- Email: [email protected]
Computing stimulation voltage in a bipolar electrode configuration to avoid side effects during deep brain stimulation
Accepted Manuscript Title: Computing stimulation voltage in a bipolar electrode configuration to avoid side effects during deep brain stimulation Author: Wei-Yi Chuang Paul C.-P. Chao Shin-Yuan Chen Sheng-Tzung Tsai Kuu-Young Young Ta-Wei Ting PII: DOI: Reference:
S0924-4247(15)30038-8 http://dx.doi.org/doi:10.1016/j.sna.2015.06.016 SNA 9216
To appear in:
Sensors and Actuators A
Received date: Revised date: Accepted date:
23-5-2014 22-4-2015 17-6-2015
Please cite this article as: Wei-Yi Chuang, Paul C.-P.Chao, Shin-Yuan Chen, ShengTzung Tsai, Kuu-Young Young, Ta-Wei Ting, Computing stimulation voltage in a bipolar electrode configuration to avoid side effects during deep brain stimulation, Sensors and Actuators: A Physical http://dx.doi.org/10.1016/j.sna.2015.06.016 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Title: Computing Stimulation Voltage in a Bipolar Electrode Configuration to Avoid Side Effects during Deep Brain Stimulation
Author list: Wei-Yi Chuang1, Paul C.-P. Chao1*, Shin-Yuan Chen2,3, Sheng-Tzung Tsai2,3, Kuu-Young Young1 and Ta-Wei Ting4
Affiliation of authors: 1
Department of Electrical Engineering, National Chiao-Tung University,
Hsinchu, Taiwan 2
Department of Neurosurgery, Buddhist Tzu Chi General Hospital,
Hualien, Taiwan 3
Department of Medicine, Tzu Chi University, Hualien, Taiwan
4
Department of Biomedical Engineering, Yuanpei University, Hsinchu,
Taiwan
*Corresponding author: Paul C.-P. Chao Tel: +886-3-5131377 Fax: +886-3-5752469 E-mail: [email protected] Postal address: Room No. 735, Engineering Building No. 5, 1001 University Rd., Highlights ► Hsinchu City, 300, Taiwan. Highlights
The methodology of computing stimulation voltage while transformation stimulation configuration from monopolar to bipolar is proposed.
Graphical abstractBiography Wei-Yi Chuang was born in, Miaoli, Taiwan, in 1984. She received the B.S. degree in electrical engineering from Chung Yuan Christian University, Jongli, Taiwan, in 2006, and the M.S. and Ph.D. degrees in electrical control engineering from National Chiao-Tung University, Hsinchu, Taiwan, in 2008 and 2014, respectively. Her research interests focus on modelling, analysis and control of Parkinson’s disease utilizing deep brain stimulation system. Paul C.-P. Chao received M.S. and Ph.D. degree from Michigan State University, USA, respectively. After graduation, he worked for the CAE department of Chrysler Corp in Auburn Hill, Detroit, USA for two years. He is currently a university distiguished professor at National Chiao Tung University (NCTU) and a faculty member of the electrical engineering department. Prof. Chao was the recipient of the 1999 Arch T. Colwell Merit Best Paper Award from Society of Automotive Engineering, Detroit, USA; the 2004 Long-Wen Tsai Best Paper Award from National Society of Machine Theory and Mechanism, Taiwan; the 2005 Best Paper Award from National Society of Engineers, Taiwan; and the 2002/2003/2004 CYCU Innovative Research Award; 2006 the AUO Award; 2007 the Acer Long-Term 2nd-prize Award; The 2007/2008/2009 NCTU EEC Outstanding Research Award; the 2009 Best Paper Award from the Symposium on Nano-Device Technology; the 2010 Best Paper Award from the 20th Annual IEEE/ASME Conference on Information Storage and Processing Systems (ISPS). He is the University Associate VP of NCTU. For editorial services, he is the Technical Editor of IEEE Sensors Journal, and also the Associate Editors of ASME Journal of Vibration and Acoustics, IEEE IoT Journal and Journal of Circuit, System and Computer. In recent years, his research interests focus on dynamic analysis and designs for sensors and actuators; micro-mechatronics and control technology. He is an ASME Fellow and IEEE senior member.
Shin-Yuan Chen received the M.D. and M.S. degrees from Taipei Medical College, Taipei Medical University, Taipei, Taiwan, in 1988 and 1995. He is currently a professor at Tzu Chi Medical College, Tzu Chi University, Hualien, Taiwan. His research interests focus on Parkinson’s disease, functional neuro surgeru and spinesurgery. Sheng-Tzung Tsai received the M.D. degree from Tzu Chi Medical College, Tzu Chi University, Hualien, Taiwan, in 2003. He is currently a clinical fellow form Parkinson Disease Research Center, Division of Functional Neurosurgery, Tzu-Chi General Hospital, Hualien, Taiwan. His research interests focus on Parkinson’s disease, consciousnessdisease, neckandlowbackpain, and spinesurgery. Kuu-young Young was born in Kaohsiung, Taiwan, 1961. He received his B.S. degree in Electrical Engineering from National Taiwan University, Taiwan, in 1983, and M.S. and Ph.D. degrees in Electrical Engineering from Northwestern University, Evanston, IL, U.S.A., in 1987 and 1990, respectively. Between 1983 and 1985, he served as an electronic officer in Taiwan Navy. Since 1990, he has been with the Department of Electrical Engineering at National Chiao Tung University, Hsinchu, Taiwan, where he is currently a Professor. He served as the chairman of the department from 2003 to 2006, and the associate dean of Electrical and Computer Engineering College, NCTU, from 2007 to 2010 and also 2014 to 2015. His research interests include robot compliance control, robot learning control, robot calibration and path planning, teleoperation, robot walking helper, and Science, Technology, and Society (STS). Ta-Wei Ting received the Ph.D. degrees in biomedical engineering from Johns Hopkins University, USA in 2000. He is currently an associate professor at Yuanpei University, Hsinchu, Taiwan. His research interests focus on 3D image reconstruction, bioreactor design and biofluid mechanics. ABSTRACT Deep brain stimulation (DBS) in the subthalamic nucleus (STN) has been applied for advanced Parkinson’s disease (PD). In clinical practices, monopolar
configuration is often utilized based on clinical efficiency. Meanwhile, the volume of tissue activated (VTA) by DBS is estimated for controlling and/or limiting the stimulated region. However, side effects like dyskinesia and tremor were accompanied during stimulation. Most studies had found that bipolar stimulation is an alternative approach for side effects avoidance. The goal of this paper is to develop a quantified relationship of stimulation voltage from monopolar to bipolar configuration to improve neural stimulation. In this paper, an electromagnetic finite element model is first built for a patient-specific physiological brain model, which is established by magnetic resonance imaging (MRI) data. The model is then used for finite element analysis (FEA) to estimate VTA with varied electrode voltage in monopolar and bipolar configurations. With the goals to avoid side effects and achieve symptoms suppression, the stimulation voltage of bipolar configuration is successfully computerized based on the electromagnetic FEA simulations. Experiments are conducted successfully to validate simulation results.
