Anthropomorphic Liver Phantom for CT and Ultrasound



Anthropomorphic Liver Phantom for CT and Ultrasound

Katelyn Herbert

Advisor: Dr. Robert Galloway (BME)

Department of Biomedical Engineering

Vanderbilt University

BME 273, Senior Design

Abstract

The purpose of this project is to create an anthropomorphic liver phantom in which liver tumors can be imaged using both CT (Computed Tomography) and Ultrasound to investigate whether imaging liver tumors with CT and Ultrasound is comparable. We tested the different imaging modalities by locating the tumors in the images, and then finding the actual physical space of the tumor using proximity detectors embedded in the tumors in the liver. We found that the tumors were observable in CT and Ultrasound images, and the proximity detectors worked in a range of 32 mm or less.

Introduction

Liver cancer is a very serious and prevalent health problem in the world today. It is the fifth most common cancer in the world, and has the third highest occurrence of cancer related death (Llovet et al, pg. 1907). Worldwide, over 500,000 people are diagnosed with liver cancer per year (Llovet et al, pg. 1907). A very invasive and difficult type of liver tumor to treat is the Hepatocellular Carcinomas (HCCs). These types of tumors are the result of chronic liver disease most often caused by Hepatitis C and liver disease caused by alcoholism (M. Sherman, pg. 1208). The HCC tumors are often very small (~3 cm) and very difficult to find using either CT or Ultrasound. Fortunately, with the more recent advent of RF ablation, 80-90% of HCC tumors can be completely eradicated using RF Ablation (Meloni, pg. 375).

RF Ablation

RF ablation is a method used to eliminate tumors. This method is especially effective for liver tumors given their placement and lack of response to other forms of treatment (Lencioni, pg. 38). When a patient undergoes RF ablation, a needle-like ablation tool is placed inside the tumor. A high frequency RF signal is sent down the ablation tool, causing an alternating electric field in the tissue of the patient (Lencioni, pg. 41). The disturbance of the ions in the tissue caused by the alternating electric field generates frictional heat (Lencioni, pg. 41). The tissue needs to be heated from 50 degrees Celsius to 100 degrees Celsius for 10 to 30 minutes in order to insure tumor cell death (Lencioni, pg. 41).

Before surgery, the tumors are located using either CT or Ultrasound; during surgery, the RF ablation tool is guided using Ultrasound imaging. However, CT and Ultrasound are relatively insensitive to locating and diagnosing these types of tumors (Cioni, pg. 163). Using Ultrasound, the tumors often appear spherical with sharp and smooth boundaries, but US only has a sensitivity rate ranging from 20-70% (Cioni, pg. 163). CT sensitivity is only 52-79% (Cioni, pg. 165) with a substantial limitation to the being able to locate the periphery of these types of tumors.

As state previously, the ablation tool is guided using these imaging techniques, so it is a significant disadvantage that the imaging used to guide the tool is at times inadequate to correctly differentiate between healthy tissue and cancerous tissue. Inaccurate placement of an ablation tool while attempting to destroy cancerous tissue could result in the recurrence of tumors or destruction of healthy liver tissue. Therefore the objective of this project was to investigate whether CT and Ultrasound imaging guidance could be comparable when attempting to locate liver tumors, and determine the accuracy with which an ablation tool is placed. To do this, we created an anthropomorphic liver phantom to be imaged using both CT and Ultrasound in order to locate "tumors" embedded in the "liver tissue".

Methodology

For this project, we chose to create an anthropomorphic liver phantom. The phantom was constructed out of a 1:1 ratio of Ecoflex® Silicone mixture. A mold was of an actual human liver was used to create the phantom. Actual tumors are very asymmetric, but are roughly spherical in shape. Therefore we made the tumors using a disposable contact lens case to mold the tumors, creating roughly spherical tumors about 3 cm in diameter. The "liver tumors" needed to contrast with the "liver tissue" for them to be imageable using CT and Ultrasound and to mimic the properties of actual tumors. In order to insure the tumors were visible in a CT image, a small amount of barium powder was added to the same 1:1 silicone mixture that was used for the "liver tissue". The barium molecules are large enough to allow the x-rays created by the CT scanner to bounce off them, making them highly visible in the image. In order to make the tumors visible to Ultrasound, we created air bubbles in the silicone mixture by frothing the mixture when it was stirred. Ultrasound uses sound waves that bounce back from interfaces between materials with different material properties. Sound waves do not propagate through air very well, so by creating air bubbles in the mixture, we insured that the sound waves would bounce back from the liver tissue/tumor interface and create a clear image.

