SAFETY ASSESSMENT OF BREAST CANCER DETECTION VIA ULTRAWIDEBAND MICROWAVE RADAR OPERATING IN PULSED-RADIATION MODE

Transcription

1 against the Refs. 7, 8, and 9, respectively. In addition, the proposed filter can be implemented by using a PCB technology to achieve low-cost, easily integrated design, and fabrication. Figure 5 Photograph of the proposed filter (29.75 ), s 0.1 mm and g 0.46 mm, as shown in Figure 3, and its simulated and measured responses are shown in Figure 4. The simulated and measured insertion losses (S 21 ) are 2.5 and 3.2 db, return losses (S 11 ) are 15 and 12 db at 2.45 GHz, respectively. The simulated and measured S 21 are 1.4 and 3.4 db, and S 11 are 20 and 10 db at 5.2 GHz, respectively. From these results, it can be seen that a little frequency shift between the simulation and measurement can be observed. The slight difference between the simulated and measured results might be because of fabrication error and this could be controlled by the improvement of fabrication technology. The proposed filter was built with Rogers RO3003 substrate with a relative dielectric constant of 3, a loss tangent of , and a thickness of 0.76 mm by using a standard PCB process. Figure 5 shows photograph of the proposed filter, and the total size of the filter is less than 14 mm 14 mm. Moreover, the proposed filter has smaller size as compared with the basic topology of stopband filters [3, 4] and SIR filters [5, 6] as listed in Table 1. The proposed filter also provides the size reduction of about 98, 77, and 15% against the Refs. 7 9, respectively. The proposed filter has attractive features, including dual-band applications, smaller size with respect to the other microstrip ring bandpass filters is very suitable for mobile cellphone and IEEE a/b/g. 4. CONCLUSION In this letter, a minimized closed-loop dual-band filters using trisection SIR structure is constructed and implemented. A trisection SIR is derived and designed to have identical fundamental and the first higher-order resonant frequencies. Moreover, the proposed filter also provides the size reduction of about 98, 77, and 15% TABLE 1 The Size of the Proposed Filter Compared with Those in Other References This work [7] [8] [9] r h (mm) Size f o (GHz) 2.4/ / / /5.2 Converter Area(mm 2 ) Converter 1 2 ref. x, mm relative x, mm ref. y, mm relative y, mm). REFERENCES 1. H. Miyake, S. Kitazawa, T. Ishizaki, T. Yamada, and Y. Nagatom, A miniaturized monolithic dual band filter using ceramic lamination technique for dual mode portable telephones, IEEE MTT-S Int Dig, Denver, CO, (1997), S. Wu and B. Razavi, A 900-MHz/1.8-GHz CMOS receiver for dual-band, IEEE J Solid-State Circuits 30 (1998), J.S. Hong and M.J. Lancaster, Microstrip filter for RF/microwave applications, Wiley, New York, 2001, Ch S.B. Cohn, Parallel coupled transmission-line resonator filters, IRE Trans Microwave Theory Tech MTT-6 (1958), E.G. Cristal and S. Frankel, Hairpin-line and hybrid hairpin-line/ halfwave parallel-coupled-line filters, IEEE Trans Microwave Theory Tech MTT-20 (1972), C. Quendo, E. Rius, and C. Person, An original topology of dual-band filter with transmission zeros, IEEE MTT-S Int Microwave Symp Dig WE1D-7 (2003), L.C. Tsai and C.W. Hsue, Dual-band bandpass filters using equallength coupled-serial-shunted lines and Z-transform technique, IEEE Trans Microwave Theory Tech 52 (2004), S.F. Chang, Y.H. Jeng, and J.L. Chen, Dual-band step-impedance bandpass filter for multimode wireless LANs, Electron Lett 40 (2004), S. Sun and L. Zhu, Compact dual-band microstrip bandpass filter without external feeds, IEEE Microwave Wireless Compon Lett 15 (2005), I. Wolff, Microstrip bandpass filter using degenerate modes of a microstrip ring resonator, Electron Lett 8 (1972), M. Makimoto and S. Yamashita, Bandpass filters using parallel coupled stripline stepped impedance resonators, IEEE Trans Microwave Theory Tech 28 (1980), J.S. Hong and M.J. Lancaster, Microstrip bandpass filter using degenerate modes of a novel meander loop resonator, IEEE Microwave Guided Wave Lett 5 (1995), L.H. Hsieh and K. Chang, Dual-mode quasi-elliptic-function bandpass filters using ring resonators with enhanced-coupling tuning stubs, IEEE Trans Microwave Theory Tech 50 (2002), Wiley Periodicals, Inc. SAFETY ASSESSMENT OF BREAST CANCER DETECTION VIA ULTRAWIDEBAND MICROWAVE RADAR OPERATING IN PULSED-RADIATION MODE Earl Zastrow, Shakti K. Davis, and Susan C. Hagness Department of Electrical and Computer Engineering University of Wisconsin-Madison Madison, WI Received 13 May 2006 ABSTRACT: We conduct an finite-difference time-domain computational electromagnetic analysis of the safety of ultrawideband microwave breast cancer detection techniques. Pulses are transmitted into anatomically realistic MRI-derived numerical breast phantoms with different levels of tissue heterogeneity as well as foreign entities such as implants and surgical clips, and the specific absorption is computed Wiley Periodicals, Inc. Microwave Opt Technol Lett 49: , 2007; Published online in Wiley InterScience ( com). DOI /mop DOI /mop MICROWAVE AND OPTICAL TECHNOLOGY LETTERS / Vol. 49, No. 1, January

