Background: Deep brain stimulation has become an established technology for the treatment of patients with a wide variety of conditions, including movement disorders, psychiatric disorders, epilepsy, and pain. Surgery for implantation of DBS devices has enhanced our understanding of human physiology, which in turn has led to advances in DBS technology. Our group has previously published on these advances, proposed future developments, and examined evolving indications for DBS. Summary: The crucial roles of structural MR imaging pre-, intra-, and post-DBS procedure in target visualization and confirmation of targeting are described, with discussion of new MR sequences and higher field strength MRI enabling direct visualization of brain targets. The incorporation of functional and connectivity imaging in procedural workup and their contribution to anatomical modelling is reviewed. Various tools for targeting and implanting electrodes, including frame-based, frameless, and robot-assisted, are surveyed, and their pros and cons are described. Updates on brain atlases and various software used for planning target coordinates and trajectories are presented. The pros and cons of asleep versus awake surgery are discussed. The role and value of microelectrode recording and local field potentials are described, as well as the role of intraoperative stimulation. Technical aspects of novel electrode designs and implantable pulse generators are presented and compared.

Key Messages

DBS as a therapeutic procedure is rapidly expanding, and the techniques of DBS system implantation have been greatly improved by developments in stereotactic technology, surgical techniques, and imaging. Innovations in DBS technology and surgery promise to make DBS surgery more accurate and effective, with potentially improved outcomes for a wider spectrum of patients, provided that technological advancements do not become so complicated and cumbersome that their dissemination and widespread use are hindered. Additional multidisciplinary work is needed, aiming at making these technological advances more user-friendly, which should result in further patient benefits.

Deep brain stimulation (DBS) is an established technology that enables direct intervention on neural circuits by implantation of electrodes in specific intracranial targets followed by local neuromodulation. The ability to directly modulate regions though a minimally invasive procedure has generated proven benefits for patients with a wide variety of neurological disorders. Over 200,000 patients have had a DBS system implanted for such conditions as epilepsy, Parkinson’s disease, dystonia, and essential tremor, with ongoing investigations examining its effectiveness in obsessive-compulsive disorder, depression, and more [1‒4]. Hence, DBS has a utility profile that is rapidly expanding as we continue to gain understanding of the neuromodulation of neural circuitry.

The modern era of DBS has brought substantial growth and technical innovations that have improved the technique and delivery of stimulation. The authors previously have addressed current and emerging indications for DBS, and the need for specific improvements in the technology of implanted systems [5, 6]. The purpose of this review was to evaluate the current state of DBS implantation techniques and to define future modifications that will increase the safety, efficacy, and ease – for patients and surgeons – of these procedures. We begin by examining developments in preoperative imaging, the array of novel stereotactic techniques, the emergence of various atlases and their functionality, and planning software that has streamlined the surgical process. Next, we examine perioperative developments, including the emergence of asleep surgery (and compare it to awake surgery), the implementation of microelectrode recordings (MERs) to better understand neural circuitry, types of electrodes and their function in stimulation, implantable pulse generators, and the emergence of robotics within the field. Last, we examine postoperative imaging as a confirmatory mechanism after implantation.

The parallel rise of novel brain imaging, new targeting techniques, improved electrode and stereotactic frame hardware, the rise of robotics, and enhanced atlas guides have enhanced technique, but may also come at a cost. This review serves to summarize these developments, examine them through a critical lens, and suggest further necessary refinements.

Preoperative Imaging

From the pioneering work of Spiegel [7] and Tasker [8] with targeting based on pneumoencephalography and Talairach with the implementation of anterior (AC) and posterior commissure (PC) lines in stereotaxis [9] to the modern use of computed tomography (CT) and magnetic resonance imaging (MRI) in directly visualizing brain structures, neuroimaging has played a central role in guiding stereotactic surgeries. Although optimized MRI sequences may enable sufficient visualization of certain DBS structures for direct targeting, for example, the globus pallidus internus (GPi) [10] and the subthalamic nucleus (STN) [11], clinical MRI sequences remain a suboptimal means of pure direct visualization of, for example, the ventral intermediate nucleus (Vim) for surgical planning [12]. Thus, indirect targeting using coordinates relative to a given anatomical landmark is still used. Disadvantages of this method include failure to account for interindividual variability in the location of target structures and inter-surgeon variability when identifying these landmarks [13, 14]. Accurate targeting is crucial since the therapeutic effects achieved with DBS are predicated on selective stimulation of the intended structure through accurate and precise placement of the electrodes [15, 16]. Here, we review neuroimaging advancements that have enabled improved direct targeting of the most common grey matter nuclei targeted with DBS.