Keywords: Deep brain stimulation (DBS), Volume of tissue activated (VTA), Side effects, Stimulation configuration, Magnetic resonance imaging (MRI), Finite element analysis (FEA). 1. INTRODUCTION Over the last decade, deep brain stimulation (DBS) has been a useful and efficient approach to treat Parkinson’s disease (PD) with implanting electrodes
chronically into a selected brain target, subthalamic nucleus (STN). Having electrode implanted, a brain-electrode interface (BEI), which consists of three main components: implanted electrode, layer of peri-electrode space surrounding the electrode, and surrounding brain tissue, is formed [1, 2]. Because the brain tissue content in the peri-electrode space varies with time, stimulation parameters should also be adjusted during DBS to maintain the efficacy of therapy [3]. In addition, the quantified representation of neuronal responses during stimulation is defined as the volume of tissue activated (VTA). Besides stimulation parameters including the stimulation voltage, the pulse width, and the frequency, stimulation configurations including monopolar and bipolar may be selected based on clinical conditions to achieve optimal therapeutic improvement. In monopolar stimulation, one of the electrode contacts is programmed as cathode and implantable pulse generator (IPG) as anode. Whereas in bipolar settings, two contacts of the electrodes are activated. One is cathode and the other anode. In clinical practices, monopolar stimulation is frequently used because it requires lower stimulation intensity than bipolar stimulation to achieve approximately the same clinical benefit [4]. Meanwhile, the corresponding side effects are invoked when current diffusion covers undesired part of the brain. In this case, changing the activated electrode contact is an intuitive approach to address this issue. However, the suppression of PD symptoms will not achieve the same level as that with the original electrode setting. An alternative approach is to change the
stimulation configuration from monopolar to bipolar stimulation. Some previous studies had found that bipolar stimulation can reduce the occurrence of side effects due to a narrower and more focused current field during stimulation [4, 5]. In this way, the issue becomes how to adjust stimulation voltage when stimulation configuration is changed from monopolar to bipolar. The aim of this work is to quantify the difference between monopolar and bipolar stimulation. An electromagnetic finite element model is first built for a patient-specific physiological brain model, established by magnetic resonance imaging (MRI) data. The model is then utilized for finite element analysis (FEA) to estimate VTA with varied stimulation parameters, like applied electrode voltage. Moreover, the relationship between stimulation voltage and the VTA is developed via regression models. To reduce side effects, stimulation configuration is changed from monopolar to bipolar stimulation in clinical trials. Furthermore, to achieve the approximately same level of suppression of movement symptoms, the stimulation voltage in a bipolar configuration is estimated by the proposed mathematical model based on the electromagnetic FEA simulations. 2. MATERIALS AND METHODS 2.1 Patient Specific Brain Model for Estimating the VTA A BEI finite element model is constructed, referring to [1, 2]. The model consists of three main components: the implanted electrode, the surrounding brain tissue, and the layer of peri-electrode space. Model 3389 DBS lead, which con-
tains four 1.5 mm-in-length and 1.27 mm-in-diameters electrodes, marked as contacts 0, 1, 2, and 3, respectively, is selected for the BEI model simulation. The brain tissue is then modelled as a cylinder surrounding the electrode lead with a radius of 5 cm and a height of 14 cm. And, the layer of peri-electrode space has a thickness of 0.25 mm [2, 3]. Doubling the density of the mesh or doubling the distance of the boundary from the electrode generates a potential distribution that differs by < 2% when compared to the default model. In clinical practice, the range of the stimulation frequency is about 90-170 Hz for optimal motor control with Parkinson’s disease patients [6]. The medium is thus taken as a quasi-static model. The equation for computing the electrical potential distribution of the BEI model can accordingly be reduced to Laplace’s equation: ∇ ⋅ σ∇V = 0
(1)
Where V is the potential (V) and σ the conductivity (S/m), which is a constant for different types of brain tissue. Meanwhile, the Dirichlet boundary condition (V=0) is set on the sides of the brain tissue. The non-active electrode contact abides by a continuity constraint and the insulating surfaces are treated as electrically insulated. In addition, electrical-related parameters are listed in Table 1 and the resulting mesh of the BEI model contains 41,615 nodes for FEA. Smaller elements of the resulting mesh are near the lead and larger elements are at the boundaries of the brain tissue. According to [7], the second difference of the elec-
trical potential distribution in the direction perpendicular to the electrode contact is used to define the neuron activation threshold of the tissue around the electrode. And, the activation threshold E* can be computed as [8] E * = 12.3 × D -0.43
(2)
where D is the normalized distance on monopolar and bipolar configurations: 1 / D = contact 1 / axon 1 + contact 2 / axon 2
(3)
where contacts 1 and 2 are the voltages on electrode contacts 1 and 2, respectively, and axons 1 and 2 are the distance from axon to contacts 1 and 2, respectively. To locate the VTA, a correct patient physiological construction model should be established. The patient physiological construction model can be reconstructed via the prior- and post-operation magnetic resonance imaging (MRI) data by the medical software package, Mimics. The estimated VTA is thus superimposed upon the patient physiological construction model, shown in Fig. 1 [3]. Based on this model, the doctor could accordingly modify the stimulation region within the desired region of STN by selecting optimal electrode contact(s) during clinical practice.
2.2 Regression Model of the VTA The quantified representation of the VTA is essential for controlling the stimulation region. According to [7], the region of the VTA in the 2D spatial contours is similar to an ellipse, which is formulated as
(x − x0 )2 (y − y0 )2 + =1 a2 b2
(4)
where ( x0, y0 ) is the center of the ellipse, and a and b are the semi-major and semi-minor axes, respectively. Variation in the stimulation voltage leads to change in the size of the VTA, consequently the shape of the ellipse. To formulate the relationship between the stimulation voltage and the VTA, we derive linear regression models with the stimulation voltage as an independent variable and a and 2b as response variables based on curve fitting. The response variables are formulated as a = f1 ( xV )
(5)
2b = f 2 ( xV )
(6)
where xV is the stimulation voltage and f1 and f2 are linear regression models of the VTA.
2.3 Optimal Stimulation Voltage in Bipolar Stimulation To reduce side effects, the stimulation configuration is changed from monopolar to bipolar in clinical trials. With bipolar stimulation, the current diffusion is limited to the specific region in STN [9]. Besides that, to achieve the same level of symptom suppression as that in monopolar stimulation, stimulation voltage for the bipolar configuration should be adjusted.
In this study, the direction of axons is assumed to be fixed according to [5, 7, 8]. Because the device of the clinical DBS system is voltage controlled, the current distribution around the activated electrode contact(s) depends on the impedance of the tissue around the activated electrode contact(s). According to [10], the impedance would be influenced by varying pulse widths and frequencies. As this study focused on the stimulation effects with various stimulation voltages, it is assumed that if stimulation regions are the same, the effects of stimulation should also be the same. In other words, to avoid side effects and maintain the benefit of monopolar stimulation, the region and location of the VTA in bipolar stimulation should be similar to that in monopolar. In this study, the electrode contacts of bipolar configuration are selected based on the following rules while stimulation configurations are changed. The cathode of bipolar stimulation is unchanged as that in monopolar stimulation and the anode of bipolar stimulation is selected as an adjacent contact. Based on Eqs. (5) and (6), the shape of the VTA is ellipsoid-like characterized with lengths depending on the stimulation voltage. The VTA can then be quantified as an ellipsoid. To maintain the same benefit as monopolar stimulation, the size difference of the VTAs with two stimulation configurations should be as small as possible. We formulate it as 4 2 2 min π (abipolar × bbipolar − amonopolar × bmonopolar ) V 3
(7)
With the Newton’s method based on optimization theory, the suitable stimulation voltage of bipolar stimulation is determined via finding the minimum value of volume difference. 2.4 Clinical Evaluation Two PD patients undergoing DBS surgery (1 female and 1 male, age 64 and 66) were included. All patients signed informed consents for STN-DBS surgery, the procedures involved in the study and scientific publication of contents and results. The evaluation procedures in the study were carried out with the ethical approval of the institutional review board (Tzu Chi General Hospital, Hualien, Taiwan) and in compliance with the Helsinki Declaration. The inclusion criteria were as follows: (1) good levodopa response on the Unified Parkinson’s Disease Rating Scale (UPDRS) part III (motor) (>30%), (2) drug-related complications (e.g.,
dyskinesia
and
“on-off
phenomenon”),
even
under
optimal
anti-parkinsonian medication adjustment, (3) no structural lesions in the brain MRI, and (4) absence of dementia. The severities of the motor symptoms of PD (UPDRS parts III) were evaluated 1 month prior to surgery and also 3 months after. An acute stimulation test was performed 1 week after operation to identify the optimal contact (among four electrodes) with good motor benefit and avoid neuropsychological effects (such as facial twitching or depression) for chronic stimulation. The test items include speech, facial expression, tremor at rest, action tremor, rigidity, finger taps, hand movements, hand pronation, supination, leg
agility, arising from chair, posture stability, dyskinesia and body bradykinesia [11-13]. Each measurement was on a 5-point scale (range 0-4) with higher scores indicating more severely affected symptoms. The total UPDRS part III (motor) scores of these two patients were 18 and 23 with DBS ON. In this study, we focused on dyskinesia side effect improvement while stimulation configuration was changed from monopolar to bipolar. Having their post-operative status remained stable for at least 6 months, these two patients were included and underwent the experiment with proposed DBS parameters adjustment again. Note that the UPDRS ratings were single-blinded. We retested their UPDRS motor scores and dyskinesia side effect scores after new stimulation voltage and stimulation configuration had worked more than one hour to eliminate the delayed effect of stimulation.
