Clinical Study Stereotact Funct Neurosurg 2014;92:365–371 DOI: 10.1159/000366002

Received: February 25, 2014 Accepted after revision: July 14, 2014 Published online: October 29, 2014

Treatment of Chronic Pain: Diffusion Tensor Imaging Identification of the Ventroposterolateral Nucleus Confirmed with Successful Deep Brain Stimulation Ilhami Kovanlikaya a Linda Heier a Michael Kaplitt b Departments of a Radiology and b Neurological Surgery, Weill Cornell Medical College, New York, N.Y., USA

Key Words Thalamic nuclei · Chronic pain · Deep brain stimulation · Diffusion tensor imaging · Fiber tractography

ing. Postoperative DBS electrode placement and the affected cortical areas can be confirmed with coregistration of CT and FT using the electrode as a seed ROI. © 2014 S. Karger AG, Basel

© 2014 S. Karger AG, Basel 1011–6125/14/0926–0365$39.50/0 E-Mail [email protected]


Deep brain stimulation (DBS) is an invasive neurosurgical intervention which is well established in treating movement disorders. It has also been applied to a variety of other conditions such as epilepsy, pain, Tourette’s syndrome, obsessive-compulsive disorders, depression and cluster headaches. Several recent studies have reported successful treatment of chronic pain syndromes with DBS [1, 2]. Treatment of pain syndromes using DBS initially focused on the sensory nuclei of the thalamus. The ventroposterolateral (VPL) and ventroposteromedial nuclei were the most commonly targeted areas for neuropathic pain [3]. Since conventional MRI is not able to identify thalamic nuclei in detail yet, DBS treatment of pain usually relies on indirect targeting based on atlas-derived coordinates. Although indirect targeting of the thalamus is generally robust, variability in anatomy and function of thalamic nuclei across individuals and across disease states would suggest that methods to identify patient-specific targets could enhance the safety and efficacy of DBS surgery [4–6]. Ilhami Kovanlikaya, MD Weill Cornell Medical College 515 East 71st Street, Room S-119 New York, NY 10021 (USA) E-Mail ilk2002 @

Downloaded by: Fudan University Library - 5/16/2015 5:34:51 PM

Abstract Background/Aims: A variety of pain syndromes have been treated successfully with deep brain stimulation (DBS) by targeting the thalamic ventroposterolateral (VPL) nucleus. The purpose of this study was to preoperatively identify the thalamic VPL nucleus by diffusion tensor imaging (DTI) fiber tractography (FT) and confirm it intraoperatively. Methods and Results: FT was performed to identify the thalamic VPL nucleus in 6 healthy volunteers and a patient with intractable chronic pain. The patient had preoperative DTI followed by DBS with the electrode placed by conventional stereotactic methods. Postoperative CT images of the DBS electrode tip were fused with the preoperative DTI and the electrode was noted to be in the position of the VPL nucleus predicted preoperatively by FT. The electrode was then used as a seed region of interest (ROI) to confirm FT back to the somatosensory cortex. Clinical confirmation was also achieved with the patient’s pain relief. In all volunteers, VPL nuclei were identified in similar locations in both thalami, although slight interand intrasubject differences were observed. Conclusion: DTI has the potential to identify the thalamic nuclei in individuals, which would be more accurate than anatomical localization and likely identical to intraoperative physiological test-

Material and Methods Patient Demographics and Medical History Six healthy volunteers, 3 female and 3 male, with an average age of 40 and a 44-year-old female patient were included in this study. The patient was diagnosed with reflex sympathetic dystrophy with


