band applications with a reasonable LO power of 0 dBm, moderate OP1dB and high LO-to-RF suppression, compared with previously reported mixers. 4. CONCLUSION In this article, a wide IF bandwidth up-conversion mixer has been realized using the commercial standard bulk 90-nm 1P9M CMOS process. Compared to the conventional double-balanced Gilbert-cell mixer structure, the proposed mixer consists of an active balun to avoid extra power consumption. The RCfeedback results in a phase difference below 58 and a magnitude difference below 1 dB, from dc to 10 GHz. Moreover, the series inductive-peaking techniques used by the integrated transformer achieve a wide IF 3-dB bandwidth from dc to 3.8 GHz. The LO 3-dB frequency bandwidth ranges from 53 to 65 GHz. The proposed mixer is suitable for high data rate wireless applications. ACKNOWLEDGMENTS The authors acknowledge chip fabrication and measurement support provided by National Chip Implementation Center (CIC) and National Nano Device Laboratories (NDL), Taiwan. REFERENCES 1. M.C. Chen, H.S. Chen, T.C. Yan and C.N. Kuo, A CMOS Upconversion mixer with wide if bandwidth for 60-GHz applications, Proc IEEE Silicon Monolith Integr Circuits RF Syst (2009), 1–4. 2. M. Kraemer, D. Dragomiresucu, and R. Plana, A dual-gate 60 GHz direct up-conversion mixer with active IF balun in 65 nm CMOS, 2010 Int Conf Wireless Information Tech Syst (2010), 1–4. 3. J.C. Jeong, I.B. Yom, and K.W. Yeom, An active IF balun for a doubly balance resistive mixer, IEEE Microwave Wireless Compon Lett 19 (2009), 224–226. 4. H. Ma, S.J. Fang, F. Lin, and H. Nakamura, Novel action differential phase splitters in RFIC for wireless applications, IEEE Trans Microwave Theory Tech 46 (2005), 2597–2603. 5. I. Song, J. Lee, C. Byeon, S. Cho, H. Kim, I. Oh, and C. Park, 60 GHz double-balanced drain-pumped up-conversion mixer using 90 nm CMOS, 2013 IEEE MTT-S Int Microwave Symp Dig (2013), 1–4. 6. T.M. Tsai and Y.-S. Lin, A 15.1 mW up-conversion mixer with 4.5 dB gain and 57.5 dB LO-RF isolation, Electron Lett 48 (2012), 844– 845. 7. P.S. Wu, C.H. Wang, C.S. Lin, K.Y. Lin, and H. Wang, A compact 60 GHz integrated up-converter using miniature transformer coupler with 5 dB conversion gain, IEEE Microwave Wireless Compon Lett 18 (2008), 641–643. C 2016 Wiley Periodicals, Inc. V UWB PATCH ANTENNA AND BREAST MIMICKING PHANTOM DESIGN AND IMPLEMENTATION FOR MICROWAVE BREAST CANCER DETECTION USING TIME REVERSAL MUSIC Mamadou Hady Bah, Jing-Song Hong, and Deedar Ali Jamro School of Physical Electronics, University of Electronic Science and Technology of China, Chengdu 610054, China; Corresponding author: sintaba@yahoo.com Received 9 July 2015 ABSTRACT: In this manuscript, a small size UWB patch antenna and a breast mimicking phantom are designed and implemented for microwave breast cancer detection using Time Reversal MUSIC. Some new DOI 10.1002/mop techniques are applied to the antenna in order to achieve a broad bandwidth, high gain, and also to avoid the energy radiation toward the ground plane of the antenna. The resulting dielectric constant of the breast phantom is relatively close to the real normal breast tissues. After the design has been completed, some techniques of Time Reversal MUSIC were employed to mimic the breast cancer detection. The experimental results show that both temporal and spacial images of the target C 2016 Wiley Periodicals, Inc. Microwave (tumor) are well represented. V Opt Technol Lett 58:549–554, 2016; View this article online at wileyonlinelibrary.com. DOI 10.1002/mop.29613 Key words: UWB patch antenna design; antenna broad bandwidth; breast mimicking phantom; microwave breast cancer detection; Time Reversal MUSIC 1. INTRODUCTION Breast cancer is among others, one of the major diseases that women fear the most. It is considered one of the main causes of world width women mortality [1]. In the last decade, many efforts have been made and still are in progress in order to overcome this issue [2]. However detection and treatment in earlier stages remain the best hope to long-term survival from the disease [3–8]. The X-ray mammography, considered as the standard screening method for earlier breast cancer diagnosis, the results turn out still with many limitations. Approximately up to 34% of breast tumors are failed to be diagnosed by this method [6,7] whereas almost 70% of the breast tumors are