3. RESULTS AND DISCUSSION With the goal of avoiding side effects and controlling the VTA to be within the interested region in the brain to achieve symptoms suppression, we apply the following procedure to determine the optimal stimulation voltage in a bipolar configuration. The first step is to simulate the VTA with monopolar and bipolar stimulation after implanting the DBS lead (Sec. 3.1) based on the BEI model of patient-specific physiological construction. The next step is to construct the regression model which describes the relationship between the stimulation voltages
and the size of VTAs (Sec. 3.2). In the third step, the stimulation configuration is changed from monopolar to bipolar. And, the optimization method is used to find the suitable stimulation voltages in bipolar stimulation which can achieve the same level of symptoms suppression as that in monopolar stimulation (Sec. 3.3). Finally, the in vivo experiment results are given (Sec. 3.4). 3.1 The Estimated VTAs in Monopolar and Bipolar Stimulation The specific brain region is stimulated with the voltage (1.5 V), pulse width (110 µs), and frequency (130 Hz). Note that the stimulation voltage is taken as absolute value in this study. The electrical potential distribution could be generated by the Laplace’s equation. The stimulation voltage of bipolar in this work is selected with the same amount but opposite polarization on two electrode contacts, a useful way in clinical practices. When the ellipse is fitted to the VTA based on Eq. (3), a is determined to be the distance from the electrode contact to the VTA boundary in the direction perpendicular to the electrode contact, and 2b the distance within the VTA boundary in the direction parallel to the electrode contact, as shown in Fig. 2(a) and (b), respectively. For this case of voltage (1.5 V), pulse width (110 µs), and frequency (130 Hz), a and 2b are found to be 2.385 mm and 1.800 mm, respectively, in monopolar stimulation and 1.995 mm and 5.150 mm, respectively, in bipolar. The estimated VTA is then used for the MRI patient-specific physiological model. The estimated VTA via FEA and the surrounding STN are shown in Fig.
2(c). The BEI and patient-specific physiological models are found to be effective in evaluating whether the side effects occurred during the stimulation, thereby providing valuable information for the selection among the electrode contacts. 3.2 Relationship between VTA and Stimulation Voltage To investigate the effects of VTAs due to various stimulation voltages, the stimulation voltage is set in the range of 0.5 to 5.0 V with a step size of 0.25 V, the pulse width at 110 µs, and the frequency at 130 Hz, consistent with typical clinical DBS parameters settings. Figures 3 and 4 show the resultant size of the VTA represented by a and 2b (for the semi-major and minor axes of the ellipse) versus the stimulation voltage in monopolar and bipolar stimulation, respectively. It is found that larger stimulation voltages result in larger VTAs in monopolar stimulation, so do in bipolar. And, a is more sensitive to variations in the stimulation voltage than is 2b in monopolar stimulation. In bipolar stimulation, the sensitivity of a and 2b to variations in the stimulation voltage are similar, as shown in Fig. 4. In other words, the size of the estimated VTA is more focused to variation in the stimulation voltage in bipolar stimulation than that in monopolar. To quantify the relationship between the stimulation voltage and the size of the VTA, curve fitting that minimizes the squared error is used to obtain the linear regression model below: amonopolar = 0.34 + 1.51xV − 0.15xV2
(8)
2bmonopolar = 1.38 + 0.27 xV
(9)
abipolar = 1.61 + 0.211xV
(10)
2bbipolar = 4.80 + 0.19 xV
(11)
where xV is the stimulation voltage and R2 values in fitting the model are 0.98, 0.98, 0.93, and 0.94 for Eqs. (8)-(11), respectively. In clinical trials, the doctor tunes the stimulation parameters in a monopolar configuration based on the clinical efficiency. The size of VTA can then be predicted by stimulation voltage via Eqs. (8) and (9). However, the corresponding side effects occur in monopolar during DBS parameters adjusting. The stimulation configuration is thus changed from monopolar to bipolar. In next step, the optimization method is used for finding the optimal stimulation voltage in bipolar stimulation to maintain the same benefit as that in monopolar stimulation.
3.3 Optimal Stimulation Voltage in a Bipolar Configuration To avoid side effects and maintain the benefit of monopolar stimulation aforementioned, the size difference of VTAs controlled by the stimulation voltage in monopolar and bipolar configurations should be minimized based on Eqs. (7)-(11). When a and 2b in monopolar stimulation are 2.385 mm and 1.800 mm, respectively, the optimal stimulation voltage is found at 1.6 V in bipolar stimulation.
To estimate the relationship of stimulation voltage between monopolar and bipolar stimulation, the optimal stimulation voltages in bipolar are calculated via Eqs. (7)-(11). The relationship of stimulation voltage is shown in Fig. 5. It is found that the amplitude of stimulation voltage in bipolar stimulation is higher (0.3-0.4 V on average) than that in monopolar with the minimal size difference of VTAs. Indeed, to maintain the same level of symptoms suppression in bipolar stimulation, the stimulation voltage is higher than that in monopolar in clinical practices [4, 9]. The size difference of VTAs between monopolar and bipolar stimulation is larger when the potential of stimulation voltage is larger. Figure 6 shows the difference of the distance perpendicular (a) and parallel (2b) to the electrode contact between monopolar and bipolar stimulation vs. the stimulation voltages in monopolar stimulation. In Fig. 6, the difference of distance perpendicular to the electrode contact is over 1 mm when monopolar stimulation voltage is over 3 V. The larger size difference of the VTAs is caused by the higher stimulation potential. The benefit of monopolar stimulation could then not be maintained by bipolar stimulation.
As a fact, the symptoms suppression in bipolar stimulation
is not able to reach as that in monopolar stimulation when stimulation voltage is higher than 3.6 V in clinical trials [9]. According to our primary results, the trend of stimulation voltage adjustment
between monopolar and bipolar configurations can be estimated. Moreover, the platform of modeling the VTA via different stimulation voltages during DBS is constructed. In clinical practices, physicians tune the stimulation parameters depending on the subjective response and the objective examination of patients during DBS therapy.Through the platform, the DBS mechanism including neuron activities and loops in the STN can be explored and further provide the development of close-loop stimulation scheme of DBS. Furthermore, the patient-specific modeling platform could also be extended to other pathological diseases like depression and obsessive compulsive disorders which are also treated via DBS. However, the narrow spectrum of stimulation parameter combinations in this work is the limitation for practical applications. Thus, the efficacy with various stimulation voltage, pulse width and frequency combinations in monopolar and bipolar configurations in the in vivo test are explored in future studies. 3.4 In Vivo Test Results Table 2 lists UPDRS motor scores and dyskinesia side effect scores for each patient during clinical test of DBS. Both patients had no significant UPDRS motor scores difference between monopolar and bipolar stimulation configurations. In addition, UPDRS dyskinesia side effect scores of each patient after stimulation voltage and configuration adjustment were improved. This observation implies that the estimated region of the VTA was maintained, and current diffusion was
limited with bipolar stimulation to improve side effects [14-16]. Meanwhile, the amount of stimulation voltage in a bipolar configuration is higher than that in monopolar stimulation to achieve similar clinical (motor) improvement. That also complied with the simulation results that proposed in this paper. Note that, the monopolar configuration of each patient was with the cathode on electrode contact 1, respectively, and the anode on the implantable pulse generator. Moreover, the bipolar configurations of both patients were with the cathodes on electrode contact 1 and 3, and the anodes on electrode contact 2 and 1, respectively. Although there were only two patients enrolled in the clinical test, the variance of stimulation voltage between monopolar and bipolar stimulation is consistent with our stimulation results and references [4, 9, 14-16]. In future work, a larger group of patients needs to be included to verify and refine our modeling methods. 4. CONCLUSIONS This paper is dedicated to find optimal stimulation voltage in a bipolar configuration with aim to avoid side effects and maintain the same level of improvement in monopolar stimulation. To avoid the side effects during deep brain stimulation (DBS), the stimulation configuration is selected to be bipolar instead of monopolar, which is often used in clinical trials. In this way, the potential of stimulation voltage in the bipolar configuration should be adjusted to maintain the
benefit achieved by monopolar stimulation. To quantify the adjusted amount of stimulation voltage, the volume of tissue activated (VTA) has been first estimated through stimulation voltage using these regression models based on curve fitting. The size difference of VTAs between monopolar and bipolar configuration is then minimized. From the simulation results, bipolar stimulation can avoid side effects and achieve the benefit of monopolar stimulation via adjusting stimulation voltage. Furthermore, the relationship of stimulation voltage between monopolar and bipolar configuration can be formulated as a transformation function integrated into the DBS programming system to determine suitable stimulation parameters for clinical trials. Finally, experiments are conducted with results successfully validating simulation counterparts.