Optimization of commonly used MRI sequences such as T1-weighted (T1w) imaging can improve visualization of DBS targets. For example, the inversion time of T1w sequences may be optimized, allowing suppression of gray matter and enabling identification of the main thalamic groups [17]. Other MRI sequences such as inversion recovery (IR) have also improved visualization of DBS targets. IR sequences aim to enhance the visualization of a given structure by selectively suppressing certain tissues with a specific composition. One such sequence, fast grey matter T1 inversion recovery (FGATIR), nulls white matter signal and generates improved visualization of the GPi with delineation of the internal medullary lamina [18]. FGATIR has been further modified to enhance the distinction between GPi and the globus pallidus externus by suppressing fluid in addition to white matter (i.e., fluid and white matter suppression sequence) (Fig. 1a) [19]. This is accomplished through the registration of two contrasts: the standard T1w anatomical contrast of the brain (i.e., magnetization-prepared rapid gradient-echo sequence) and the images with suppression of the white matter signal (i.e., FGATIR). Another IR sequence that suppresses signal from white matter (i.e., white matter attenuated inversion recovery) has demonstrated promise in visualizing the internal subdivisions of the thalamus due to more difference in contrast between different gray matter territories in PD and ET patients (Fig. 1b) [20].

Fig. 1.

Optimized 3T MRI images for visualizing DBS targets. a Zoomed out and zoomed in MRI of the globus pallidus internus (axial 3T FGATIR image); (b) thalamus (coronal 3T WAIR image sequence); and (c) subthalamic nucleus (coronal 3T QSM image) shown in addition to the corresponding slice from the Mai et al. atlas. Insets depict magnified target structures. Atlas slices were reprinted by permission from Academic Press, Mai JK, Majtanik M, Paxinos G, Atlas of the Human Brain (4th Edition), 239 & 406, Copyright Elsevier (2015). b Was reprinted by permission from: [20], with permission from Elsevier. c Was reprinted by permission from [21].

Fig. 1.

Optimized 3T MRI images for visualizing DBS targets. a Zoomed out and zoomed in MRI of the globus pallidus internus (axial 3T FGATIR image); (b) thalamus (coronal 3T WAIR image sequence); and (c) subthalamic nucleus (coronal 3T QSM image) shown in addition to the corresponding slice from the Mai et al. atlas. Insets depict magnified target structures. Atlas slices were reprinted by permission from Academic Press, Mai JK, Majtanik M, Paxinos G, Atlas of the Human Brain (4th Edition), 239 & 406, Copyright Elsevier (2015). b Was reprinted by permission from: [20], with permission from Elsevier. c Was reprinted by permission from [21].

Close modal

T2*-weighted sequences have also improved visualization of DBS targets. Sequences exploiting the T2* effect enhance the magnetic susceptibility differences between tissues. These sequences are particularly sensitive to iron content. T2* images have been successfully used to visualize all STN boundaries and to discern GPi and globus pallidus externus [22, 23]. Quantitative susceptibility maps can be computed with post-processing of T2* images, which allows for quantification and correction of geometrical distortions inherent to T2*-weighted imaging [12, 21]. Quantitative susceptibility map has shown promise for visualizing the STN and, to a lesser extent, other structures such as the GPi (Fig. 1c) [21, 22, 24].

Despite encouraging results, the adoption of these sequences in clinical settings remains relatively low to date, perhaps due to the specialized knowledge base required, single vendor implementation, high costs, and the need for replication of relevant findings in larger studies [12, 25]. Second, demonstrable improvements in image quality and the use of novel sequences are limited by the persistent requirement of using specific head coils that may not be available on all MRI scanners and may not physically accommodate stereotactic head frames. Finally, geometrical distortions associated with most of these optimized sequences remain to be quantified [12, 26, 27].