identified as benign [8]. It is characterized by a high false positive and negative rate. In addition, it uses ionizing radiation and breast compression during the examination process. These drawbacks are the most inspiring that brought many researchers to seek for a new and promising method to complement the mammography in the early stage breast cancer diagnosis. Alternatively, microwave imaging is a low cost and improved safety technology system for breast cancer imaging. More details can be found in Refs. 5–9. Due to its wide applicability, it can be divided into passive and active microwave imaging. Many investigations have been made for both two techniques in the case of breast cancer detection [10]. Passive microwave imaging referred to as radiometry [11,12] uses the differences in temperature between healthy and timorous tissues. For active microwave imaging, it relays upon the difference in dielectric properties of the human tissues [4,7,13]. In the literature’s studies, the permittivity of malignant tissues (cancer) was found to be 10–20% higher than the healthy tissues [14]. With the increase of breast cancer mortality in the recent decades, many methods using breast phantoms have been studied and experimented in order to estimate the real breast tissues properties [2,7,14]. The use of oil in gelatin has been implemented by Lazebnik et al. in Ref. 14. This method consists of mixing these two materials and adding surfactant liquids. It has been also utilized by Madsen et al. [15] in the field of ultrasound for tissues mimicking. For this study, a mixture of some daily life use materials having different dielectric properties is considered as breast phantom. In microwave imaging system, the EM wave propagation through biological tissues is characterized by some refraction and multipath properties. This could lead the back-scattered signals to reach the detector by multipath. These conditions return some limitations to the identification of the target location. To overcome this issue we make use of time reversal multiple signals classification (TR_MUSIC). It has the ability to employ the MICROWAVE AND OPTICAL TECHNOLOGY LETTERS / Vol. 58, No. 3, March 2016 549 Figure 1 (a) Parametric design; (b) fabricated antenna with reflector. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary. com] multipath propagation to achieve good results in target detection such as breast cancer detection [1]. In TR_MUSIC more complex is the imaging medium more the imaging result is obvious. As the human body has different type of tissues having different electrical properties which make it complex, we have found that TR_MUSIC is one the most effective methods for breast cancer detection. However it has some limitations especially for dense breast tissues. In this imaging system, antenna is a very important tool which facilitates the transmission and/or reception of the EM waves. Ever since engineers started using microwaves for medical applications, the search for a suitable antenna has strongly attracted their attention. Many antennas have been used in microwave medical imaging. Among them, patch is a compact and versatile design that is characterized by the potential advantages to be low cost, light weight, easy design, broad bandwidth property, and has greater availability. 2. MATERIALS AND METHODS In this manuscript, our aim is to carry out an experimental study on breast cancer detection based on time reversal (TR_MUSIC) method using a previously designed patch antenna and a breast mimicking phantom. For safety proposes, the designed antenna and breast phantom have been chosen to operate within the range allocated by ISM (Industrial, Science and Medical) which is from 10 MHz to 20 GHz. For this paper, all design and simulations were performed using computer simulation technology Software (CST Microwave studio), Matlab, and origin Pro. To confirm the results obtained through simulation, some experimental measurements are performed bellow. 