ACKNOWLEDGMENTS The authors would like to thank Dr. Shin-Yuan Chen and Dr. Sheng-Tzung Tsai in the Department of Neurosurgery, Division of Functional Neuroscience, at Tzu-Chi General Hospital for helpful discussions. References [1] N. Yousif, R. Bayford, P. G. Bain, and X. Liu, “The peri-electrode space is a significant element of the electrode-brain interface in deep brain stimulation: A computational study,” Brain Res Bull., 74 (2007) 361-368.
[2] N. Yousif and X. Liu, “Investigating the depth electrode–brain interface in deep brain stimulation using finite element models with graded complexity in structure and solution,” J Neurosci Methods, 184 (2009) 142-151. [3] Wei-Yi Chuang, Paul C.-P. Chao, and Kuu-Young Young, “Computerized stimulation parameter adjustment that avoids side effects and minimizes power consumption for deep brain stimulation in Parkinson's disease,” Journal of Medical and Biological Engineering, doi:10.5405/jmbe.1328, 2013. [4] J. Volkmann, E. Moro, and R. Pahwa, “Basic algorithms for the programming of deep brain stimulation in Parkinson’s disease,” Mov. Disord., 21 (2006) 284-289. [5] C. C. McIntyre, S. Miocinovic, D. L. Sherman, N. V. Thakor, and J. L. Vitek, “Electric field and stimulating influence generated by deep brain stimulation of the subthalamic nucleus,” Clin. Neurophysiol., 115 (2004) 589-595. [6] L. Lopiano, M. Rizzon, B. Bergamasco, A. Tavella, E. Torre, P. Perozzo, M. C. Valentini, and M. Lanotte, “Deep brain stimulation of subthalamic nucleus: clinical effectiveness and safety,” Neurology, 56 (2001) 552-554. [7] C. R. Butson and C. C. McIntyre, “Role of electrode design on the volume of tissue activated during deep brain stimulation,” J. Neural Eng., 3 (2006) 1-8. [8] C. R. Butson and C. C. McIntyre, “Current steering to control the volume of tissue activated during deep brain stimulation,” Brain Stimul., 1 (2008) 7-15.
[9] G. Deli, I. Balas, F. Nagy, E. Balazs, J. Janszky, S. Komoly, and N. Kovacs, “Comparison of the efficacy of unipolar and bipolar electrode configuration during subthalamic deep brain stimulation,” Parkinsonism Relat. Disord., 17 (2011) 50-54. [10] J. Holsheimer, E. A. Dijkstra, H. Demeulemeester and B. Nuttin, “Chronaxie calculated from current-duration and voltage-duration data,” J. Neurosci. Methods, 97 (2000) 45-50. [11] S. Fahn and R. Elton, “The unified parkinson’s disease rating scale,” Recent Developments in Parkinson’s Disease, 2 (1987) 153-163. [12] J. C. Morgan, S. H. Mehta and K. D. Sethi, “Biomarkers in parkinson's disease,” Curr. Neurol. Neurosci. Rep., 10 (2010) 423-430. [13] S. T. Tsai, S. H. Lin, S. Z. Lin, J. Y. Chen, C. W. Lee and S. Y. Chen, “Neuropsychological effects after chronic subthalamic stimulation and the topography of the nucleus in parkinson's disease,” Neurosurgery, 61 (2007) 1024-1029. [14] C. Schmidt and U. Van Rienen, “Modeling the field distribution in deep brain stimulation: the influence of anisotropy of brain tissue,” IEEE Trans. Biomed. Eng., 59 (2012) 1583-1592. [15] M. Sobstyl, M. Ząbek, S. Dzierzęcki, H. Koziara and Z. Mossakowski, “Chronic bilateral pallidal stimulation in patients with generalized primary
dystonia - multi-contact cathodal stimulation is superior to bipolar stimulation mode. Preliminary results,” Neurol Neurochir Pol., 45 (2011) 252-259. [16] T. C. Zhang and W. M. Grill, “Modeling deep brain stimulation: point source approximation versus realistic representation of the electrode,” J. Neural Eng., 7 (2010) doi:10.1088/1741-2560/7/6/066009.
TABLE CAPTIONS
Table 1
Parameters of Laplace’s equation
Table 2
Dyskinesia Side Effect Improvement from trials
Table:
Table 1
Symbol
Quantity
Value [Unit]
σt
Electrical conductivity of the brain tissue
0.15 [S /m ]
σi
The conductivity of the insulation portion of the lead
10-10 [S /m ]
σe
The conductivity of the electrode portion of the lead
4 × 106 [S /m ]
σp
The conductivity of the peri-electrode space
0.2 [S /m ]
Table 2
Patient
Stimulation Voltage
UPDRS
UPDRS
Stimulation Voltage
UPDRS
UPDRS
(Monopolar) (V)
(Motor)
(Dyskinesia)
(Bipolar) (V)
(Motor)
(Dyskinesia)
1
1.2
18
4
2.1
19
2
2
3.0
23
3
3.6
22
1
UPDRS: unified Parkinson disease rating scale
FIGURE CAPTIONS Figure 1. Patient-specific brain model construction: (a) Segmentation of the MRI data. (b) The probe with electrodes (model 3389) with STN marked. R stands for the right side of the patient and L left. (c) 3D brain model.
Figure 2. (a) and (b) show the estimated VTA region based on the BEI model on electrode contact 1 in monopolar and electrode contact 0 and 1 in bipolar stimulation, respectively. (c) The STN (brown) with the electrodes implanted and the VTA (red).
Figure 3. The relationship between the stimulation voltage and the size of VTA in monopolar stimulation.
Figure 4. The relationship between the stimulation voltage and the size of VTA in bipolar stimulation.
Figure 5. The relationship of stimulation voltage in monopolar and bipolar configurations.
Figure 6. The size difference of VTA between monopolar and bipolar stimulation.