Stereotact Funct Neurosurg 2014;92:365–371 DOI: 10.1159/000366002

intractable burning pain and recurrent skin ulcers over her left shin following a motor vehicle accident that resulted in a left leg compartment syndrome. She had failed medication, multiple spinal cord stimulator placements and intrathecal drug delivery. The DTI sequence was performed on a 3T General Electric scanner (GE Signa HDxt) with the following parameters: TR 10,000 ms, TE 95.8 ms, and matrix 128 × 128, FOV 140 mm, slice thickness 2.4 mm, number of gradient directions 33, spatial resolution 2.5 × 1.8 × 1.8 mm. For the patient, MRI data was acquired a week before the surgery consisting of three-dimensional sagittal spoiled gradient echo, coronal and axial high-resolution SE T2, and DTI sequences. Postoperative CT exam was obtained on a General Electric (GE Light SpeedXtra) with 1.3 mm slice thickness, 512 × 512 matrix, 120 kV tube voltage and 0.9375 spiral pitch factor. DTI Fiber Tractography DTI FT was performed using the deterministic algorithm on the BrainLab (BrainLab, Munich, Germany) workstation. Fractional anisotropy (FA), vector maps, and color-coded maps were generated. Orientation-based color coding, a two-dimensional visualization approach, was used to identify specific white matter tracts. In this approach, image brightness represents diffusion anisotropy, with a red-green-blue color scheme indicating tract orientation (red, revealing fibers with lateral orientation; green, anteroposterior, and blue, craniocaudal). First, the connecting fibers of the thalamus with cortical somatomotor, somatosensory and cerebellar areas were demonstrated by using the whole of the thalamus as a seed ROI for FT. Subcortical white matter of the precentral and postcentral gyrus, plus the red nucleus were used as targeting ROIs for cortical somatomotor, somatosensory and dentato-rubro-thalamic fiber tracts, respectively. Thalamic-somatosensory, thalamic-somatomotor and dentato-rubro-thalamic connecting fibers traversing the thalamus were chosen for guidance to estimate the location of the VPL and the ventral intermedius (ViM) nuclei of the thalamus [11, 12, 16]. In order to better define these thalamic subregions in which low FA values can be seen due to crossing fibers, we reran the analyses with various levels of FA thresholds until the anatomically related cerebral cortical and cerebellar areas were reached (fig. 1, 2). Image Fusion For coregistration of MRI, FT and postoperative CT images, an automatic image fusion module was used on the BrainLab workstation. The automatic image fusion module computes all rigid transformations from one image set into another such that the image sets match [17]. Central parts of the brain such as choroid plexus calcifications and lateral ventricles were taken as anatomical reference points to check image fusion and to fine-tune it if necessary (fig. 3). After fusion of CT and DTI images, the precise location of the electrode and its relation to the VPL and ViM nuclei were also assessed on MRI images. Object Creation The contacts of the electrode were defined on CT images by selecting appropriate contrast on the window levels. The distal two electrode contacts were selected to create an object which was used as a seed ROI for the second set of FT. A second FT was performed to depict the target areas of the brain connected to the particular

Kovanlikaya /Heier /Kaplitt  



Downloaded by: Fudan University Library - 5/16/2015 5:34:51 PM

The noninvasive investigation of the structural connectivity of thalamic nuclei with other nuclei, cerebral and cerebellar cortical areas of the brain has been facilitated by diffusion tensor imaging (DTI) [6]. DTI is a relatively new MRI technique that characterizes the apparent diffusion properties of water in tissue with a high degree of directional organization. The principal diffusion direction corresponds well with the orientation of the major fibers in each voxel of the white matter. Based on this data, fiber tracts can be reconstructed in vivo with DTI tractography providing a visual demonstration of brain structural connectivity [4, 6–9]. Studies in non-human primates have shown that the thalamic ventrobasal complex (the caudal division of the VPL nucleus and the large-celled part of the ventroposteromedial nucleus) project to primary and secondary somatosensory areas (S1 and S2) [10]. The ventral lateral and ventral anterior nuclei project to primary motor cortex (M1) and premotor cortex [10]. In humans, the thalamus is further subdivided into numerous nuclei [10]. Although the nomenclature is not yet consistent, studies revealed similar findings as to which ventral posterior nuclei (VPL, ventroposteromedial) had a strong probability of somatosensory cortical connectivity while ventral lateral, ventral anterior and ventral medial nuclei had motor cortical connectivity [10, 11]. The ventroposterior nuclei group representing the principal thalamic relay for somatosensory information has been a target in multiple stereotactic interventions to treat neurogenic pain [1, 2, 12–15]. Recently, using a probabilistic tractography algorithm, anatomical connectivity has been projected fully into gray matter. Classification of the thalamic gray matter based on cortical connectivity patterns revealed distinct subregions whose locations correspond to the thalamic nuclei described previously in histological studies [4, 8, 11]. In this study, by using DTI fiber tractography (FT) we identified the thalamic VPL nuclei in 6 healthy volunteers and a patient suffering from intractable pain before DBS intervention. Furthermore, the distal electrode contact and its expanded area were used as a seed region of interest (ROI) for FT to demonstrate the connecting fibers between the thalamic target and the cortex.