3. GEOMETRY OF THE SINGLE MONOPOLE ANTENNA Due to the stability in the impedance bandwidth of rectangular patch antenna compared to elliptical and circular monopoles [16], we choose to use the rectangular shaped antenna, operating in the UWB range 3.1–10.6 GHz authorized by FCC. The initial dimensions of the radiating patch LP and WP are taken as k=4 at the lower frequency point. In Figure 1 the radiating antenna (colored in pink) is designed on top of a FR_4 substrate having a dielectric constant of 4.6 and a height of 1.2 mm. LS 535, WS 5 44, WP 5 24, and LP 5 17 are length and width of the substrate and the radiating patch, respectively. In its opposite face a reduced and blended ground plane is placed. The radius of the Figure 2 (a) Antenna array surrounding the breast model. (b) The excitation signal. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com] 550 MICROWAVE AND OPTICAL TECHNOLOGY LETTERS / Vol. 58, No. 3, March 2016 DOI 10.1002/mop Figure 3 Compared S-parameters of the UWB antenna. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary. com] blended edges is 16 mm. The side length of the square slot is a 5 7 mm. Figure 1 illustrates the parametric design and the fabricated antenna with and without reflector. LF 510.5 is the length of the feed-line; Lb 2.5 and Wb 5 15 are length and width of the feed-step which also contributes to the input port matching; and finally WG 5 40 is the width of the ground. All dimensions are in millimeter (mm). Figure 5 Ground blending effect of the antenna. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com] D2 Eðr; tÞ 2 lðrÞeðrÞ @2 Eðr; tÞ 5 0 @t2 (1) An electromagnetic wave is called time reversal if the wave can propagate backward, in such case the signal would refocus to the point source. In another words, Time Reversal is a new technique which satisfies the reciprocity theorem EðtÞ5Eð2tÞ; meaning that if Eðr; tÞ is a solution of Eq. (1), then the existence of another solution will be observed so-called Time Reversed Eðr; tÞ5Eðr; 2tÞ . For simplicity, here we consider the electromagnetic wave equation in a uniform and loss-free medium. In this equation, Eðr; tÞ is the electric field; ~ r is the vector position; l; e are permeability and permittivity, respectively. The reciprocity theorem governing this wave equation allows the scattered wave to reverse and propagate backwards to the original source. The TR wave would focus at the point source regardless of the nature of the medium. The basic idea about TR consist of recording the different received wave signals using an array of transducers, time reverse, and send it back to the original point source. It can exploit the multipath effect to achieve good results [17]. TR microwave imaging techniques use the multi-static data matrix (MDM). Different imaging methods for microwave breast cancer imaging are investigated in Refs. 3,4,17 using Time Figure 4 Compared Gain of the antenna. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com] Figure 6 Slot effect of the monopole antenna. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com] 4. TR_MUSIC AND MULTI-STATIC IMAGING ALGORITHM DOI 10.1002/mop MICROWAVE AND OPTICAL TECHNOLOGY LETTERS / Vol. 58, No. 3, March 2016 551 K5UuV H (6) After substituting the singular value decomposition, K, into the time reversal operator defined as T5K H K; then we will getK5UuV H VuUH where the upper script H denotes conjugate transpose operation. Here U and V are both unitary matrixes; so the finally time reversal operator can be deduced as T5u2 : For DORT, every eigen value of u2 signifies the presence of the cancer in the investigated region. The back propagated corresponding eigenvectors facilitate the identification of the targets in that region. There is a strong relationship between the eigenvectors (v) and the Green’s (g) function vector of the medium [11,18] as shown in (11) . vm ðxÞ  eju Figure 7 The dielectric constant of the breast phantom Reversal. Among the well-known methods, DORT and TR MUSIC are based on either the eigen values or its equivalent singular value decomposition of the MDM matrix. Here, we propose the use of Time Reversal multiple signals classification (known as TR-MUSIC) technique to estimate the location of a tumor in a breast phantom. For that, let us consider an N antenna elements located at rj. To facilitate the localization of M targets in a two dimensional plane, we apply an excitation signal ej (x) to the jth element of the antenna array with j 5 1, 2, 3, . . . N. ej (x) is the excited electric field with respect to the frequency domain. The electric field, after being propagated in the region under investigation, the scattered energy by the Mtarget is recorded by the lth element of the array and it can be expressed