Further improvement may be expected as 7T MRI becomes more widely available. Higher magnetic field strengths offer increased signal-to-noise ratio, which in turn allows increased spatial resolution, permitting the delineation of smaller neuroanatomical structures (Fig. 2) [28, 29]. Given these advantages, it is not surprising that 7T MRI has been shown to be superior to 1.5T and 3T MRI when visualizing DBS targets and with comparable acquisition times [30, 31]. However, these images are also more prone to susceptibility effects and image distortions, potentially leading to a greater risk of mistargeting [26, 32]. To this end, the creation of 7T MRI image correction algorithms is required for better visualization of target structures, which is needed for clinical implementation of this technique.

Fig. 2.

Optimized 7T MRI images for visualizing DBS targets and neuroanatomy. a Left, atlas image of the thalamus; right, zoomed out and (inset) zoomed in 7T white matter nulled FGATIR MRI of the thalamus. b 7T SWI (left) depicting venous system and neuroanatomic details such as Edinger’s comb system not visible in similar detail at lower field strengths such as T2* sequences (right). Atlas slices were reprinted by permission from Academic Press, Mai JK, Majtanik M, Paxinos G, Atlas of the Human Brain (4th Edition), 239 & 406, Copyright Elsevier (2015).

Fig. 2.

Optimized 7T MRI images for visualizing DBS targets and neuroanatomy. a Left, atlas image of the thalamus; right, zoomed out and (inset) zoomed in 7T white matter nulled FGATIR MRI of the thalamus. b 7T SWI (left) depicting venous system and neuroanatomic details such as Edinger’s comb system not visible in similar detail at lower field strengths such as T2* sequences (right). Atlas slices were reprinted by permission from Academic Press, Mai JK, Majtanik M, Paxinos G, Atlas of the Human Brain (4th Edition), 239 & 406, Copyright Elsevier (2015).

Close modal

Stereotactic Frames

Lars Leksell transformed stereotactic neurosurgery in 1949 with the development of a frame relying on centripetal targeting instead of rectilinear adjustments in a Cartesian coordinate system. This significant innovation dramatically improved the efficacy and flexibility of the stereotactic frame [7, 33, 34]. It is remarkable that Leksell’s frame, with minimal adaptations, has remained a reliable standard – indeed, the most widely used platform in intracranial stereotaxis for over 70 years. In the past several decades, newer technologies, such as 3D printing, have enabled production of platforms with comparable accuracy and flexibility. Here, we briefly describe these novel platforms and compare their accuracy and advantages to the “traditional” frame standard.

The STarFix System

The STarFix microTargeting Platform (FHC, Bowdoin, ME, USA) (shown in Fig. 3a, b) is a customized miniframe that attaches rigidly to previously placed bone fiducial screws. This platform was enabled by early additive manufacture (3D printing) technology that allows for precise customization and the possibility of multiple simultaneous trajectories incorporated into the same frame. Konrad et al. [35] measured a targeting error for this platform of 1.24 ± 0.4 mm in DBS cases when brain shift was minimal [36]. These numbers compare favorably with even the best reported accuracy for frame-based techniques, which was a vector deviation of 1.2 ± 0.6 mm reported by Bjartmarz and Rehncrona for the Leksell G-frame [37]. The achievement of accuracy comparable to traditional frames has made use of custom-made platforms a viable option for surgeries requiring optimal accuracy. The multiple trajectories that can be achieved provide the opportunity of time savings in bilateral DBS procedures.

Fig. 3.

StarFix Apparatus. a Computer visualization of a dual-trajectory StarFix platform for bilateral DBS. b Visualization of a multi-trajectory STarFix platform. c Microtable with mounting hardware for STarFix, denoted by white arrow.

Fig. 3.

StarFix Apparatus. a Computer visualization of a dual-trajectory StarFix platform for bilateral DBS. b Visualization of a multi-trajectory STarFix platform. c Microtable with mounting hardware for STarFix, denoted by white arrow.

Close modal

A variation of this platform technology is known as the Microtable (shown in Fig. 3c), which is cut from a flat piece of polycarbonate and attached to four legs of varying lengths to give the desired trajectory. The Microtable can be produced rapidly (the platform is created within 15 min). Ball et al. [38] showed a Euclidean error of 0.97 ± 0.37 mm for the Microtable, which was reduced to 0.75 ± 0.17 mm for patients who had 1-month delayed CT scans with no pneumocephalus.