Fig. 1. Thalamocortical somatosensory connecting fibers passing through the VPL nucleus in 6 volunteers were demonstrated by using deterministic FT. The thalami are colored red and the VPL are the oval yellow regions in the mid-lateral thalami contiguous with the thalamocortical sensory fiber tracts.

Chronic Pain: DTI Identification of the VPL Nucleus

Stereotact Funct Neurosurg 2014;92:365–371 DOI: 10.1159/000366002


Downloaded by: Fudan University Library - 5/16/2015 5:34:51 PM

Fig. 2. Preoperative FT of the patient: thalamocortical sensory fiber tracts (yellow), dentato-rubro-thalamic-cortical fiber tracts (red; red arrow: ViM), and corticothalamic motor fiber tracts (blue) are shown. The distal tip of the DBS electrode (purple, purple arrow) located in the ventrodorsolateral nucleus (yellow area in the right thalamus).

Fig. 3. Image fusion: preoperative MRI and postoperative CT. Choroid plexus calcifications and lateral ventricles were taken as reference points for fine-tuning for image coregistration which are perfectly matched. The purple circle shows the distal end of the electrode (arrow).

contacts of the DBS electrode (fig. 4). In addition, to assess the effects of increasing the electric field of the electrode in the areas around the contacts, we were able to create an expanded ROI from two distal contacts for the third set of FT (fig. 5). An expanded ROI was chosen by arbitrarily increasing it by 2 mm from the borders of the first ROI (distal electrode contacts), which can be tailored depending on the electric power in the electrode contacts. The same FA threshold value (= 2.4) was used for the second and third set of FT in order to compare affected brain areas due to volume changes of ROI alone.

In the third set of FT performed from expanded ROI, similar thalamocortical and thalamocerebellar fiber tracts were depicted but in conjunction with larger areas in the sensory/motor cortices, although the same FA threshold was used (fig. 5). The patient has been pain-free for more than a year since the surgery.

Discussion Findings


Stereotact Funct Neurosurg 2014;92:365–371 DOI: 10.1159/000366002

Kovanlikaya /Heier /Kaplitt  



Downloaded by: Fudan University Library - 5/16/2015 5:34:51 PM

In all volunteers, the VPL nucleus was located in the mid-lateral thalami as expected, although slight interand intrasubject differences were observed in shape and localization of the nuclei (fig. 1). In the patient, the precise location of the DBS electrode which was in the VPL nucleus of the right thalamus was confirmed by using presurgical DTI tractograms fused with CT images after surgery. With the second set of FT, we were able to demonstrate the reverse connection from the two distal electrode contacts on the VPL nucleus to the right somatosensory cortex and the cerebellum (fig. 4).

In this study, we retrospectively identified the thalamic VPL nuclei in 6 healthy volunteers with DTI FT from the sensory cortex. Postoperatively a DBS electrode contact in the VPL nucleus of a patient with intractable pain was used as a seed ROI to track fibers from the thalamus to the cerebral cortex, confirming VPL placement. Preoperative FT from the sensory cortex predicted the VPL location in this patient. DTI FT has been used in DBS before with single case reports and a series of 4 patients have been published. This work plus ours suggests a more proactive role for DTI in intraoperative navigation [18–20]. We obtained similar results indicating that DTI FT may be useful not only to identify the nucleus in order to place

Fig. 4. Second set of FT in which the distal electrode contacts were used as an ROI to reveal connecting fiber tracts between the contact and brain sensory and motor cortical areas and the superior cerebellar peduncle (arrow: central sulcus).

Fig. 5. The third set of DTI FT was created

Chronic Pain: DTI Identification of the VPL Nucleus

Stereotact Funct Neurosurg 2014;92:365–371 DOI: 10.1159/000366002


Downloaded by: Fudan University Library - 5/16/2015 5:34:51 PM

using an expanded ROI of the distal electrode contacts to predict affected connecting fiber tracts and cortical areas. The size of the ROI around the contacts can be tailored regarding the amount of electric power utilized in the electrode (arrow: central sulcus).