2 Key words: ultrawideband microwave radar; breast cancer detection; specific absorption 1. INTRODUCTION Dosimetric studies of human exposure to ultrawideband (UWB) electromagnetic waves are of increasing importance because of the growing interest over the past decade in the use of UWB electromagnetic radiation in medicine. For example, UWB microwave radar has received much attention recently as a promising modality for early-stage breast cancer detection [1 5]. This approach images tumor signatures in backscatter received by an antenna array in response to the transmission of low-power UWB microwave signals into the breast. While it has been assumed that low-power UWB technology poses no health risk to the patient, it is important to formally verify that electromagnetic energy absorption in the breast falls within accepted safety guidelines prior to clinical implementation of UWB microwave breast cancer detection technology. In this article, we report the results of a finite-difference timedomain (FDTD) analysis [6] of electromagnetic energy absorption in the breast due to microwave radiation in the 1 11 GHz band. We assume that the UWB signals are generated physically as time-domain impulses rather than synthetically using a sweptfrequency input. (The latter scenario is considered in [7].) We compute the specific absorption (SA) in anatomically realistic 3D numerical breast phantoms containing different levels of tissue heterogeneity, as well as foreign entities such as surgical marking clips and saline breast implants. This safety assessment is broadly applicable to several preclinical embodiments of UWB microwave breast cancer detection technology. 2. MODELS AND METHODS We consider the microwave imaging scenario wherein a patient lies prone with an antenna array encircling the breast. We represent this scenario numerically by modeling an anatomically realistic breast immersed in a low-loss coupling medium. A single UWB pyramidal horn antenna [8] is placed near the breast. We position the antenna with the aperture 2.5 cm away from the breast, centered in the xy-plane as illustrated in Figure 1. The shape and the interior tissue structure of the anatomically realistic breast phantoms are derived from 3D MRI data sets. The MRI voxel intensity in the breast interior is mapped to a dielectric model via a piecewise linear map [9] for which we specify the median dielectric properties for fat and fibroglandular tissues and assume 10% variation about each median. Although fatty and fibroglandular tissues are considered to be distinct types of normal breast tissue with distinct dielectric characteristics, the bulk of the current literature on dielectric characterization of breast tissue at microwave frequencies does not make this distinction [10 12]. We expect fibroglandular tissue to exhibit higher dielectric properties than fat. For this safety assessment study, we consider the worst-case scenario by assuming that the dielectric properties data in the literature represents predominantly fatty breast tissue. In lieu of an established range of values for fibrogandular tissue, we consider two estimates of the dielectric properties of fibroglandular tissue. Type A is a moderate estimate (also used in Ref. 9), whereas type B is an extreme estimate with dielectric properties that are significantly higher than those reported to date in the literature. We describe all of the dispersive media properties in our models using a single-pole Debye expression for the complex permittivity 0 s j 0 s 1 j (1) Figure 1 3D FDTD model comprised of an MRI-derived numerical breast phantom (Phantom II) and an UWB antenna. (a) Full view showing the breast surface, (b) Cross-sectional view showing the breast interior Table 1 summarizes the Debye parameters for the dispersive dielectric media in the models. The Debye parameters for muscle (the 1.5-cm-thick chestwall) and skin (a 2-mm-thick layer) were chosen to mimic the properties reported in Ref. 13 over the frequency range of interest. The degree of heterogeneity of breast tissue varies from patient to patient. For completeness we investigate energy depositions in four numerical phantoms based on patients with the following breast composition: (i) almost entirely fat, (ii) scattered fibroglandular tissue, (iii) heterogeneously dense, and (iv) extremely dense. A cross-sectional view of Phantom II is shown in Figure 1(b). For each of the four phantoms, we consider both type A and type B dielectric properties to represent the fibroglandular tissue. In addition to the four numerical breast phantoms with varying tissue TABLE 1 Debye Parameters Used in the FDTD Models Media s s (S/m) (ps) Coupling medium [4] N/A Breast fat [3] Fibroglandular tissue-type A [9] Fibroglandular tissue-type B Skin[9] Muscular chestwall Physiological saline (0.15 M) 14] MICROWAVE AND OPTICAL TECHNOLOGY LETTERS / Vol. 49, No. 1, January 2007 DOI /mop