as M X ESlj 5 GðXm ; rl ; xÞsm ðxÞGðrl ; Xm ; xÞej ðxÞ (2) m51 g ðxm Þ kgðxm Þk (7) where u is the phase arising from singular value decomposition and upper script star denotes conjugate operation. Consequently, the DORT imaging function that utilizes the signal subspace can be formulated as IDORT ðrÞ5 M X 2 jhvn jgðr; xÞij (8) n51 Due to the orthogonality between noise and signal subspaces, time reversal can make use of the singular value decomposition of the multi-static data matrix to separate them. Then the scattering object located at r will be estimated by the time reversal MUSIC pseudo-spectrum formulated in Eq. (9). ITR MUSIC ðrÞ5 X N 1 2 This condition is satisfied only when the term hvn jgðrÞi is set to be null. Here vn stands for the n eigenvectors and gðrÞ is the vector Green’s function. where sm (x) is the scattering potential, representing the backscattering strength as a function of angular frequency; Xm and rj represent the location of mth target, and the jth receiving element; k is the wave propagation constant and G is the background Green’s function of the medium. For simplicity, in the following section, we omit the frequency term in the mathematical formulation. The l,jth element of the multi-static data matrix can be written as M X Klj 5ESlj 5 sm gr ðXm ÞgTt ðXm Þ (3) m51 where the upper script T denotes transpose operation; gr and gt are vectors Green’s function for transmission (subscripts t) and reception (subscripts r) respectively. gr and gt can be define as gr ðXm Þ5½Gðr1 ; Xm Þ; Gðr2 ; Xm Þ; :::; GðrN ; Xm Þ (4) gt ðXm Þ5½GðXm ; r1 Þ; GðXm ; r2 Þ; :::; GðXm ; rN Þ (5) By taking into account the reciprocity theorem in the propagation medium, the MDM becomes symmetric and can be express as Klj 5Kjl . The singular value decomposition, K, can be represented as 552 (9) jhvn jgðrÞij n5M11 Figure 8 Time reversal focused signal MICROWAVE AND OPTICAL TECHNOLOGY LETTERS / Vol. 58, No. 3, March 2016 DOI 10.1002/mop Figure 9 The location of the tumor (cancer) inside the breast phantom at 6.85 GHz. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com] 5. EXPERIMENTAL SETUP For our experimental study, four antennas have been considered, symmetrically placed around the breast phantom that has been predesigned. The antennas lie on a circle with a radius of 100 mm. So that the distance between two symmetric antennas is 200 mm. An image illustrating the experimental set up is shown in Figure 2. During the experiment, these antennas play both the role of transmitting and receiving; however only one antenna transmit at a time. In the process, the antenna elements (numbered from 1 to 4) around the breast phantom transmit a short pulse signal into the imaging medium, after its propagation, the backscattered signals are recorded by the receiving antennas and then retransmitted back to the original source. The more the reversed signal tends to the point source the more the signal is compressed. The maximum amplitude of the time reversed signal is observed at the point source. In Figure 2, the breast phantom is composed of three different materials: corn flour, soy bean oil, and water in a ratio of 6.5:2:1.5. These materials have been chosen because when they are mixed in this ratio, it displays a dielectric constant similar to that of the real normal breast tissues. In the case of the breast cancer (tumor), due to the high dielectric constant of the malignant tissues (cancer or tumor), mineral water has been considered, put in a unit cell glass and imbedded in the phantom. The top of the cell glass was covered with a wood stick (see Fig. 2). 6. SIMULATION AND EXPERIMENTAL RESULTS This section shows a good agreement between the simulation and measured results. In Figure 3, the results illustrate a return loss less than 210 dB over the entire frequency range. As expected, a broad bandwidth is obtained for both simulated and measured results. The little variation observed in the results is due to some errors acquired during the port welding and the surrounding environment in the experiment. In the parametrical design figure, a slot is made on the radiating patch where less electric current distribution is observed. A diamond structure is DOI 10.1002/mop cut out in order to control the surface current and match the low frequency. In the opposite plane, a reduced ground plane has been adopted. In this