While high accuracy and precision are fundamental requirements of stereotactic devices, there are also clinical considerations. Patient comfort is a significant factor in choosing a stereotactic system, and studies have reported on the discomfort of frame-based procedures [39]. Longer surgeries, like awake bilateral DBS with MER, can be quite difficult for patients in a frame, especially when they have a significant axial tremor. The StarFix and Microtable platforms do not require the patient to be rigidly affixed to the operating table. Additionally, since the fiducial markers are placed in a previous visit, operations typically proceed earlier on the day of surgery without the delay of CT or MRI [40, 41]. This is critical for Parkinson’s patients who will have discontinued medication before surgery, as surgery can be conducted sooner and without the encumbrances of the frame.

The NexFrame System

The Medtronic NexFrame stereotactic system (NexFrame, Medtronic, Minneapolis, MN, USA) is a skull-mounted device for DBS lead placement. It is used in conjunction with bone-anchored fiducials to offer an alternative to frame-based stereotaxy [42]. The bone fiducials provide a rigid base for paired-point registration (Fig. 4). Compared to other registration methods used for frameless targeting (i.e., adhesive fiducials or surface registration), registration with bone fiducials produces the registration accuracy required for DBS. A minimum of 5 bone fiducials are required to register the NexFrame for surgical navigation.

Fig. 4.

Paired-point using the bone-anchored fiducials. The tip of the image-guided probe (white) is placed in the divot of one of the five bone-anchored fiducials. The optical camera of the navigation system (not pictured) triangulates the relative position of the image-guided probe to the passive reference frame (blue).

Fig. 4.

Paired-point using the bone-anchored fiducials. The tip of the image-guided probe (white) is placed in the divot of one of the five bone-anchored fiducials. The optical camera of the navigation system (not pictured) triangulates the relative position of the image-guided probe to the passive reference frame (blue).

Close modal

The NexFrame consists of 3 components: an image-guided probe, a passive reference frame, and a stereotactic tower (Fig. 5). The tower is comprised of a ring assembly, which is mounted to the skull at the site of the burr hole; a socket assembly, which rotates on the ring and contains a sweep mechanism; and a reference frame bracket assembly, which secures the passive reference frame to the ring assembly. Alignment with the target is performed by rotating and sweeping the socket until the trajectory intersects with the target. Once the trajectory is locked in, the navigation software provides the distance to target and the DBS lead is placed.

Fig. 5.

The NexFrame tower. On the right, the ring assembly is mounted to the skull and the reference frame bracket assembly connects the passive reference frame to the ring assembly; there is no socket assembly. On the left, the socket assembly has been placed on the ring assembly, and the image-guided probe is secured to the socket assembly. The socket assembly is able to rotate on the ring assembly. The socket sweep is illustrated with the red arrow.

Fig. 5.

The NexFrame tower. On the right, the ring assembly is mounted to the skull and the reference frame bracket assembly connects the passive reference frame to the ring assembly; there is no socket assembly. On the left, the socket assembly has been placed on the ring assembly, and the image-guided probe is secured to the socket assembly. The socket assembly is able to rotate on the ring assembly. The socket sweep is illustrated with the red arrow.

Close modal

The SmartFrame System

Many patients who are good candidates for DBS, such as children or adults with significant anxiety or severe involuntary movements, may not be able to tolerate awake surgery. An interventional MRI (iMRI)-guided procedure that allows for real-time anatomical imaging, with the goal of achieving very accurate lead placement in patients who are under general anesthesia, is one possible solution for these patients.

The procedure is performed within the isocenter of a high-field diagnostic magnet, often in a radiology suite rather than in an operating room. A disposable skull-mounted aiming device is used instead of a stereotactic frame. Initially, this was done using an aiming device with two degrees of freedom [43], not specifically designed for iMRI applications. Based on this experience, a second-generation device was developed to improve ease of use and accuracy of targeting; this device included improved mechanical controls and an integrated software package (SmartFrame, ClearPoint Neuro, Solana Beach, CA) [44, 45]. The SmartFrame (Fig. 6) has four degrees of freedom: “pitch” and “roll” controls for performing an initial rapid approximate alignment in conjunction with oblique axial imaging orthogonal to the alignment stem of the device, and finer X and Y controls used in conjunction with oblique coronal and sagittal imaging through the long axis of the device (Fig. 6) for fine adjustment of the final aim.

