining process was complete, totaled approximately six hours.

Similar to the rest of the project, there was no cost incurred to develop the stimulator lead sheath. The Delran core given to Mr. Davis for the machining process was excess found in the TGT laboratory, and Mr. Davis generously donated his spare time to work on the machining of the sheath, as the project would have likely proven too difficult and detailed to be adequately completed by an inexperienced machinist.

Results

Following the completion of the machining of the sheath, the entire prototype was ready to undergo testing. The same research that suggested implementing a sheath to reduce the stimulator lead flexion also suggested a methodology for evaluating the accuracy of the bipolar stimulator tracking system [9].

The testing process began with the use of a program to create a rigid-body file. This data file contains the matrix information that relates the locations of the 24 IREDs to the location of the tip of the bipolar stimulator in free space. This information is then transformed using a previously-constructed rig file so that the relationship between the IREDs and the stimulator tip are known in sensor space. This transformed data file is then passed to ORION®, which is used to gather the locations of seven points on the surface of an accurately machined (1/10,000th of an inch precision) cube (A-2, Fig. 8).

A final program called Target-Registration Error/Fiducial-Registration Error (TREFRE) is then used to determine the error involved in the measurement of the points. Four of the seven points are assigned as fiducials while the remaining three are assigned as targets. The program then calculates the least squares fit of the distance between the measured and calculated locations of the two groups of points. Because the “targets” are measured along a face perpendicular to the face containing the four fiducial points, the TRE is considered the true test of error.

As revealed in Fig. 9 and Table 1 in appendix A-3, both the maximum and average registration error were significantly reduced for both sets of points. This reduction would indicate a strong increase in accuracy in association with the implementation of the sheath. Evaluating the prototype as a whole, however, indicates the need for additional work in order to improve the overall accuracy of the system. Despite the 268% decrease in error as a result of the implementation of the sheath, a maximum error of less than 1.00 millimeters is generally desired. The maximum target registration error of the sheathed stimulator was approximately 1.62 millimeters, indicating that significant work remains to decrease this error to the desired threshold.

Conclusions

Many lofty goals were established at the onset of this project. The designing of a three-dimensional tracking system for a bipolar stimulator began as a one-facet project. Quickly, however, it incorporated many more fundamental engineering design facets as issues arose concerning the connection of the stimulator to the localizer probe without permanent alterations to the stimulator. Similar issues surfaced as the lack of rigidity in the stimulator leads began to have a negative effect on the accuracy of the tracking system as a whole.

Despite the occurrence of unexpected problems and new dimensions to seemingly one-dimensional issues, every goal and specification initially detailed and encountered along the course of the designing of the project was met. A method for tracking the bipolar stimulator was developed and it is fully compatible with the existing ORION® software package. The tracking system can be connected to the bipolar stimulator without altering the stimulator in any way. The accuracy-destroying flexion of the stimulator leads was significantly reduced through the implementation of a sheath that can be sanitized safely using existing methods at the Vanderbilt Medical Center and has been machined well enough to prevent its accidental dislodging during point localization. Finally, the entire system has been implemented in a package that is still easily manipulated and handled by the operator in an OR setting without affecting the operator’s line of sight. Similarly, all designsafe hazards and risks were adequately compensated for and eliminated where possible.

In short, it can be concluded that the tracking of the bipolar stimulator was designed, implemented, and tested successfully.

Recommendations

A great deal of work remains to be done to totally complete this project. Some concerns exist about the off-set nature of the carriage and its consequence on the accuracy of the overall tracking system. A design for a new, linear carriage was developed and dropped off to Mr. Davis in the machine shop. However, Mr. Davis was unable to complete the carriage prior to the presentation of the tracking system and this report. Upon completion of the new carriage, the same TREFRE analysis as described above needs to be implemented in order to determine the accuracy gained, if any, as a result of the new carriage.

Also, the two-lead design of bipolar stimulators makes it difficult to accurately obtain a rigid-body file, a process that requires rotation of the entire tracking system about a single, fixed point. Development of an apparatus that would accept two leads on one end and allow the entire apparatus to rotate about a 3mm ball on the other end would likely increase the overall system accuracy by allowing for the creation of a more accurate rigid-body file.

It is likely that the new carriage and rotation design will decrease the error below the one millimeter threshold. However, it was also desired that the system be universal so as to accommodate for a variety of different stimulator handles. Unfortunately, the more universal the carriage design, the less accurate it is likely to be. That wish, as of now, will have to be postponed for the sake of the accuracy of the system. However, all of the needs placed on the design of the system have either been achieved or are easily within reach.

Bibliography

[1] “Nerve Conduction Studies”, Date Accessed: April 23, 2001, .

[2] “Preoperative Images”, Date Accessed: November 20, 2000,

.

[3] T. M. Peters, B. Davey, P. Munger, R. M. Comeau, A. C. Evans, and A. Olivier, “Three-dimensional multimodal image-guidance for neurosurgery,” IEEE Trans. Med. Imag., vol. 15, pp. 121-128, Apr. 1996.

[4] S. L. Hartmann, R. L. Galloway, “Depth-Buffer Targeting for Spatially Accurate 3-D Visualization of Medical Images”, IEEE Trans. Med. Imag., vol. 19, pp. 1024-1031, Oct. 2000.

[5] R. L. Galloway, R. J. Maciunas. W. A. Bass, and W. J. Carpini, ”Optical localization for interactive, image-guided neurosurgery,” SPIE Med. Imag., vol. 2164, pp. 137-145, 1994.

[6] “Intraoperative Findings”, Date Accessed: November 20, 2000,

.

[7] “Intrasurgical Guidance, Image Display and Data Gathering”, Date Accessed: November 20, 2000, .

[8] D. M. Muratore, B. M. Dawant, R. L. Galloway, “Vertebral Surface Extraction from Ultrasound Images for Image-Guided Surgery,” Medical Imaging, vol. 3661, pp. 1499-1510, 1999.

[9] “Error Reduction for Radiofrequency Ablation Probe Calibration and the ORION Software for Guidance Using a Preplanned Trajectory”, Date Accessed: November 20, 2000, .

Appendix

A-1

[pic] [pic]

[pic] [pic]

A-2

[pic] [pic] [pic]

[pic]

A-3

[pic]

-----------------------

[pic]

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download