3 Figure 2 Unaveraged SA distribution of the coronal slice containing the global peak absorption (mj/kg) for Phantom I (type A) through IV (type A) compositions, we consider three scenarios where foreign materials are present in the breast phantoms. For the first scenario, we obtain an MRI of a patient with breast implants and assign the dielectric properties of physiological saline (0.15 M) [14] to the region of the implant. We use the type B properties to represent the fibroglandular tissue in this phantom. In the second and third scenarios, titanium surgical marking clips commonly deployed in excisional biopsy procedures are artificially inserted in Phantom III. We assign the type B properties to the fibroglandular tissue in the phantom for both of these scenarios. The two surgical clips we consider are a 2 2mm 2 horseshoe-shaped clip [15], and a 5 1mm 2 line-shaped clip. The clips are centered at x 5.0 cm, y 5.0 cm, and z 6.1 cm ( 0.5 cm beneath the skin layer and on the axis of the antenna). An FDTD simulation is conducted for each of the 11 breast phantoms described above. The excitation waveform applied at the feedpoint of the UWB horn antenna is a modulated Gaussian pulse with a temporal full width at half-maximum of 120 ps and a spectral peak at 6 GHz. The unaveraged SA during single-pulse exposure is computed at every voxel in the breast phantom as follows SA 1 TE t J t dt (2) Figure 3 Peak unaveraged SA in each coronal slice (corresponding to depth x) for each tissue type for Phantom I (type A) through IV (type A) peak unaveraged SA in a coronal slice at depth x for a specified tissue type. These figures show that the spatial distribution of unaveraged SA varies with the distance from the antenna as well as the tissue type. As expected, the absorption in the interior tissues (fat and fibroglandular) is less than the absorption in the skin. Variations in the absorption in the breast interior for different phantoms are relatively small. The peak unaveraged SA consistently occurs in the skin layer and has similar magnitude for all four phantoms. We further investigate the absorption in fibroglandular tissue as a function of its dielectric properties. Figure 4 shows the peak unaveraged SA value for the fibroglandular tissue in each coronal slice of Phantom III with type A and type B tissue properties. The energy absorbed in type B fibroglandular tissue is approximately twice the energy absorbed in type A, but remains significantly lower than the absorption in the skin and fat. Next, we evaluate absorption when foreign objects are present in the breast. In our first scenario, we consider the absorption in the breast phantom containing a saline implant as depicted by Figure 5. Figure 6(a) shows the unaveraged SA of the coronal slice containing the peak absorption for this phantom. The absorption level at the tissue-implant interface is elevated above other interior regions inside the breast. Figure 6(b) shows the peak unaveraged SA in each coronal slice for each tissue type and for saline. The saline implant alters the breast shape and distribution of the fibroglandular tissue, thereby modifying the interior scattering and using the six-field-component approach [16]. The tissue density, is assumed to be 1 g/cm 3. E and J are the electric field and total current density vectors, respectively, and T is the duration of the simulation. We also compute the peak 1-g SA for each phantom. All absorption values are normalized to 1 mj of radiated energy. 3. RESULTS Figure 2 shows the spatial distribution of unaveraged SA over the coronal slice containing the global peak absorption in mj/kg for Phantoms I through IV with type A fibroglandular tissue properties. The energy deposition as a function of tissue type within each phantom is shown in Figure 3, where each curve represents the Figure 4 Peak unaveraged SA in the fibroglandular tissue in each coronal slice (corresponding to depth x) of Phantom III for two types of fibroglandular tissue properties (type A and B) DOI /mop MICROWAVE AND OPTICAL TECHNOLOGY LETTERS / Vol. 49, No. 1, January