Figures:
Figure 1
Figure 2
Semi-major axis (a) Minor axis (2b)
Size of the ellipse (mm)
5
4
3
2
1
0 0
1
2
3
4
5
6
Stimulation voltage (V)
Figure 3
Size of the ellipse (mm)
7
6
5
Semi-major axis (a) Minor axis (2b)
4
3
2
1 0
1
2
3
4
Stimulation voltage (V)
Figure 4
5
6
Stimulation voltage in bipolar mode (V)
5
4
3
2
1
0 0
1
2
3
4
5
Stimulation voltage in monopolar mode (V)
Size difference of the ellipse (mm)
Figure 5
Semi-major axis (a) Minor axis (2b)
4
3
2
1
0
0
1
2
3
4
5
Stimulation voltage in monopolar mode (V)
Figure 6
6
S0924-4247(15)30038-8 http://dx.doi.org/doi:10.1016/j.sna.2015.06.016 SNA 9216
To appear in:
Sensors and Actuators A
Received date: Revised date: Accepted date:
23-5-2014 22-4-2015 17-6-2015
Please cite this article as: Wei-Yi Chuang, Paul C.-P.Chao, Shin-Yuan Chen, ShengTzung Tsai, Kuu-Young Young, Ta-Wei Ting, Computing stimulation voltage in a bipolar electrode configuration to avoid side effects during deep brain stimulation, Sensors and Actuators: A Physical http://dx.doi.org/10.1016/j.sna.2015.06.016 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Title: Computing Stimulation Voltage in a Bipolar Electrode Configuration to Avoid Side Effects during Deep Brain Stimulation
Author list: Wei-Yi Chuang1, Paul C.-P. Chao1*, Shin-Yuan Chen2,3, Sheng-Tzung Tsai2,3, Kuu-Young Young1 and Ta-Wei Ting4
Affiliation of authors: 1
Department of Electrical Engineering, National Chiao-Tung University,
Hsinchu, Taiwan 2
Department of Neurosurgery, Buddhist Tzu Chi General Hospital,
Hualien, Taiwan 3
Department of Medicine, Tzu Chi University, Hualien, Taiwan
4
Department of Biomedical Engineering, Yuanpei University, Hsinchu,
Taiwan
*Corresponding author: Paul C.-P. Chao Tel: +886-3-5131377 Fax: +886-3-5752469 E-mail: [email protected] Postal address: Room No. 735, Engineering Building No. 5, 1001 University Rd., Highlights ► Hsinchu City, 300, Taiwan. Highlights
The methodology of computing stimulation voltage while transformation stimulation configuration from monopolar to bipolar is proposed.
Graphical abstractBiography Wei-Yi Chuang was born in, Miaoli, Taiwan, in 1984. She received the B.S. degree in electrical engineering from Chung Yuan Christian University, Jongli, Taiwan, in 2006, and the M.S. and Ph.D. degrees in electrical control engineering from National Chiao-Tung University, Hsinchu, Taiwan, in 2008 and 2014, respectively. Her research interests focus on modelling, analysis and control of Parkinson’s disease utilizing deep brain stimulation system. Paul C.-P. Chao received M.S. and Ph.D. degree from Michigan State University, USA, respectively. After graduation, he worked for the CAE department of Chrysler Corp in Auburn Hill, Detroit, USA for two years. He is currently a university distiguished professor at National Chiao Tung University (NCTU) and a faculty member of the electrical engineering department. Prof. Chao was the recipient of the 1999 Arch T. Colwell Merit Best Paper Award from Society of Automotive Engineering, Detroit, USA; the 2004 Long-Wen Tsai Best Paper Award from National Society of Machine Theory and Mechanism, Taiwan; the 2005 Best Paper Award from National Society of Engineers, Taiwan; and the 2002/2003/2004 CYCU Innovative Research Award; 2006 the AUO Award; 2007 the Acer Long-Term 2nd-prize Award; The 2007/2008/2009 NCTU EEC Outstanding Research Award; the 2009 Best Paper Award from the Symposium on Nano-Device Technology; the 2010 Best Paper Award from the 20th Annual IEEE/ASME Conference on Information Storage and Processing Systems (ISPS). He is the University Associate VP of NCTU. For editorial services, he is the Technical Editor of IEEE Sensors Journal, and also the Associate Editors of ASME Journal of Vibration and Acoustics, IEEE IoT Journal and Journal of Circuit, System and Computer. In recent years, his research interests focus on dynamic analysis and designs for sensors and actuators; micro-mechatronics and control technology. He is an ASME Fellow and IEEE senior member.
Shin-Yuan Chen received the M.D. and M.S. degrees from Taipei Medical College, Taipei Medical University, Taipei, Taiwan, in 1988 and 1995. He is currently a professor at Tzu Chi Medical College, Tzu Chi University, Hualien, Taiwan. His research interests focus on Parkinson’s disease, functional neuro surgeru and spinesurgery. Sheng-Tzung Tsai received the M.D. degree from Tzu Chi Medical College, Tzu Chi University, Hualien, Taiwan, in 2003. He is currently a clinical fellow form Parkinson Disease Research Center, Division of Functional Neurosurgery, Tzu-Chi General Hospital, Hualien, Taiwan. His research interests focus on Parkinson’s disease, consciousnessdisease, neckandlowbackpain, and spinesurgery. Kuu-young Young was born in Kaohsiung, Taiwan, 1961. He received his B.S. degree in Electrical Engineering from National Taiwan University, Taiwan, in 1983, and M.S. and Ph.D. degrees in Electrical Engineering from Northwestern University, Evanston, IL, U.S.A., in 1987 and 1990, respectively. Between 1983 and 1985, he served as an electronic officer in Taiwan Navy. Since 1990, he has been with the Department of Electrical Engineering at National Chiao Tung University, Hsinchu, Taiwan, where he is currently a Professor. He served as the chairman of the department from 2003 to 2006, and the associate dean of Electrical and Computer Engineering College, NCTU, from 2007 to 2010 and also 2014 to 2015. His research interests include robot compliance control, robot learning control, robot calibration and path planning, teleoperation, robot walking helper, and Science, Technology, and Society (STS). Ta-Wei Ting received the Ph.D. degrees in biomedical engineering from Johns Hopkins University, USA in 2000. He is currently an associate professor at Yuanpei University, Hsinchu, Taiwan. His research interests focus on 3D image reconstruction, bioreactor design and biofluid mechanics. ABSTRACT Deep brain stimulation (DBS) in the subthalamic nucleus (STN) has been applied for advanced Parkinson’s disease (PD). In clinical practices, monopolar
configuration is often utilized based on clinical efficiency. Meanwhile, the volume of tissue activated (VTA) by DBS is estimated for controlling and/or limiting the stimulated region. However, side effects like dyskinesia and tremor were accompanied during stimulation. Most studies had found that bipolar stimulation is an alternative approach for side effects avoidance. The goal of this paper is to develop a quantified relationship of stimulation voltage from monopolar to bipolar configuration to improve neural stimulation. In this paper, an electromagnetic finite element model is first built for a patient-specific physiological brain model, which is established by magnetic resonance imaging (MRI) data. The model is then used for finite element analysis (FEA) to estimate VTA with varied electrode voltage in monopolar and bipolar configurations. With the goals to avoid side effects and achieve symptoms suppression, the stimulation voltage of bipolar configuration is successfully computerized based on the electromagnetic FEA simulations. Experiments are conducted successfully to validate simulation results.