Proximity Detectors

In order to test the accuracy of the placement of the ablation tool, we developed two distinct methods. The first method was to embed an antenna in the tumor that consisted of a coil of wire. A 3MHz signal from the function generator would be sent down the ablation tool. The high frequency signal would create a current in the coil of wire. The signal created by the coil of wire would be amplified using a differential amplifier and filtered with a bandpass filter for only the 3MHz signal. A multimeter would measure the output voltage with an LED that would light up to inform us when then ablation tool was within a few millimeters of the tumor. The higher the output voltage, the closer the ablation tool was to the tumor.

The second method to test the accuracy of the placement of the ablation tool was to embed a 4.7 mm reed switch (GR200) from Standex Electronics in the tumor which was then embedded in the liver. Reed switches composed of two very tiny, thin strips of metal that bend and contact each other, closing the circuit when a magnetic field is present. Using this idea, we placed a very small magnet from K&J Magnetics (D11SH) in the tip of an ablation tool. As the ablation tool nears the reed switch (the tumor) the metal pieces in the switch contact each other, close the circuit, and turn on a LED that is attached to the circuit.

Imaging

The phantom was first imaged using CT and Ultrasound in order to locate the tumor using both modalities. The CT image was created using a 120 kV signal at 79 mAs, and a slice thickness of 2mm. The phantom was then imaged using Ultrasound.

Results and Discussion

The phantom was first imaged using CT. The CT image (Figure 2) revealed that the tumor was in fact visible. The silicone of the liver tissue appeared gray while the silicone and barium of the tumor tissue appeared bright white in the CT image. The phantom was successful in showing a significant difference in the two different types of tissue. The clear difference may be in fact unrealistic when compared to actual tumor tissue and liver tissue. One of the main challenges of using CT imaging to locate tumors is finding the barrier where the diseased tissue stops and healthy tissue starts. In our phantom, there is a very clear barrier. For the purpose of this model, this is not a vital issue, but for models it may be something that might be considered.

The phantom was next imaged using Ultrasound. The Ultrasound image was able to show the tumor in the liver phantom as indicated by the red circle (Figure 3). The image was somewhat unclear, but Ultrasound images are very often unclear due to tissue

heterogeneity and anisotropy. Therefore this image was very realistic, but still was able to achieve the goals for this project. If needed, a further improvement to the phantom would be to not only create air bubbles within the tumor mixture, but to pay more attention to the removal of bubbles within the liver tissue mixture. This would create more of a contrast in the US images. However, there are limitations due to the setting time of the silicone mixture used (10 minutes), and this improvement may not be feasible.

The first method of the proximity detection of the ablation tool was using the antenna embedded in the liver and measure the signal created in it to find the distance the ablation tool was from the tumor. Using the circuit in Figure 1, we were able to create a signal that increased as the ablation tool got closer to the tumor (Figure 4). However, it was only on the order of mV. In order for us to have used an LED and observe a significant signal change in the output, the output voltage to be on the order of a few volts. Using this knowledge, it was realized that the antenna would have to be tuned and further modifications would have to be made. Given that an electrical engineer was not a part of the team, extensive research, outside input, and time beyond the scope of this project would be required to perform these adjustments. Further development of this method would be required to determine whether this would be an acceptable solution to the problem at hand.

The second attempt involved embedding the reed switch in the tumor with the tumor then embedded in the liver phantom. Before the tumor was placed in the liver, it was tested at what distance the magnet needed to be when the two pieces of metal in the reed switch came together to close the circuit and turn on the LED. It was found that the reed switch closed when the magnet was 10mm-32mm away from the switch (Figure 5). This was significantly dependent on the angle of approach of the magnet (ablation tool). The magnet had to be 10mm away in order for it to close the switch when the tumor was approached perpendicular (90 degrees) from the front of the reed switch, but the ablation tool only needed to be about 30mm away when approached from 45 degrees or about 32mm away when approached from 90 degrees from the top of the reed switch.

Our method of proximity detection was therefore successful, but instead of giving us a signal reading that could tell us the exact distance the ablation tool was from the tumor based on signal strength, it instead could give only a binary result. The LED is either on and therefore in the range of 10mm-32mm, or it was off, meaning it was out of that range. Therefore, the sensitivity of this method is hindered by the fact that it can give us only a range rather than an actual value. The HCC tumors are about 3cm in diameter, this method may not be sensitive enough as it stands to determine the accuracy of the placement of the ablation tool. Further adjustments such as using a magnet that would still be able to make the circuit close but at a smaller distance might be done to allow the test to be more sensitive.

