Transcript Impedance Imaging for Breast Cancer Diagnosis

Impedance Imaging for Breast Cancer Diagnosis
Tzu-Jen
1
Kao ,
Ning
3
Liu ,
1
Xia ,
1
Kim ,
2
Isaacson ,
3
Saulnier ,
Hongjun
Bong Seok
David
Gary J.
Jonathan C.
1Departments of Biomedical Engineering, 2Mathematical Sciences
3Department of Electrical, Computer and Systems Engineering Rensselaer Polytechnic Institute
Model of the mammogram geometry:
Introduction:
State of the art & dummy load test:
The long-term goal of this project is to develop Electrical Impedance
Tomography (EIT) technology to improve the diagnosis of breast cancer.
The spectral characteristics of excised breast tissue provided by Jossinet & Schmitt [6]
Xitron
Theoreticals
ACT4 measurement
300
500
Importance of this work:
The reconstruction of conductivity images in a mammogram geometry will
provide a foundation for the comparison of EIT as a breast cancer detection tool
with the X-ray mammography.
CoCo Plot with 600 ohm // ( 300 ohm + 0.01 uF)
Breast Cancer is responsible for over 40,000 deaths annually among women in
the United Stated. Early detection of breast cancer improves the chances that it
can be treated successfully. Also, improved detection techniques may be able to
reduce the number of biopsies that are performed. Even using combined
diagnostic techniques, such as mammography, MRI, clinical breast examination,
only 20-40 % of the biopsies that are performed actually reveal cancer [4].
1
Newell
Reactance
200
150
100
50
0
100
200
400
600
700
-50
Electrical Impedance Tomography, also called Electrical Impedance Imaging, is
a non-invasive technique used to image the electrical conductivity and
permittivity within a body from measurements taken on the body surface [5].
Small electric currents are passed through the body using electrodes attached to
the skin; the resulting voltages are then measured and used by mathematical
algorithms to reconstruct the values of the electrical conductivity and
permittivity within the body.
It has been known for some time that breast tumors have a significantly higher
conductivity and permittivity than surrounding normal tissue. Hence, by
forming images of the electrical conductivity of the breast, tumors may be
detected and differentiated from normal tissue. EIT does not use ionizing
radiation and employs relatively low-cost instrumentation, making it less
expensive and/or safer than x-ray mammography, MRI and CT.
Hardware and Software: [1, 2]
Resistance
The results suggest that EIT is a promising modality to image the breast for
malignancies.
Experimental Design: [3]
Reconstruction Algorithms:
Size of Electrode Array: 5 x 5 cm
Dim. of the box : 5 cm x 5 cm x 5 cm
Size of Electrode: 1.0 x 1.0 cm
Gap between Electrodes: 2 mm
Reconstructing an EIT image involves solving the inverse problem of
determining the conductivity and permittivity distribution within a body that
produces the observed relationship between the voltage and current at the
surface. The problem is ill-posed, meaning that large changes in the
conductivity and permittivity can produce only small changes at the surface.
This ill-posedness makes the reconstruction problem more difficult and results in
the need for very high-precision instrumentation.
Conductivity σ0 = 0.364 S/m
Agar Target size: 5 mm cube
Conductivity σ = 0.901 S/m
In the low frequency limit ( ω = 0 ) Maxwell’s equations reduce to
The ACT 4 hardware and software supports 64 electrodes that will produce 3-D
images of conductivity and permittivity in real time using applied excitations
with frequencies between 1 KHz and 1 MHz. It will be able to apply arbitrary
patterns of voltage or currents while measuring the resulting currents and
voltages, respectively. The operator will be able to select various combinations
of image rate and measurement precision, including real-time operation to
visualization cardiac-frequency events.
MBLT
Slave
FPGA
ESM(0)
DSP1
   E
Therefore, in this case
E  U
This work is supported in part by CenSSIS, the Center for Subsurface Sensing
and Imaging Systems, under the Engineering Research Centers Program of the
National Science Foundation (Award Number EEC-9986821).
This work is supported in part by NIBIB, the National Institute of Biomedical
Imaging and Bioengineering under Grant Number R01-EB000456-01.
Future Plans:
 E  0
  E  0
Where U denotes the electrical potential or voltage throughout the interior of the
body. Hence in the low frequency limit we have that the voltage inside the body,
B, satisfies
   .U  0
On the body’s surface, S, one has   U  j and V  U . Here υ denotes
the outward pointing unit normal vector to the body B on the surface S, and j
denotes the applied current density on the surface. V denotes the measured
voltage on the surface.
The 4x4 Electrode Array
Agar Target
The Agar Target with
Electrode Array
We are presently testing the ACT 4 hardware and implementing the initial version
of the system software and firmware. Over the next few weeks we expect to
begin collecting complete data sets and testing the instrument performance. Initial
testing will use laboratory test phantoms with conductivity similar to that of the
human body. Once the instrument is fully tested and its performance
characterized, it will be used to study breast cancer patients at Massachusetts
General Hospital. The study, involving a small number of patients who will be
undergoing biopsies, will establish the ability of EIT to detect cancer by directly
comparing EIT results with biopsy results.
DSP2
PCI
User
Interface
PCI
CON B
DSP3
PMC
DSP4
Reef
DSP
Reef
FPGA
CON A
Master
FPGA
MBLT
VME
Bus
MBLT
Slave
FPGA
ESM(8)
Reef PMC+ Board
Hammerhead
Board
MBLT
Slave
FPGA
CAL
To support a high-speed, high-precision, multi-channel instrument, we designed
a distributed digital system, including a computer, digital signal processors
(DSPs) and field-programmable gate arrays (FPGAs). The computer provides a
user interface to control the instrument and to display the results for analysis.
Four DSPs are working in parallel to implement the real-time reconstruction
algorithm. We use FPGAs mainly to implement the signal generation, voltmeter
and control for the ACT 4 analog circuits. FPGAs are also used to implement
the interface between multiple channels through a VME-64x backplane, and
between the computer and the instrument.
Reconstruction
Algorithm
Data
Collection
Hardware
Hardware
Control
The following steps are used to reconstruct an approximation to the conductivity
σ at low frequencies:
1. Introducing a guess for the conductivity, σ0.
2. Relating the potentials or fields on the surface S to the electrical tissue
properties and field inside the body B by the identity
S
Image
Display
Import
Export
Layer 2
B
Here the subscript 0 denotes fields due to the conductivity σ0 . The
superscripts denote the fields that result from different current densities.
3. Apply an electrode model [7] relating currents and voltages on electrodes
labeled with the subscript l = 1, …,L into the above exact relation. Use the
notation δσ = σ - σ0 and the approximation Uk = U0k + O(δσ) to obtain the
equations relating measured current voltages to moments of the unknown
conductivity
L
D(k , x)   Vl k ( 0 ) I lx  Vl x ( ) I lk   U 0k  U 0x dp  O( 2 )
Layer 3
Layer 4
5
cm
Layer 5
References
Publications Acknowledging NSF Support:
1. Ning Liu, Gary J. Saulnier, J.C. Newell, D. Isaacson and T-J Kao. “ACT4: A
High-Precision, Multi-frequency Electrical Impedance Tomography” Conference
on Biomedical Applications of Electrical Impedance Tomography, University
College London, June 22-24th, 2005.
2. Hongjun Xia, A. S. Ross E. Brevdo, T-J Kao, Ning Liu, B. S. Kim, J.C. Newell,
G. J. Saulnier and D. Isaacson. “The Application software of ACT4.” Conference
on Biomedical Applications of Electrical Impedance Tomography, University
College London, June 22-24th, 2005.
3. Choi, M.H., T-J Kao, D. Isaacson, G.J. Saulnier and J.C. Newell. “A Simplified
Model of a Mammography Geometry for Breast Cancer Imaging with Electrical
Impedance Tomography” Proc. IEEE-EMBS Conf. 26, In Press 2005.
Layer 6
Layer 7
B
4. Measure and compute the components of the “data” matrix D.
5. Choose a basis of functions  n ( p)nN1 for approximation δσ by
6. Display σ = δσ + σ0
Data
Management
Layer 1
x k
k x
x
k
U
j