antenna structure, the length of the ground is reduced to k=10 over a width of less than k=2 at the low operating frequency; a blending edge is applied to the upper edges of the ground. It is well known that a microstrip patch antenna fully ground bounded is more directional; however due to the confinement of the fringing fields in between the radiating patch and the ground plane in that case, the bandwidth of the antenna could be affected, leading to a narrow bandwidth of the radiator as well as increasing the size of the antenna. By taking into account the relationship between the skin depth and the operating frequency [see Eq. (10)], some techniques are applied to the design in order to achieve good penetration depth and high resolution as well. kp 5 c sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  2 11 e0r =er 21 2pf 2er (10) The blending edges of the ground together with the slot on the radiator contribute to the improvement of the bandwidth and the miniaturization of the antenna as well. Figure 4 shows up the simulated and measured high gain. The effects of blending the edges of the ground plane and the slot in the middle of the radiator are depicted in Figures 5 and 6, respectively. Below are shown the dielectric constant of the breast phantom (Fig. 7) and the resulting waveform with good temporal focusing signals (see Fig. 8). For this breast phantom, we have been able to achieve a relative permittivity between 9 to 12 which is relatively close to the real normal breast tissues as presented in Refs. 7,19. This is due to the low water content in it. In Figure 8, the resulting waveform explains the ability of the Time Reversal technique to refocus the signals from the receiving positions back to the source. In a more realistic approach, a comparison between the simulated and experimental results is MICROWAVE AND OPTICAL TECHNOLOGY LETTERS / Vol. 58, No. 3, March 2016 553 carried out in this study using TR-Music algorithm formulated in Eq. (9). A good estimation of the target location is shown in Figure 9. Refer to Figure 2, we can see that the target is located in the middle of the imaging medium with a set of antennas placed around the breast. 7. CONCLUSION Through this manuscript, a UWB patch antenna operating in the range of 3.1–10.6 GHz (according to FCC) is designed and implemented for breast cancer detection using time reversal MUSIC. For a further understanding approach, some comparisons are conducted between the results obtained from simulation and measurement. One of the big advantages of this antenna is that its size has been reduced without the use of high dielectric constant or high frequency as employed in Ref. 10. As characteristic of this design, a broad bandwidth and a relatively high gain are achieved. In microwave breast cancer imaging, the commonly used technique is to rotate the transmitting antenna around the breast and this is to facilitate the cancer (tumor) detection. As the target is located in front of the transmitting antenna, we need to focus almost all the radiated energy in forward direction. By placing the reflector in the back side of the antenna, it contributes to improve not only the gain but the directivity of the antenna as well. This experimental test was carried out in free space as depicted in Figure 2. The designed antenna can be used for multiple applications using broadband in the UWB frequency range from 2 to 12 GHz. A dielectric constant relatively close to the real normal breast tissue is achieved, refer to Refs. 7,19. In this experiment, all materials used are safe, non-toxic even through inhalation. The waveform in Figure 9 shows up a very narrow peak at the focusing point, meaning that the time reversed signals from the receiving points are well focused. This will facilitate the detection and localization of the cancer (tumor) from the breast. Figures 9(a) and (b) explain about the capability of the antenna to transmit and receive a narrow pulse signal, which is very important in target detection such as breast cancer detection through medical imaging. In Figure 9(b), the measured result presents some noise, this is due to the migration of the water and oil from top to the bottom of the bottle during one week deposit after the breast phantom has been modeled. From this research, we can conclude that even though TR_MUSIC has a lot of advantages in breast