4 Figure 5 Cross-sectional view of the 3D FDTD model comprised of an MRI-derived numerical breast phantom (type B) containing a saline implant absorption properties. As a result, the SA levels in fibroglandular tissue (type B) in this phantom are higher than those observed in Figure 4 for Phantom III (type B). In our second scenario, we evaluate the situation where metallic surgical clips are introduced at a shallow location in the subcutaneous fat layer of Phantom III (type B). We considered several orientations of the two clips described in the previous section and found that the absorption is highest when the clips are oriented in the x-direction. We report the results for this orientation. Figure 7 shows that the presence of these clips raises the level of unaveraged SA in the vicinity of the clips. Clearly the absorption characteristics depend on the specific shape of the clip. The line clip induces two localized absorption peaks at the ends of the clip. The horseshoe clip induces a single localized absorption peak between the two prongs. The values of the peak absorption are comparable for both clips. These sharp increases in absorption are highly Figure 7 Peak unaveraged SA in each coronal slice for each tissue type of Phantom III (type B) with a (a) line-shaped clip and (b) horseshoeshaped clip centered at x 5.0 cm, y 5.0 cm, and z 6.1 cm localized and therefore have little impact on the 1-g SA around the clip region. The peak 1-g SA remains in the skin region. Table 2 summarizes the peak unaveraged and peak 1-g SA normalized to 1 mj of radiated energy from a single UWB pulse for all numerical phantoms. The variation in the shape, size and tissue composition of the phantoms results in only minor variation in the peak absorption values, and the foreign objects have no noticeable impact on the peak 1-g SA. The results of Table 2 along with the exposure limits established by IEEE standard C95.1 [17] make it possible to assess the safety of future clinical systems. The IEEE peak 1-g SA limit for a 6-min exposure is 576 J/kg [17]. This limit is 28,000 times higher than the peak 1-g SA from a single 1-mJ UWB pulse in Table 2. With these numbers, we can predict limits on the approximate amount of energy radiated in each pulse, the number of TABLE 2 SA in mj/kg for 1 mj of Radiated Energy Computed During Single-pulse Exposure Figure 6 (a) Unaveraged SA distribution of the coronal slice containing the global peak absorption (mj/kg) for the phantom (type B) containing the saline implant. (b) Peak unaveraged SA in each coronal slice (corresponding to depth x) for each tissue type and the implant EM Model Peak Unaveraged SA Peak 1-g SA Phantom I (type A) Phantom I (type B) Phantom II (type A) Phantom II (type B) Phantom III (type A) Phantom III (type B) Phantom IV (type A) Phantom IV (type B) Phantom (type B) with breast implant Phantom III (type B) with horseshoe-shaped clip Phantom III (type B) with line-shaped clip MICROWAVE AND OPTICAL TECHNOLOGY LETTERS / Vol. 49, No. 1, January 2007 DOI /mop

5 pulses radiated per antenna, and the number of antennas in an array. For example, if each pulse carries 1 mj of energy, it would be permissible to couple up to 28,000 pulses into the breast over a 6 min period. In a practical UWB microwave radar imaging scenario, an array of antennas surrounds the breast volume and each element sequentially illuminates the breast. A typical array may contain on the order of 50 antennas, suggesting that up to 560 pulses could be transmitted by each antenna without exceeding the peak 1-g SA limit suggested in the IEEE exposure guideline. Note that the level of exposure of 560 pulses per antenna is much higher than we anticipate needing for this application. 4. CONCLUSIONS We have investigated the absorption of short (120 ps, 6-GHzcarrier) microwave pulses in anatomically realistic numerical breast phantoms in an effort to formally evaluate the safety of UWB microwave breast cancer detection technology operating in the 1 11 GHz range. We have found that the SA does not vary greatly with patient-to-patient variations in breast shape, tissue composition, or fibroglandular dielectric properties. While the specific characteristics (antenna radiation patterns, coupling media properties, etc.) of future clinical systems may differ from those assumed for this computational study, the SA values are not expected to vary significantly as a function of those characteristics. The normalized SA values reported in this paper can be scaled to account for the total number of pulses radiated and different pulse energies, providing valuable guidance in the design of future clinical systems that are in compliance with safety standards. For anticipated