Keywords: Deep brain stimulation (DBS), Volume of tissue activated (VTA), Side effects, Stimulation configuration, Magnetic resonance imaging (MRI), Finite element analysis (FEA). 1. INTRODUCTION Over the last decade, deep brain stimulation (DBS) has been a useful and efficient approach to treat Parkinson’s disease (PD) with implanting electrodes
chronically into a selected brain target, subthalamic nucleus (STN). Having electrode implanted, a brain-electrode interface (BEI), which consists of three main components: implanted electrode, layer of peri-electrode space surrounding the electrode, and surrounding brain tissue, is formed [1, 2]. Because the brain tissue content in the peri-electrode space varies with time, stimulation parameters should also be adjusted during DBS to maintain the efficacy of therapy [3]. In addition, the quantified representation of neuronal responses during stimulation is defined as the volume of tissue activated (VTA). Besides stimulation parameters including the stimulation voltage, the pulse width, and the frequency, stimulation configurations including monopolar and bipolar may be selected based on clinical conditions to achieve optimal therapeutic improvement. In monopolar stimulation, one of the electrode contacts is programmed as cathode and implantable pulse generator (IPG) as anode. Whereas in bipolar settings, two contacts of the electrodes are activated. One is cathode and the other anode. In clinical practices, monopolar stimulation is frequently used because it requires lower stimulation intensity than bipolar stimulation to achieve approximately the same clinical benefit [4]. Meanwhile, the corresponding side effects are invoked when current diffusion covers undesired part of the brain. In this case, changing the activated electrode contact is an intuitive approach to address this issue. However, the suppression of PD symptoms will not achieve the same level as that with the original electrode setting. An alternative approach is to change the
stimulation configuration from monopolar to bipolar stimulation. Some previous studies had found that bipolar stimulation can reduce the occurrence of side effects due to a narrower and more focused current field during stimulation [4, 5]. In this way, the issue becomes how to adjust stimulation voltage when stimulation configuration is changed from monopolar to bipolar. The aim of this work is to quantify the difference between monopolar and bipolar stimulation. An electromagnetic finite element model is first built for a patient-specific physiological brain model, established by magnetic resonance imaging (MRI) data. The model is then utilized for finite element analysis (FEA) to estimate VTA with varied stimulation parameters, like applied electrode voltage. Moreover, the relationship between stimulation voltage and the VTA is developed via regression models. To reduce side effects, stimulation configuration is changed from monopolar to bipolar stimulation in clinical trials. Furthermore, to achieve the approximately same level of suppression of movement symptoms, the stimulation voltage in a bipolar configuration is estimated by the proposed mathematical model based on the electromagnetic FEA simulations. 2. MATERIALS AND METHODS 2.1 Patient Specific Brain Model for Estimating the VTA A BEI finite element model is constructed, referring to [1, 2]. The model consists of three main components: the implanted electrode, the surrounding brain tissue, and the layer of peri-electrode space. Model 3389 DBS lead, which con-
tains four 1.5 mm-in-length and 1.27 mm-in-diameters electrodes, marked as contacts 0, 1, 2, and 3, respectively, is selected for the BEI model simulation. The brain tissue is then modelled as a cylinder surrounding the electrode lead with a radius of 5 cm and a height of 14 cm. And, the layer of peri-electrode space has a thickness of 0.25 mm [2, 3]. Doubling the density of the mesh or doubling the distance of the boundary from the electrode generates a potential distribution that differs by < 2% when compared to the default model. In clinical practice, the range of the stimulation frequency is about 90-170 Hz for optimal motor control with Parkinson’s disease patients [6]. The medium is thus taken as a quasi-static model. The equation for computing the electrical potential distribution of the BEI model can accordingly be reduced to Laplace’s equation: ∇ ⋅ σ∇V = 0
(1)
Where V is the potential (V) and σ the conductivity (S/m), which is a constant for different types of brain tissue. Meanwhile, the Dirichlet boundary condition (V=0) is set on the sides of the brain tissue. The non-active electrode contact abides by a continuity constraint and the insulating surfaces are treated as electrically insulated. In addition, electrical-related parameters are listed in Table 1 and the resulting mesh of the BEI model contains 41,615 nodes for FEA. Smaller elements of the resulting mesh are near the lead and larger elements are at the boundaries of the brain tissue. According to [7], the second difference of the elec-
trical potential distribution in the direction perpendicular to the electrode contact is used to define the neuron activation threshold of the tissue around the electrode. And, the activation threshold E* can be computed as [8] E * = 12.3 × D -0.43
(2)
where D is the normalized distance on monopolar and bipolar configurations: 1 / D = contact 1 / axon 1 + contact 2 / axon 2
(3)
where contacts 1 and 2 are the voltages on electrode contacts 1 and 2, respectively, and axons 1 and 2 are the distance from axon to contacts 1 and 2, respectively. To locate the VTA, a correct patient physiological construction model should be established. The patient physiological construction model can be reconstructed via the prior- and post-operation magnetic resonance imaging (MRI) data by the medical software package, Mimics. The estimated VTA is thus superimposed upon the patient physiological construction model, shown in Fig. 1 [3]. Based on this model, the doctor could accordingly modify the stimulation region within the desired region of STN by selecting optimal electrode contact(s) during clinical practice.
2.2 Regression Model of the VTA The quantified representation of the VTA is essential for controlling the stimulation region. According to [7], the region of the VTA in the 2D spatial contours is similar to an ellipse, which is formulated as
(x − x0 )2 (y − y0 )2 + =1 a2 b2
(4)
where ( x0, y0 ) is the center of the ellipse, and a and b are the semi-major and semi-minor axes, respectively. Variation in the stimulation voltage leads to change in the size of the VTA, consequently the shape of the ellipse. To formulate the relationship between the stimulation voltage and the VTA, we derive linear regression models with the stimulation voltage as an independent variable and a and 2b as response variables based on curve fitting. The response variables are formulated as a = f1 ( xV )
(5)
2b = f 2 ( xV )
(6)
where xV is the stimulation voltage and f1 and f2 are linear regression models of the VTA.
2.3 Optimal Stimulation Voltage in Bipolar Stimulation To reduce side effects, the stimulation configuration is changed from monopolar to bipolar in clinical trials. With bipolar stimulation, the current diffusion is limited to the specific region in STN [9]. Besides that, to achieve the same level of symptom suppression as that in monopolar stimulation, stimulation voltage for the bipolar configuration should be adjusted.
In this study, the direction of axons is assumed to be fixed according to [5, 7, 8]. Because the device of the clinical DBS system is voltage controlled, the current distribution around the activated electrode contact(s) depends on the impedance of the tissue around the activated electrode contact(s). According to [10], the impedance would be influenced by varying pulse widths and frequencies. As this study focused on the stimulation effects with various stimulation voltages, it is assumed that if stimulation regions are the same, the effects of stimulation should also be the same. In other words, to avoid side effects and maintain the benefit of monopolar stimulation, the region and location of the VTA in bipolar stimulation should be similar to that in monopolar. In this study, the electrode contacts of bipolar configuration are selected based on the following rules while stimulation configurations are changed. The cathode of bipolar stimulation is unchanged as that in monopolar stimulation and the anode of bipolar stimulation is selected as an adjacent contact. Based on Eqs. (5) and (6), the shape of the VTA is ellipsoid-like characterized with lengths depending on the stimulation voltage. The VTA can then be quantified as an ellipsoid. To maintain the same benefit as monopolar stimulation, the size difference of the VTAs with two stimulation configurations should be as small as possible. We formulate it as 4 2 2 min π (abipolar × bbipolar − amonopolar × bmonopolar ) V 3
(7)
With the Newton’s method based on optimization theory, the suitable stimulation voltage of bipolar stimulation is determined via finding the minimum value of volume difference. 2.4 Clinical Evaluation Two PD patients undergoing DBS surgery (1 female and 1 male, age 64 and 66) were included. All patients signed informed consents for STN-DBS surgery, the procedures involved in the study and scientific publication of contents and results. The evaluation procedures in the study were carried out with the ethical approval of the institutional review board (Tzu Chi General Hospital, Hualien, Taiwan) and in compliance with the Helsinki Declaration. The inclusion criteria were as follows: (1) good levodopa response on the Unified Parkinson’s Disease Rating Scale (UPDRS) part III (motor) (>30%), (2) drug-related complications (e.g.,
dyskinesia
and
“on-off
phenomenon”),
even
under
optimal
anti-parkinsonian medication adjustment, (3) no structural lesions in the brain MRI, and (4) absence of dementia. The severities of the motor symptoms of PD (UPDRS parts III) were evaluated 1 month prior to surgery and also 3 months after. An acute stimulation test was performed 1 week after operation to identify the optimal contact (among four electrodes) with good motor benefit and avoid neuropsychological effects (such as facial twitching or depression) for chronic stimulation. The test items include speech, facial expression, tremor at rest, action tremor, rigidity, finger taps, hand movements, hand pronation, supination, leg
agility, arising from chair, posture stability, dyskinesia and body bradykinesia [11-13]. Each measurement was on a 5-point scale (range 0-4) with higher scores indicating more severely affected symptoms. The total UPDRS part III (motor) scores of these two patients were 18 and 23 with DBS ON. In this study, we focused on dyskinesia side effect improvement while stimulation configuration was changed from monopolar to bipolar. Having their post-operative status remained stable for at least 6 months, these two patients were included and underwent the experiment with proposed DBS parameters adjustment again. Note that the UPDRS ratings were single-blinded. We retested their UPDRS motor scores and dyskinesia side effect scores after new stimulation voltage and stimulation configuration had worked more than one hour to eliminate the delayed effect of stimulation.