cancer detection, the experimental work remains with some limitations such as carrying out an experiment with highly dense breast and a breast mimicking phantom resulting from mixed materials of different dielectric properties. In simulation it would be easier to achieve good results through breast cancer detection. Finally we found it very challenging to achieve a dielectric constant relatively close to the breast tissues by mixing non-chemical materials, with different dielectric constants. Also we think there is a need of a more advanced method to complement the limitation of Time Reversal MUSIC in the case of very dense tissues. ACKNOWLEDGMENTS This work was supported by the National Natural Science Foundation of China (No. 61172115 and No. 60872029); the HighTech Research and Development Program of China (No. 2008AA01Z206); the Aeronautics Foundation of China (No. 20100180003); and the Fundamental Research Funds for the 554 Central Universities (No. 9140A07030513DZ02098. ZY-GX2009J037) and Project REFERENCES 1. M. Sajjadieh, F. Foroozan, and A. Asif, Breast cancer detection using time reversal signal processing, In: IEEE 13th International Multioptic Conference, INMIC 2009, 2009. 2. N. Nadine Joachimowicz, C. Conessa, T. Henriksson, and B. Duch^ene, Breast phantoms for microwave imaging, IEEE Antennas Wireless Propag Lett 13 (2014), 1333–1336. 3. M.D. Hossain and A.S. Mohan, Breast cancer localization in three dimensions using time reversal DORT method, In: IEEE Antennas and Propagation (ISAP), Chicago, IL 2012. 4. Y.W. Jin, J.M.F. Moura, and Y. Jiang, Breast cancer detection by time reversal imaging, In: IEEE Biomedical Imaging: From Nano to Macro, ISBI 2008, 2008. 5. W. Shao and R.S. Adams, Two antipodal vivaldi antennas and an antenna array for microwave early breast cancer detection, Microwave Opt Technol Lett 55 (2013), 670–674. 6. P.T. Huynh, A.M. Jarolimek, and S. Daye, The false-negative mammogram, Radiograph 18 (1998), 1137–1154. 7. E.C. Fear, X. Li, S.C. Hagness, and M.A. Stuchly, Confocal microwave imaging for breast cancer detection: Localization of tumors in three dimensions, IEEE Trans Biomed Eng 49 (2002), 812–822. 8. J.G. Elmore, M.B. Barton, V.M. Moceri, S. Polk, P.J. Arena, and S.W. Fletcher, Ten-year risk of false positive screening mammograms and clinical breast examinations, New Engl J Med 338 (1998), 1089–1096. 9. E.C. Fear, S.C. Hagness, P.M. Meaney, M. Okoniewski, and M.A. Stuchly, Enhancing breast tumour detection with nearfield imaging, IEEE Microwave Mag 3 (2002), 48–56. 10. K.L. Carr, Microwave radiometry: Its importance to the detection of cancer, IEEE Trans Microwave Theory Tech 37 (1989), 1862–1869. 11. B. Bocquet, J.C. van de Velde, A. Mamouni, Y. Leroy, G. Giaux, J. Delannoy, and D. Del Valee, Microwave radiometric imaging at 3 GHz for the exploration of breast tumors, IEEE Trans Microwave Theory Tech 38 (1990), 791–793. 12. S.S. Chaudhary, R.K. Mishra, A. Swarup, and J.M. Thomas, Dielectric properties of normal and malignant human breast tissues at radiowave and microwave frequencies, Indian J Biochem Biophys 21 (1984), 76–79. 13. 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), 257–263. 14. M. Lazebnik, E.L. Madsen, G.R. Frank, and S.C. Hagness, Tissue-mimicking phantom materials for narrowband and ultrawideband microwave applications, Phys Med Biol 50 (2005), 4245–4258. 15. E.L. Madsen, J.A. Zagzebski, and G.R. Frank, Oil-in-gelatin dispersions for use as ultrasonically tissue-mimicking materials, Ultrasound Med Biol 8 (1982), 277–287. 16. Abbosh and M.E. Bialkowski, Design of ultra-wideband planar monopole antennas of circular and elliptical shape, IEEE Trans Antennas Propag 56 (2008), 17–23. 17. M.E. Yavuz and F.L. Teixeira, Ultra wideband microwave sensing and imaging using time-reversal techniques: A review, Remote Sens 9 (2009), 466–495. 18. M. Delwar Hossain, F. Yang, M.J. Abedin, and A. Sanagavarapu Mohan, Time reversal microwave imaging for the localization and classification of early stage breast cancer, In: Proceedings of the Asia-Pacific Microwave Conference, 2011, IEEE, Australia, p, 477 19. S.M. Salvador and G. Vecchi, Experimental tests of microwave breast cancer detection on phantoms, IEEE Trans Antennas Propag 57 (2009), 1705–1712. C 2016 Wiley Periodicals, Inc. V MICROWAVE AND OPTICAL TECHNOLOGY LETTERS / Vol. 58, No. 3, March 2016 DOI 10.1002/mop