Stereotact Funct Neurosurg 2014;92:365–371 DOI: 10.1159/000366002

lective medial thalamotomies in patients with chronic neuropathic pain instead of DBS. Since there is no microelectrode monitoring in HIFU, DTI-FT has great potential to provide guidance in targeting the thalamic subregions [26, 27]. The second set of FT was performed to visualize thalamocortical and cerebellar connecting fibers by using a seed ROI which defines the distal two electrode contacts. This second FT has enabled us to assess brain targets for DBS by showing specific connecting fiber tracts between the distal two contacts of the electrode and the sensory, motor cortical areas and other parts of the brain (fig. 4). The anatomical extension of connecting fiber tracts of the thalamic subregions in our case was parallel to findings reported in studies using DTI to define thalamic connecting fibers [7–9, 11, 16, 21, 22, 25]. In order to evaluate the effects of any particular or combined electrode contacts on targeted brain areas, we were able to create additional seed ROIs from these particular contacts and perform an FT. We propose that the effects of increasing the power of the electric stimulus may also be assessed by enlarging the seeding ROI around the contacts which can be tailored to the size of the anticipated affected area around it. The size of the ROI was increased arbitrarily by 2 mm from the margins in three dimensions for the third set of FT. It revealed similar connecting fiber tracts between the thalamus and cortical areas but in a larger amount as one might expect, although the same FA threshold values were used (fig. 5). The highfrequency stimulation creates not only a functional lesion inside the nucleus but also may activate afferent, efferent and nearby axonal pathways [28]. If any relation can be found, this approach can allow us to select an adequate number of contacts and to adjust the power of the electrode for more effective stimulation. There are several limitations to the present study. The first is the well-known ‘crossing fiber’ problem that can occur with deterministic fiber tracking. This algorithm fails to demonstrate some crossing fiber tracts or may show erroneous fibers due to an incorrect principal diffusion direction. In this respect, the thalamocortical association fiber tracts shown in this study are mostly confirmatory by knowledge-based tractography [9, 28–31]. On the other hand, the deterministic algorithm is still considered advantageous for clinical applications compared to many alternative computationally intensive probabilistic fiber tracking methods [7, 8]. Secondly, regarding fused images created with preoperative MRI and postoperative CT, one must consider that displacement of the brain structures after the implanKovanlikaya /Heier /Kaplitt  



Downloaded by: Fudan University Library - 5/16/2015 5:34:51 PM

the electrodes but also to understand which neuronal network and cortices are affected by DBS [18–20]. Although localization of the ViM and the VPL nuclei in the thalamus is crucial for DBS electrode placement, it relies mainly on atlas-based stereotactic coordinates via CT guidance. Conventional MR imaging is not yet capable of directly visualizing these thalamic nuclei [4, 5, 21]. On the other hand, probabilistic tractography obtained through post-processing of DTI data has been able to successfully segment the thalamus on the basis of its connectivity to various regions of the cerebral cortex [4, 6, 11, 22, 23]. However, even anatomy- and physiology-based studies have shown that difficulties exist in defining the subregions of the thalamus [12]. Additionally, substantial variations between subjects seemed to be somewhat problematic when attempting to use stereotactic atlases or thalamic segmentation based on probabilistic tractography data to guide the electrodes for DBS [16]. Thus, we have used a deterministic post-processing algorithm for FT to define the thalamic subregions after visualizing thalamocortical and dentato-rubro-thalamic pathways passing through the thalamus for each individual subject (fig. 1, 2). The fiber connections and the thalamic subregions, such as the probable VPL and ViM nuclei in the thalamus, were easily demonstrated in the volunteers and the patient (fig.  1, 2). Findings related to localization of the thalamic segmentation were largely similar to those previously found in tracer studies in human and non-human primates [3, 6, 8, 11, 12, 16, 24, 25]. Although the same algorithm and similar FA threshold values were used for FT, slight differences were observed in the images regarding the localization and shape of the VPL nucleus between subjects and even nuclei in the same subject. We think that this observation may be due to the anatomical variability of the thalamic nuclei as pointed out above [4, 5, 16], although it is impossible to rule out some inherent technical problems of FT and MR echo-planar imaging, which might be the cause of this observation. After image fusion of the MRI and postoperative CT images, we were able to define the distinct contacts of the electrode and find their precise location in the brain. The distal contacts in this patient were just located in the area of the right VPL nucleus as demarcated on the preoperative FT (fig. 2). In the future we hope to use these fused images of thalamic FT and CT for image guidance as well, instead of the standard atlas-based coordinates alone. In recent studies, transcranial MR-guided high-intensity focused ultrasound (tcM-RgHIFU) was used for se-