3. RESULTS AND DISCUSSION With the goal of avoiding side effects and controlling the VTA to be within the interested region in the brain to achieve symptoms suppression, we apply the following procedure to determine the optimal stimulation voltage in a bipolar configuration. The first step is to simulate the VTA with monopolar and bipolar stimulation after implanting the DBS lead (Sec. 3.1) based on the BEI model of patient-specific physiological construction. The next step is to construct the regression model which describes the relationship between the stimulation voltages
and the size of VTAs (Sec. 3.2). In the third step, the stimulation configuration is changed from monopolar to bipolar. And, the optimization method is used to find the suitable stimulation voltages in bipolar stimulation which can achieve the same level of symptoms suppression as that in monopolar stimulation (Sec. 3.3). Finally, the in vivo experiment results are given (Sec. 3.4). 3.1 The Estimated VTAs in Monopolar and Bipolar Stimulation The specific brain region is stimulated with the voltage (1.5 V), pulse width (110 µs), and frequency (130 Hz). Note that the stimulation voltage is taken as absolute value in this study. The electrical potential distribution could be generated by the Laplace’s equation. The stimulation voltage of bipolar in this work is selected with the same amount but opposite polarization on two electrode contacts, a useful way in clinical practices. When the ellipse is fitted to the VTA based on Eq. (3), a is determined to be the distance from the electrode contact to the VTA boundary in the direction perpendicular to the electrode contact, and 2b the distance within the VTA boundary in the direction parallel to the electrode contact, as shown in Fig. 2(a) and (b), respectively. For this case of voltage (1.5 V), pulse width (110 µs), and frequency (130 Hz), a and 2b are found to be 2.385 mm and 1.800 mm, respectively, in monopolar stimulation and 1.995 mm and 5.150 mm, respectively, in bipolar. The estimated VTA is then used for the MRI patient-specific physiological model. The estimated VTA via FEA and the surrounding STN are shown in Fig.
2(c). The BEI and patient-specific physiological models are found to be effective in evaluating whether the side effects occurred during the stimulation, thereby providing valuable information for the selection among the electrode contacts. 3.2 Relationship between VTA and Stimulation Voltage To investigate the effects of VTAs due to various stimulation voltages, the stimulation voltage is set in the range of 0.5 to 5.0 V with a step size of 0.25 V, the pulse width at 110 µs, and the frequency at 130 Hz, consistent with typical clinical DBS parameters settings. Figures 3 and 4 show the resultant size of the VTA represented by a and 2b (for the semi-major and minor axes of the ellipse) versus the stimulation voltage in monopolar and bipolar stimulation, respectively. It is found that larger stimulation voltages result in larger VTAs in monopolar stimulation, so do in bipolar. And, a is more sensitive to variations in the stimulation voltage than is 2b in monopolar stimulation. In bipolar stimulation, the sensitivity of a and 2b to variations in the stimulation voltage are similar, as shown in Fig. 4. In other words, the size of the estimated VTA is more focused to variation in the stimulation voltage in bipolar stimulation than that in monopolar. To quantify the relationship between the stimulation voltage and the size of the VTA, curve fitting that minimizes the squared error is used to obtain the linear regression model below: amonopolar = 0.34 + 1.51xV − 0.15xV2
(8)
2bmonopolar = 1.38 + 0.27 xV
(9)
abipolar = 1.61 + 0.211xV
(10)
2bbipolar = 4.80 + 0.19 xV
(11)
where xV is the stimulation voltage and R2 values in fitting the model are 0.98, 0.98, 0.93, and 0.94 for Eqs. (8)-(11), respectively. In clinical trials, the doctor tunes the stimulation parameters in a monopolar configuration based on the clinical efficiency. The size of VTA can then be predicted by stimulation voltage via Eqs. (8) and (9). However, the corresponding side effects occur in monopolar during DBS parameters adjusting. The stimulation configuration is thus changed from monopolar to bipolar. In next step, the optimization method is used for finding the optimal stimulation voltage in bipolar stimulation to maintain the same benefit as that in monopolar stimulation.
3.3 Optimal Stimulation Voltage in a Bipolar Configuration To avoid side effects and maintain the benefit of monopolar stimulation aforementioned, the size difference of VTAs controlled by the stimulation voltage in monopolar and bipolar configurations should be minimized based on Eqs. (7)-(11). When a and 2b in monopolar stimulation are 2.385 mm and 1.800 mm, respectively, the optimal stimulation voltage is found at 1.6 V in bipolar stimulation.
To estimate the relationship of stimulation voltage between monopolar and bipolar stimulation, the optimal stimulation voltages in bipolar are calculated via Eqs. (7)-(11). The relationship of stimulation voltage is shown in Fig. 5. It is found that the amplitude of stimulation voltage in bipolar stimulation is higher (0.3-0.4 V on average) than that in monopolar with the minimal size difference of VTAs. Indeed, to maintain the same level of symptoms suppression in bipolar stimulation, the stimulation voltage is higher than that in monopolar in clinical practices [4, 9]. The size difference of VTAs between monopolar and bipolar stimulation is larger when the potential of stimulation voltage is larger. Figure 6 shows the difference of the distance perpendicular (a) and parallel (2b) to the electrode contact between monopolar and bipolar stimulation vs. the stimulation voltages in monopolar stimulation. In Fig. 6, the difference of distance perpendicular to the electrode contact is over 1 mm when monopolar stimulation voltage is over 3 V. The larger size difference of the VTAs is caused by the higher stimulation potential. The benefit of monopolar stimulation could then not be maintained by bipolar stimulation.
As a fact, the symptoms suppression in bipolar stimulation
is not able to reach as that in monopolar stimulation when stimulation voltage is higher than 3.6 V in clinical trials [9]. According to our primary results, the trend of stimulation voltage adjustment
between monopolar and bipolar configurations can be estimated. Moreover, the platform of modeling the VTA via different stimulation voltages during DBS is constructed. In clinical practices, physicians tune the stimulation parameters depending on the subjective response and the objective examination of patients during DBS therapy.Through the platform, the DBS mechanism including neuron activities and loops in the STN can be explored and further provide the development of close-loop stimulation scheme of DBS. Furthermore, the patient-specific modeling platform could also be extended to other pathological diseases like depression and obsessive compulsive disorders which are also treated via DBS. However, the narrow spectrum of stimulation parameter combinations in this work is the limitation for practical applications. Thus, the efficacy with various stimulation voltage, pulse width and frequency combinations in monopolar and bipolar configurations in the in vivo test are explored in future studies. 3.4 In Vivo Test Results Table 2 lists UPDRS motor scores and dyskinesia side effect scores for each patient during clinical test of DBS. Both patients had no significant UPDRS motor scores difference between monopolar and bipolar stimulation configurations. In addition, UPDRS dyskinesia side effect scores of each patient after stimulation voltage and configuration adjustment were improved. This observation implies that the estimated region of the VTA was maintained, and current diffusion was
limited with bipolar stimulation to improve side effects [14-16]. Meanwhile, the amount of stimulation voltage in a bipolar configuration is higher than that in monopolar stimulation to achieve similar clinical (motor) improvement. That also complied with the simulation results that proposed in this paper. Note that, the monopolar configuration of each patient was with the cathode on electrode contact 1, respectively, and the anode on the implantable pulse generator. Moreover, the bipolar configurations of both patients were with the cathodes on electrode contact 1 and 3, and the anodes on electrode contact 2 and 1, respectively. Although there were only two patients enrolled in the clinical test, the variance of stimulation voltage between monopolar and bipolar stimulation is consistent with our stimulation results and references [4, 9, 14-16]. In future work, a larger group of patients needs to be included to verify and refine our modeling methods. 4. CONCLUSIONS This paper is dedicated to find optimal stimulation voltage in a bipolar configuration with aim to avoid side effects and maintain the same level of improvement in monopolar stimulation. To avoid the side effects during deep brain stimulation (DBS), the stimulation configuration is selected to be bipolar instead of monopolar, which is often used in clinical trials. In this way, the potential of stimulation voltage in the bipolar configuration should be adjusted to maintain the
benefit achieved by monopolar stimulation. To quantify the adjusted amount of stimulation voltage, the volume of tissue activated (VTA) has been first estimated through stimulation voltage using these regression models based on curve fitting. The size difference of VTAs between monopolar and bipolar configuration is then minimized. From the simulation results, bipolar stimulation can avoid side effects and achieve the benefit of monopolar stimulation via adjusting stimulation voltage. Furthermore, the relationship of stimulation voltage between monopolar and bipolar configuration can be formulated as a transformation function integrated into the DBS programming system to determine suitable stimulation parameters for clinical trials. Finally, experiments are conducted with results successfully validating simulation counterparts.