Safety Issues

Luckily, this project has very few real dangers associated with it. One danger is the use of the ablation tool. These are very sharp and may require some additional care to insure no injury when using it to test the proximity of its placement. As with all electrical equipment, care must be taken to insure when handling the phantom and the power source used (3V battery). Additional insulation on the wires may be applied to insure no injury will occur due to electric shock. When the phantom is imaged, precautions must be taken to avoid radiation from the CT machine. This can be prevented by using appropriate safety equipment or clothing.

Economic Estimates

The materials for this project are relatively inexpensive. The phantom may be used multiple times, but should be punctured by the ablation tool as little as possible. Therefore after a few uses, it will need to be replaced. The silicone mixture is roughly $180.00 for a gallon of the mixture, but uses only about 40oz per phantom. The magnets were roughly a dollar for 10 magnets, while the reed switches were given to Vanderbilt for this purpose by the Standex company. Therefore, these are relatively inexpensive to make.

Conclusions

We were able to create a phantom that had tumors that were visible using both CT imaging and Ultrasound imaging, and were able to accurately locate them in the physical space of the phantom. We embedded proximity detectors in the tumors to give a range of distances the ablation tool (magnet) was away from the tumor by using the reed switches that close the circuit when a magnetic field is present. However, this is a binary result (the reed switch closes when the magnet is at least 32 mm away). Therefore we are not able to get the exact distance of the ablation tool to the tumor and further work would need to be done to achieve that result.

Recommendations

This project will not need to be tested with humans or animals, so there are no societal issues or ethical issues that could be raised by overstepping boundaries by creating this phantom. This project is intended purely for research only, and it will not reach a clinical setting. Therefore, it is not foreseen that FDA approval will be required after further development.

This project could be further developed by attempting the first method of proximity detection. With more extensive electrical knowledge, it would be possible to tune the antenna, and make it possible to have a signal output that could give the actual distance of the ablation tool to the tumor. This would provide more information and better results to more clearly identify whether CT imaging and ultrasound imaging are comparable in locating liver tumors.

References

1. Lencioni et al. Image-guided thermal ablation of hepatocellular carcinoma. Crit Rev. Oncol. Hematol. 2008 June; 66(3) 200-7

2. J.M. Llovet, A. Burroughs and J. Bruix, Hepatocellular carcinoma, Lancet 362 (2003), pp. 1907–1917

3.Cioni and C. Della Pina et al., Imaging diagnosis, Semin Liver Dis 25 (2005), pp. 162–170

4. Meloni et al, “Hepatocellular Carcinoma Treated with Radiofrequency Ablation.” American Journal of Roentgenology 2001 vol. 177 pg. 375

5. J. Bruix and M. Sherman, Management of hepatocellular carcinoma,  Hepatology  42  (2005), pp. 1208–1236

Appendix

IWB

Ideation Process

Innovation Situation Questionnaire

1. Brief description of the problem

It is reported that nearly 20,000 individuals in the United States are diagnosed with liver cancer every year. The common treatment for patients is the removal of the tumors using an ablation tool with tomographic guidance. The ablation tool is an electrode that emits radiofrequency, or a high frequency alternating current, that when placed in the center of the tumor, generates enough heat to destroy the cancerous tissue. The current best practice is to guide the tool using ultrasound; however, 44 percent of all liver tumors are indistinguishable from healthy tissue using this method. A further drawback of this method is that finding the center of the three-dimensional object (the tumor) in a two-dimensional view is difficult.

2. Information about the system

2.1 System name

* The following needs to be considered:

-the anthropomorphic liver phantom

-"tumors"

-the ablation tool

-method of testing the accuracy of placement of ablation tool

Anthropomorphic phantom- The phantom needs to be made out of silicone but needs to be free of air bubbles so that it may be used with ultrasound. It needs to have a different attenuation than the tumors so that ultrasound will be able to show the tumors in the image.

tumors- The tumors also need to be made of silicone, but should have something else mixed in to make them have a different attenuation than the surrounding silicone. They need be of appropriate size for this experiment. They will be roughly spherical in shape to imitate the actual size/shape of tumors. It will have a reed switch embedded in order to be the means of detecting the proximity of the ablation tool/magnet

ablation tool- needs to have a magnet at the tip in order to work with the reed switch to detect the proximity of it to the tumor

2.2 System structure

The tumor embedded in the phantom with the reed switch will be attached to a power source on one end and an LED at the other. As the magnet (ablation tool) is brought closer to the target (tumor) the reed switch will close, turning on the LED.