tation of the DBS electrode can affect image coregistration. We believe that these effects are minimal if any because the thalamus is located in the center of the brain and this displacement would not affect FT based on preoperative MRI. We focused mainly on the distal end of the electrode in the thalamus on postoperative CT and have used choroid plexus calcifications and the lateral ventricles for reference points for image coregistration, which in this case matched perfectly on CT and MR images (fig. 2).


Image fusion of FT with CT may be useful not only to define the location of thalamic nuclei preoperatively but to confirm postoperatively the placement of the electrode contacts and their connections to cerebral and cerebellar areas which would be affected by DBS. Larger prospective studies are required before DTI alone can identify targeted thalamic nuclei for DBS, but it already has a supporting role in intraoperative localization reducing the number of electrode passes, complications and decreasing operation time.


Chronic Pain: DTI Identification of the VPL Nucleus



14 15






Boulby PA, Barker GJ, Sillery EL, Sheehan K, Ciccarelli O, Thompson AJ, Brady JM, Matthews PM: Non-invasive mapping of connections between human thalamus and cortex using diffusion imaging. Nat Neurosci 2003; 6: 750–757. Morel A, Magnin M, Jeanmonod D: Multiarchitectonic and stereotactic atlas of the human thalamus. J Comp Neurol 1997;387:588– 630. Nizard J, Raoul S, Nguyen JP, Lefaucheur JP: Invasive stimulation therapies for the treatment of refractory pain. Discov Med 2012;14: 237–246. Levy R, Deer TR, Henderson J: Intracranial neurostimulation for pain control: a review. Pain Physician 2010;13:157–165. Owen SL, Green AL, Nandi DD, Bittar RG, Wang S, Aziz TZ: Deep brain stimulation for neuropathic pain. Acta Neurochir Suppl 2007;97:111–116. Yamada K, Akazawa K, Yuen S, Goto M, Matsushima S, Takahata A, Nakagawa M, Mineura K, Nishimura T: MR imaging of ventral thalamic nuclei. AJNR Am J Neuroradiol 2010;31:732–735. Sure U, Benes L, Riegel T, Schulte DM, Bertalanffy H: Image fusion for skull base neuronavigation. Technical note. Neurol Med Chir (Tokyo) 2002;42:458–461. Owen SLF, Heath J, Kringelbach M, Green AL, Pereira EAC, Jenkinson N, Jegan T, Stein JF, Aziz TZ: Pre-operative DTI and probabilistic tractography in four patients with deep brain stimulation for chronic pain. J Clin Neurosci 2008;15:801–805. Coenen VA, Allert N, Mädler B: A role of diffusion tensor imaging fiber tracking in deep brain stimulation surgery: DBS of the dentato-rubro-thalamic tract (drt) for the treatment of therapy-refractory tremor. Acta Neurochir (Wien) 2011;153:1579–1585. Barkhoudarian G, Klochkov T, Sedrak M, Frew A, Gorgulho A, Behnke E, De Salles A: A role of diffusion tensor imaging in movement disorder surgery. Acta Neurochir (Wien) 2010;152:2089–2095.