ACKNOWLEDGMENTS The authors would like to thank Dr. Shin-Yuan Chen and Dr. Sheng-Tzung Tsai in the Department of Neurosurgery, Division of Functional Neuroscience, at Tzu-Chi General Hospital for helpful discussions. References [1] N. Yousif, R. Bayford, P. G. Bain, and X. Liu, “The peri-electrode space is a significant element of the electrode-brain interface in deep brain stimulation: A computational study,” Brain Res Bull., 74 (2007) 361-368.
[2] N. Yousif and X. Liu, “Investigating the depth electrode–brain interface in deep brain stimulation using finite element models with graded complexity in structure and solution,” J Neurosci Methods, 184 (2009) 142-151. [3] Wei-Yi Chuang, Paul C.-P. Chao, and Kuu-Young Young, “Computerized stimulation parameter adjustment that avoids side effects and minimizes power consumption for deep brain stimulation in Parkinson's disease,” Journal of Medical and Biological Engineering, doi:10.5405/jmbe.1328, 2013. [4] J. Volkmann, E. Moro, and R. Pahwa, “Basic algorithms for the programming of deep brain stimulation in Parkinson’s disease,” Mov. Disord., 21 (2006) 284-289. [5] C. C. McIntyre, S. Miocinovic, D. L. Sherman, N. V. Thakor, and J. L. Vitek, “Electric field and stimulating influence generated by deep brain stimulation of the subthalamic nucleus,” Clin. Neurophysiol., 115 (2004) 589-595. [6] L. Lopiano, M. Rizzon, B. Bergamasco, A. Tavella, E. Torre, P. Perozzo, M. C. Valentini, and M. Lanotte, “Deep brain stimulation of subthalamic nucleus: clinical effectiveness and safety,” Neurology, 56 (2001) 552-554. [7] C. R. Butson and C. C. McIntyre, “Role of electrode design on the volume of tissue activated during deep brain stimulation,” J. Neural Eng., 3 (2006) 1-8. [8] C. R. Butson and C. C. McIntyre, “Current steering to control the volume of tissue activated during deep brain stimulation,” Brain Stimul., 1 (2008) 7-15.
[9] G. Deli, I. Balas, F. Nagy, E. Balazs, J. Janszky, S. Komoly, and N. Kovacs, “Comparison of the efficacy of unipolar and bipolar electrode configuration during subthalamic deep brain stimulation,” Parkinsonism Relat. Disord., 17 (2011) 50-54. [10] J. Holsheimer, E. A. Dijkstra, H. Demeulemeester and B. Nuttin, “Chronaxie calculated from current-duration and voltage-duration data,” J. Neurosci. Methods, 97 (2000) 45-50. [11] S. Fahn and R. Elton, “The unified parkinson’s disease rating scale,” Recent Developments in Parkinson’s Disease, 2 (1987) 153-163. [12] J. C. Morgan, S. H. Mehta and K. D. Sethi, “Biomarkers in parkinson's disease,” Curr. Neurol. Neurosci. Rep., 10 (2010) 423-430. [13] S. T. Tsai, S. H. Lin, S. Z. Lin, J. Y. Chen, C. W. Lee and S. Y. Chen, “Neuropsychological effects after chronic subthalamic stimulation and the topography of the nucleus in parkinson's disease,” Neurosurgery, 61 (2007) 1024-1029. [14] C. Schmidt and U. Van Rienen, “Modeling the field distribution in deep brain stimulation: the influence of anisotropy of brain tissue,” IEEE Trans. Biomed. Eng., 59 (2012) 1583-1592. [15] M. Sobstyl, M. Ząbek, S. Dzierzęcki, H. Koziara and Z. Mossakowski, “Chronic bilateral pallidal stimulation in patients with generalized primary
dystonia - multi-contact cathodal stimulation is superior to bipolar stimulation mode. Preliminary results,” Neurol Neurochir Pol., 45 (2011) 252-259. [16] T. C. Zhang and W. M. Grill, “Modeling deep brain stimulation: point source approximation versus realistic representation of the electrode,” J. Neural Eng., 7 (2010) doi:10.1088/1741-2560/7/6/066009.
TABLE CAPTIONS
Table 1
Parameters of Laplace’s equation
Table 2
Dyskinesia Side Effect Improvement from trials
Table:
Table 1
Symbol
Quantity
Value [Unit]
σt
Electrical conductivity of the brain tissue
0.15 [S /m ]
σi
The conductivity of the insulation portion of the lead
10-10 [S /m ]
σe
The conductivity of the electrode portion of the lead
4 × 106 [S /m ]
σp
The conductivity of the peri-electrode space
0.2 [S /m ]
Table 2
Patient
Stimulation Voltage
UPDRS
UPDRS
Stimulation Voltage
UPDRS
UPDRS
(Monopolar) (V)
(Motor)
(Dyskinesia)
(Bipolar) (V)
(Motor)
(Dyskinesia)
1
1.2
18
4
2.1
19
2
2
3.0
23
3
3.6
22
1
UPDRS: unified Parkinson disease rating scale
FIGURE CAPTIONS Figure 1. Patient-specific brain model construction: (a) Segmentation of the MRI data. (b) The probe with electrodes (model 3389) with STN marked. R stands for the right side of the patient and L left. (c) 3D brain model.
Figure 2. (a) and (b) show the estimated VTA region based on the BEI model on electrode contact 1 in monopolar and electrode contact 0 and 1 in bipolar stimulation, respectively. (c) The STN (brown) with the electrodes implanted and the VTA (red).
Figure 3. The relationship between the stimulation voltage and the size of VTA in monopolar stimulation.
Figure 4. The relationship between the stimulation voltage and the size of VTA in bipolar stimulation.
Figure 5. The relationship of stimulation voltage in monopolar and bipolar configurations.
Figure 6. The size difference of VTA between monopolar and bipolar stimulation.
Figures:
Figure 1
Figure 2
Semi-major axis (a) Minor axis (2b)
Size of the ellipse (mm)
5
4
3
2
1
0 0
1
2
3
4
5
6
Stimulation voltage (V)
Figure 3
Size of the ellipse (mm)
7
6
5
Semi-major axis (a) Minor axis (2b)
4
3
2
1 0
1
2
3
4
Stimulation voltage (V)
Figure 4
5
6
Stimulation voltage in bipolar mode (V)
5
4
3
2
1
0 0
1
2
3
4
5
Stimulation voltage in monopolar mode (V)
Size difference of the ellipse (mm)
Figure 5
Semi-major axis (a) Minor axis (2b)
4
3
2
1
0
0
1
2
3
4
5
Stimulation voltage in monopolar mode (V)
Figure 6
6