U
j
ds

(



)

U


U
dp
0
0
 0

l 1
Patient Information
Reconstructed Images:
Layer 8
 3D Image in 2D sequence (Layer 1 ~ Layer 8, thickness: 6.25 mm each )
 Agar target is located 6 mm away from the electrode array
BMP
Files
System
Configuration
Others:
4. American Cancer Society, http://www.cancer.org/docroot/home
5. M. Cheney, D. Isaacson, Newell, S. Simske, and J. Goble, “NOSER: An
Algorithm for Solving the Inverse Conductivity Problem” Int.. J. Imag. Syst.
Tech., vol. 2, pp. 66-75,1990.
6. Jossinet, J. and M. Schmitt. “A review of Parameters for Bioelectrical
characterization of Breast Tissue” Ann. NY Acad. Sci. Vol. 873:30-41, 1999
7. Cheng, K-S., D. Isaacson, J.C. Newell and D.G. Gisser. “Electrode Models for
Electric Current Computed Tomography” IEEE Trans. Biomed. Eng. 36:918-924,
1989.
Contact Info:
Jonathan Newell, Ph. D.
Professor of Biomedical Engineering
E-mail: [email protected]
Rensselaer Polytechnic Institute
Web site: http://www.rpi.edu/~newelj/eit.html
Block Diagram of Software Structure
The ACT 4 system is hosted by a Pentium-based personal computer. Through
the computer, the operator is able to control the hardware, display and
manipulate images and manage the information database. The user interface
software is designed to be used in both clinical and laboratory environments and
is implemented as a window-style program coded in Visual C++. The interface
to allow access to low-level hardware functions will be available for
laboratory/experimental use. A simplified interface offering only the functions
needed for collecting patient data will be available for clinical use.
110 Eighth St. Troy, NY 12180-3590
Phone : 518-276-6433 FAX : 518-276-3035
Solid line: reconstructed voxel conductivity Dotted line: ideal computed conductivity of voxel