Fig. 6.

SmartFrame mounting device apparatus and use. a The SmartFrame skull mounted aiming device, reprinted with permission from [44]. b Oblique coronal and sagittal images aligned with the long axis of the alignment stem to adjust or confirm the final aim.

Fig. 6.

SmartFrame mounting device apparatus and use. a The SmartFrame skull mounted aiming device, reprinted with permission from [44]. b Oblique coronal and sagittal images aligned with the long axis of the alignment stem to adjust or confirm the final aim.

Close modal

The radial error of lead placement (deviation of the lead trajectory from the intended trajectory in the axial plane) averaged 0.6 mm [46]. In nonrandomized comparisons, the clinical outcomes of iMRI-guided DBS for PD and dystonia were similar to those published for traditional awake, microelectrode-guided techniques [44, 47]. Over 95% of leads were placed with a single brain penetration. A recently introduced closed-dura technique avoids entrainment of intracranial air. While conceptually simple, the technique is critically dependent on direct visualization of the anatomic target on high-resolution MR images.

Stereotactic Atlases

Human stereotactic atlases allow surgeons to determine Cartesian coordinates for targeting structures in the thalamus or the basal ganglia by referring to anatomic landmarks shown on myelin-stained thin slice brain sections. Since their introduction in clinical practice in the early 1950s, these atlases have had a major impact on the practice of “indirect targeting” mostly by referring to structures in the third ventricle, especially the intercommissural line that “connects” the AC and PC. The x, y, and z coordinates for the anatomic target structure can be derived from overlays on the anatomic specimens represented in three orthogonal planes (axial, coronal, and sagittal) which intersect at one point, commonly defined as zero.

For a long period of time, printed atlases were the main source of target coordinates and neighboring anatomical structures [48]. Soon after publication of the first printed stereotactic brain atlas by Spiegel and Wycis in 1952, other atlases appeared, such as the Schaltenbrand and Bailey atlas in 1959, and its second edition, the Schaltenbrand and Wahren atlas in 1977. The latter contained an “electroanatomical atlas” alongside pure morphology [49, 50]. These impressive, folio-sized, expensive volumes became the “liturgical” books of the stereotactic OR. In particular, when exploring “new” targets or determining trajectories, these atlases became indispensable. Plate 47 from the Schaltenbrand and Bailey atlas became an icon for targeting the nucleus ventralis intermedius for treatment of tremor for many years.

The last two decades have brought tremendous progress in stereotactic targeting, with new atlases becoming available on the internet, concentrating on different aspects of anatomy and physiology [51]. New formats include printed atlases accompanied by digital media, purely electronic atlases, software installed on commercially available workstations, and internet-based tools available as free shareware or on a pay-per-use basis [51, 52]. Additional views can be derived directly from histology or MRI [53, 54]. In contrast to the classical print formats, these new platforms also provide pseudo or even true 3D space, and allow for segmenting, scaling, and morphing of overlay atlases according to the individual’s anatomy. An example of segmentation of the thalamus for identification of the motor subregion is depicted in Figure 7.

Fig. 7.

Registration of atlases to imaging. a FGATIR MRI scan depicted in native patient space. An equidistant grid is overlaid. b The scan is nonlinearly co-registered to Montreal Neurological (2009b) space using default parameters in Lead-DBS [55]. Note the nonlinear distortion introduced to the image as indicated by the skewed grid lines. c After registration, the human motor thalamus [56] is overlaid with the normalized image, now showing spatial agreement. Such registrations work in both directions (and are hence termed “diffeomorphic”). Once a solution is found, it can be applied to the image (porting it to template space) or to the atlas (porting it to native subject space).

Fig. 7.

Registration of atlases to imaging. a FGATIR MRI scan depicted in native patient space. An equidistant grid is overlaid. b The scan is nonlinearly co-registered to Montreal Neurological (2009b) space using default parameters in Lead-DBS [55]. Note the nonlinear distortion introduced to the image as indicated by the skewed grid lines. c After registration, the human motor thalamus [56] is overlaid with the normalized image, now showing spatial agreement. Such registrations work in both directions (and are hence termed “diffeomorphic”). Once a solution is found, it can be applied to the image (porting it to template space) or to the atlas (porting it to native subject space).