embodiments of such a system, we conclude that UWB microwave breast cancer detection modalities pose no health risk to the patient. horn antenna with curved launching plane for pulse radiation, IEEE Antennas Wireless Propag Lett (2003), M. Converse, E.J. Bond, B.D. Van Veen, and S.C. Hagness, A computational study of ultrawideband versus narrowband microwave hyperthermia for breast cancer treatment, IEEE Trans Microwave Theory Tech 54 (2006), W.T. Joines, Y. Zhang, C. Li, and R. L. Jirtle, The measured electrical properties of normal and malignant human tissues from 50 to 900 MHz, Med Phys (1994), A.J. Surowiec, S.S. Stuchly, J.R. Barr, and A. Swarup, Dielectric properties of breast carcinoma and the surrounding tissues, IEEE Trans Biomed Eng 35 (1988), S. Chaudhary, R. Mishra, A. Swarup, and J. Thomas, Dielectric properties of normal and malignant human breast tissue at radiowave and microwave frequencies, Ind J Biochem Biophys 21 (1984), S. Gabriel, R.W. Lau, and C. Gabriel, The dielectric properties of biological tissues. III. Parametric models for the dielectric spectrum of tissues, Phys Med Biol 41 (1996), J. Hilland, Simple sensor system for measuring the dielectric properties of saline solutions, Meas Sci Technol (1997), D.M. Lanners, K.K. Amrami, R.S. Jonsgaard, J.J. Gisvold, and J.P. Felmlee, Safety and MRI artifact evaluation at 1.5 T of metallic mounting sheath of a marking clip inadvertently deployed at stereotactic biopsy, Am J Roentgenol (2004), K. Caputa, M. Okoniewski, and M.A. Stuchly, An algorithm for computations of the power deposition in human tissue, IEEE Antennas Propag Mag 41 (1999), IEEE Standard for safety levels with respect to human exposure to radio frequency electromagnetic fields, 3 khz to 300 GHz, IEEE Standard C95.1, Wiley Periodicals, Inc. ACKNOWLEDGMENTS This work was supported by the National Institutes of Health under grant F31 CA awarded by the National Cancer Institute and the National Science Foundation under grant BES REFERENCES 1. S.C. Hagness, A. Taflove, and J.E. Bridges, Two-dimensional FDTD analysis of a pulsed microwave confocal system for breast cancer detection: Fixed-focus and antenna-array sensors, IEEE Trans Biomed Eng (1998), E.C. Fear and M.A. Stuchly, Microwave system for breast tumor detection, IEEE Microwave Guided Wave Lett 9 (1999), E.J. Bond, X. Li, S.C. Hagness, and B.D. Van Veen, Microwave imaging via space-time beamforming for early detection of breast cancer, IEEE Trans Antennas Propag (2003), X. Li, S.K. Davis, S.C. Hagness, D.W. van der Weide, and B.D. Van Veen, Microwave imaging via space-time beamforming: Experimental investigation of tumor detection in multilayer breast phantoms, IEEE Trans Microwave Theory Tech (2004), J.M. Sill and E.C. Fear, Tissue sensing adaptive radar for breast cancer detection Experimental investigation of simple tumor models, IEEE Trans Microwave Theory Tech 53 (2005), A. Taflove and S.C. Hagness, Computational electrodynamics: The finite-difference time-domain method, 3rd ed., Artech House, Norwood, MA, E. Zastrow, S.K. Davis, and S.C. Hagness, Safety assessment of breast cancer detection via ultrawideband microwave radar opoerating in swept frequency mode, In Proceedings of IEEE International Symposium on Antennas and Propagation, Albuquerque, NM, July 2006, in press. 8. X. Li, S.C. Hagness, M.K. Choi, and D.W. van der Weide, Numerical and experimental investigation of an ultrawideband ridged pyramidal APPLICATION OF THE FOLDY LAX MULTIPLE SCATTERING METHOD TO THE ANALYSIS OF VIAS IN BALL GRID ARRAYS AND INTERIOR LAYERS OF PRINTED CIRCUIT BOARDS C.-J. Ong, 1 D. Miller, 2 L. Tsang, 1 B. Wu, 1 and C.-C. Huang 2 1 Department of Electrical Engineering University of Washington Paul Allen Center, Room AE100R Campus Box Seattle, WA Intel Corporation Jones Farm 5, 2111 N.E., 25th Avenue Hillsboro, OR Received 12 May 2006 ABSTRACT: The authors applied the method of Foldy Lax multiple scattering equations to multiple vias in ball grid arrays and the interior layers of printed circuit boards. The method gives the scattering parameters of the array. The results are verified with Ansoft s HFSS with very little difference between the Foldy Lax approach and the HFSS results. The CPU required for the method is only a small fraction of the time that HFSS requires. For a array of vias, the method calculates all the S-parameters of the 512 ports in 4 min for 10 frequencies on a Pentium 3.2 GHz PC Wiley Periodicals, Inc. Microwave Opt Technol Lett 49: , 2007; Published online in Wiley InterScience ( DOI /mop Key words: vias; multiple scattering; signal integrity; ball grid arrays; printed circuit boards DOI /mop MICROWAVE AND OPTICAL TECHNOLOGY LETTERS / Vol. 49, No. 1, January