2.3 Functioning of the system

Primary useful function:

Detect the proximity of the ablation tool with respect to the tumors

Reason to perform useful function:

Question whether CT and Ultrasound guidance can be compared and are accurate when performing RF ablation

Functioning of the system:

The silicone and barium tumors have the proximity detectors (reed switches) placed inside them. They are embedded inside the liver phantom, attached to a power source and a LED. The reed switches close the circuit when a magnetic field comes in contact with them. A magnet on the end of the ablation tool will be brought near the tumors, and when the magnet is within 10-30mm to the tumor, the switch will close, completing the circuit and allowing the LED to turn on.

2.4 System environment

The magnetic field will need to be able to penetrate the silicone and still be strong enough to make the reed switch close.

3. Information about the problem situation

3.1 Problem that should be resolved

The problem to be resolved is to improve the accuracy of the placement of the ablation tool using tomographic guidance and an anthropomorphic liver phantom.

3.2 Mechanism causing the problem

It is difficult to tell the difference between healthy tissue and tumors in real patients. It is also difficult to determine the center of the 3-dimensional tumor in the 2-dimensional view.

3.3 Undesired consequences of unresolved problem

The tumors could not be completely destroyed, allowing them to return and then create the need for additional surgeries. It may also damage healthy tissue in the process.

3.4 History of the problem

The problem is as old as the development of the ablation tool.

Previous attempts have been made to create the anthropomorphic liver phantom out of silicone and the tumors out of styrofoam, but when the system was tested, surgeons were able to feel the styrofoam tumors because they were so much harder than the surrounding silicone. It is not known whether the location of the tumors in the phantom has been attempted in the way proposed.

3.5 Other systems in which a similar problem exists

The problem could exist with all surgical tools. It is always difficult to insure the proper placement of surgical tools.

3.6 Other problems to be solved

1. Creating a phantom with tumors that will be visible in ultrasound and CT

2. method of detecting proximity

3. creating a working demonstration

4. Ideal vision of solution

Be able to create a working demonstration with the phantom that will show as the magnet gets close enough to the switch, the LED will turn on.

5. Available resources

Substance Resources:

Silicone

molds to make phantom

molds to make tumors

barium to mix with silicone to make tumors

reed switches (smallest available to keep the correct scale)

magnets(also small enough to fit in the tip of the ablation tool, but still have a large enough magnetic field to close the switch)

wiring for circuit

LEDs

ultrasound machine

CT machine

Field Resources:

3V battery

magnetic field

Informational Resources:

Fields emitted from the system: electrical, magnetic

Functional resources:

Additional useful functions:

Way of detecting the tumors in the phantom faster

Harmful:

Ablation tool with magnet attached to it would no longer be able to be used during actual surgery

6. Allowable changes to the system

How the location of the tumor is determined may be modified.

Limitations:

• the phantom must be made out of silicone in order to best represent actual tissue.

• I won’t be able to actually glue the magnet to the ablation tool because of the cost, therefore some other type of tool may need to be used for the demonstration

• I have limited experience with electronics

• Use of the CT and ultrasound is limited due to time constraints

7. Criteria for selecting solution concepts

Desired Technological Characteristics:

-detect the presence of the magnet by the closing of the reed switch, allowing for an LED to turn on.

Desired Economic characteristics

Be able to only use those resources that are available to use in the lab at Vanderbilt

Desire Timetable

Spring Semester

Expected Degree of Novelty

Be able to create a demonstration of the phantom

8. Company business environment

Create only for the purposes of a senior design project

9. Project data

Project name:Anthropomorphic Phantom for CT and Ultrasound

Project Objectives: Create a Phantom that will be able to detect the proximity of a magnet(ablation tool)

Project Timeline:

10/29 NCIIA format proposal due electronically

11/3 web based progress report #1

11/4 make phantoms degassing one, one without degassing, develop LabView platform to generate signal and filter the output

11/5 web pages must be up by now with email links, and make ultrasound of phantoms 11/10 web progress report #2 or oral 1, start developing antenna

11/12 Corrected NCIIA proposal due by today on web, continue developing antenna

11/17 Web progress report # 2 or #3 or Oral 1, and test antenna

11/19 electronic reference links due - minimum of 3 relevant www links, and test antenna when encased in silicone