21 Stancanello J, Romanelli P, Pantelis E, Sebastiano F, Modugno N: Atlas-based functional radiosurgery: early results. Med Phys 2009;36:457–463. 22 Ziyan U, Westin CF: Joint segmentation of thalamic nuclei from a population of diffusion tensor MR images. Med Image Comput Comput Assist Interv 2008;11:279–286. 23 Ziyan U, Tuch D, Westin CF: Segmentation of thalamic nuclei from DTI using spectral clustering. Med Image Comput Comput Assist Interv 2006;9:807–814. 24 Nandi D, Aziz TZ: Deep brain stimulation in the management of neuropathic pain and multiple sclerosis tremor. J Clin Neurophysiol 2004;21:31–39. 25 Jones EG, Wise SP, Coulter JD: Differential thalamic relationships of sensory-motor and parietal cortical fields in monkeys. J Comp Neurol 1979;183:833–881. 26 Martin E, Jeanmonod D, Morel A, Zadicario E, Werner B: High-intensity focused ultrasound for noninvasive functional neurosurgery. Ann Neurol 2009;66:858–861. 27 Jeanmonod D, Werner B, Morel A, Michels L, Zadicario E, Schiff G, Martin E: Transcranial magnetic resonance imaging-guided focused ultrasound: noninvasive central lateral thalamotomy for chronic neuropathic pain. Neurosurg Focus 2012;32:E1. 28 Coenen VA, Schlaepfer TE, Allert N, Mädler B: Diffusion tensor imaging and neuromodulation: DTI as key technology for deep brain stimulation. Int Rev Neurobiol 2012;107:207–334. 29 Hua K, Oishi K, Zhang J, Wakana S, Yoshioka T, Zhang W, Akhter KD, Li X, Huang H, Jiang H, van Zijl P, Mori S: Mapping of functional areas in the human cortex based on connectivity through association fibers. Cereb Cortex 2009;19:1889–1895. 30 Wakana S, Nagae-Poetscher LM, Jiang H, van Zijl P, Golay X, Mori S: Macroscopic orientation component analysis of brain white matter and thalamus based on diffusion tensor imaging. Magn Reson Med 2005;53:649–657. 31 Mori S, van Zijl PC: Fiber tracking: principles and strategies – a technical review. NMR Biomed 2002;15:468–480.

Stereotact Funct Neurosurg 2014;92:365–371 DOI: 10.1159/000366002


Downloaded by: Fudan University Library - 5/16/2015 5:34:51 PM

1 Boccard SG, Pereira EA, Moir L, Aziz TZ, Green AL: Long-term outcomes of deep brain stimulation for neuropathic pain. Neurosurgery 2013;72:221–231. 2 Lyons MK: Deep brain stimulation: current and future clinical applications. Mayo Clin Proc 2011;86:662–672. 3 Turnbull IM, Shulman R, Woodhurst WB: Thalamic stimulation for neuropathic pain. J Neurosurg 1980;52:486–493. 4 Pouratian N, Zheng Z, Bari AA, Behnke E, Elias WJ, Desalles AA: Multi-institutional evaluation of deep brain stimulation targeting using probabilistic connectivity-based thalamic segmentation. J Neurosurg 2011; 115: 995–1004. 5 Traynor CR, Barker GJ, Crum WR, Williams SC, Richardson MP: Segmentation of the thalamus in MRI based on T1 and T2. Neuroimage 2011;56:939–950. 6 Zhang D, Snyder AZ, Shimony JS, Fox MD, Raichle ME: Noninvasive functional and structural connectivity mapping of the human thalamocortical system. Cereb Cortex 2010;20:1187–1194. 7 Wang JY, Abdi H, Bakhadirov K, Diaz-Arrastia R, Devous MD Sr: A comprehensive reliability assessment of quantitative diffusion tensor tractography. Neuroimage 2012; 60: 1127–1138. 8 Behrens TE, Woolrich MW, Jenkinson M, Johansen-Berg H, Nunes RG, Clare S, Matthews PM, Brady JM, Smith SM: Characterization and propagation of uncertainty in diffusionweighted MR imaging. Magn Reson Med 2003;50:1077–1088. 9 Mori S, Crain BJ, Chacko VP, van Zijl PC: Three-dimensional tracking of axonal projections in the brain by magnetic resonance imaging. Ann Neurol 1999;45:265–269. 10 Bosch-Bouju C, Hyland BI, Parr-Brownlie LC: Motor thalamus integration of cortical, cerebellar and basal ganglia information: implications for normal and parkinsonian conditions. Front Comput Neurosci 2013;7:163. 11 Behrens TE, Johansen-Berg H, Woolrich MW, Smith SM, Wheeler-Kingshott CA,

Treatment of chronic pain: diffusion tensor imaging identification of the ventroposterolateral nucleus confirmed with successful deep brain stimulation.

A variety of pain syndromes have been treated successfully with deep brain stimulation (DBS) by targeting the thalamic ventroposterolateral (VPL) nucl...
2MB Sizes 0 Downloads 9 Views