11/20 Thanksgiving Break

11/29 Return from Thanksgiving Break

12/1 Web progress report #3 or Oral 1 this week, perform more testing of antenna

12/8 IWB with conflict map & expansion due on web, make-up Oral

12/20 Winter Break

1/9-Return from Winter Break

1/19-Create a plot of the output to the signal source distance

2/2-Embed the mesh, tumor, antenna in anthropomorphic phantom

2/11-Test the whole system

2/16-Set up project for demonstration

3/2-Continue working on demonstration

3/9-Continue working on demonstration

3/18-Begin finishing up Project

3/25-Continue finishing up project, begin work on poster

4/6-Continue work on poster/demonstration and presentation

4/13-Continue work on poster/demonstration and presentation

Late April-Have finished project with poster complete, present project

Project Team:

Katelyn Herbert, BME

katelyn.j.herbert@vanderbilt.edu

Problem Formulation

1. Build the Diagram

[pic]

2. Directions for Innovation

12/8/2009 11:45:25 PM Diagram2

1. Find an alternative way to obtain [the] (Complete Removal of Tumor) that offers the following: provides or enhances [the] (Return to health), eliminates, reduces, or prevents [the] (Additional Surgeries), does not require [the] (Accuracy in determining location of tumor).

2. Find an alternative way to obtain [the] (Return to health) that does not require [the] (Complete Removal of Tumor).

3. Consider transitioning to the next generation of the system that will provide [the] (Return to health) in a more effective way and/or will be free of existing problems.

4. Find an alternative way to obtain [the] (Accuracy in determining location of tumor) that offers the following: provides or enhances [the] (Complete Removal of Tumor), (Quick Removal of Tumor) and (Prevention of removal of healthy tissue), does not require [the] (Coupling with mesh around tumor).

5. Find an alternative way to obtain [the] (Quick Removal of Tumor) that does not require [the] (Accuracy in determining location of tumor).

6. Consider transitioning to the next generation of the system that will provide [the] (Quick Removal of Tumor) in a more effective way and/or will be free of existing problems.

7. Find an alternative way to obtain [the] (Coupling with mesh around tumor) that offers the following: provides or enhances [the] (Accuracy in determining location of tumor), does not cause [the] (Additional heat), does not require [the] (Signal from antenna).

8. Try to resolve the following contradiction: The useful factor [the] (Coupling with mesh around tumor) should be in place in order to provide or enhance [the] (Accuracy in determining location of tumor), and should not exist in order to avoid [the] (Additional heat).

9. Find an alternative way to obtain [the] (Signal from antenna) that offers the following: provides or enhances [the] (Coupling with mesh around tumor), does not cause [the] (Interference with the radiofrequency signal from ablation tool).

10. Try to resolve the following contradiction: The useful factor [the] (Signal from antenna) should be in place in order to provide or enhance [the] (Coupling with mesh around tumor), and should not exist in order to avoid [the] (Interference with the radiofrequency signal from ablation tool).

11. Find a way to eliminate, reduce, or prevent [the] (Additional heat) under the conditions of [the] (Coupling with mesh around tumor).

12. Find a way to eliminate, reduce, or prevent [the] (Interference with the radiofrequency signal from ablation tool) under the conditions of [the] (Signal from antenna).

13. Find a way to eliminate, reduce, or prevent [the] (Additional Surgeries).

14. Find an alternative way to obtain [the] (Prevention of removal of healthy tissue) that does not require [the] (Accuracy in determining location of tumor).

15. Consider transitioning to the next generation of the system that will provide [the] (Prevention of removal of healthy tissue) in a more effective way and/or will be free of existing problems.

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

[pic]

Figure 1. Circuit diagram with differential amplifier and LED for first method.

[pic]

Figure 2. CT image of the liver phantom and tumor. The “Liver tissue” appears grey in the image, and the “tumor” appears white in the image.

[pic]

Figure 3. US Image of phantom. Red circle=location of tumor

[pic]

Figure 4. Signal to Distance graph with the 3 MHz Signal.

|Angle of Approach (degrees) |Distance From Reed Switch (mm) |

|90 from front |10 |

|90 from back |10 |

|45 from left |30.1 |

|45 from right |30.1 |

|90 from top |32.2 |

Figure 5. Tested at what the distance the magnet was when it caused the reed switch to close (caused LED to turn on). Approached the reed switch from several different angles.

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

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

Google Online Preview   Download