Close modal

The contemporary stereotactic atlases provide many features beyond morphology, with data on fiber tracts or on vessels and enhanced information on function and connectivity [51, 57, 58]. Other options include embedded probabilistic data, fitting to electrophysiological data, and hotspot maps for obtaining better clinical results for DBS electrode contact positioning [59]. More recently, age-dependent and ethnicity-specific characteristics have been taken into consideration [60, 61]. All of this information may be relevant when positioning segmented DBS electrodes.

Despite these advances in atlases, direct targeting has become the preferred method in most centers worldwide due to developments in neuroimaging. These atlases do not account for patient-specific anatomy and contain inherent margins of error due to the process of nonlinear image co-registration and atlas overlay [62‒65]. New methodologies and the development of advanced neuroimaging may diminish these errors. Implementation of atlases may still occur alongside direct targeting, especially for approaching targets which are difficult to visualize (such as the thalamic Vim), for target coordinate determination of image-derived targets, and for confirmation of DBS electrode contact location coordinates. It is expected that the different types of stereotactic atlases considering both form and function will become more widely available through the internet and that more integrative multimodal modular sets will be developed.

Planning Software

DBS planning software integrates preoperative imaging, stereotactic atlases, and a fiducial system in a user-friendly and intuitive format that optimizes direct and indirect targeting of electrodes. While the fundamentals of the various commonly used planning software platforms are similar, there are important distinctions, as summarized in Table 1, which compares two commonly used and commercially available systems.

Table 1.

Comparison of preoperative planning platforms

 Comparison of preoperative planning platforms
 Comparison of preoperative planning platforms

Most planning begins with obtaining and fusing imaging sequences. After selecting an MRI field strength (including 7T, if possible) and the ideal sequence for the target (e.g., T1, T2, FGATIR), image fusion is performed. While T1w thin-cut “stereotactic” axial imaging fused to other sequences allows for targeting of the STN or GPi, it is common to use CT as a reference image to confirm spatial accuracy.

Direct targeting is based on patient-specific MRI. Comparisons have been made between iMRI and MER for accuracy of lead placement and clinical outcomes, demonstrating comparable safety profiles [66, 67]. However, these studies have found varying conclusions regarding clinical outcomes, with one study showing superior clinical outcomes in iMRI [66], and another study demonstrating no significant difference in clinical outcomes between iMRI and MER [67].

Ideally, the planning software of the future will be intuitive to use and largely automated, auto-populating relevant imaging studies through secure, cloud-based services. Images will be overlaid with 3D anatomic and functional atlases. Optimization algorithms will be incorporated to target, based on intended lead, and safe trajectories will be automatically generated [68, 69]. MER recordings will automatically and wirelessly be communicated to the planning station and will be mapped along the anatomic trajectory. Automated algorithms will give real-time feedback to the surgeon in terms of changes to optimal target or trajectory safety. If accepted, new coordinates or trajectories will be wirelessly communicated to the frame or robot, the latter of which would automatically make the needed adjustments. None of these novel advances will obviate the need for neurosurgeons to be intimately familiar with neuroanatomy and to be able to proceed with the surgery in the inevitable event of complex system failure.

“Asleep” versus “Awake” DBS Surgery

The premise behind “asleep” DBS is that accurate anatomical placement of leads can produce equivalent functional outcomes as lead placement guided by the traditional method of awake testing. The anatomical target is visualized on MRI and the accuracy of lead placement is assessed using intraoperative imaging (Fig. 8a). The consideration of asleep DBS resulted in part from advances in MRI, increased availability of intraoperative 3D imaging modalities (e.g., due to trends in spine and brain tumor surgery; Fig. 8b), and decades of clinical data on the efficacy of DBS surgery [70].

Fig. 8.

Intraoperative imaging to determine proper lead placement. a Use of intraoperative CT to image lead position intraoperatively prior to closure. In this figure, imaging is being obtained prior to removal of the stylet from the DBS lead. b Intraoperative CT merged with preoperative MRI indicates the location of contact 2 (Medtronic 3,387 electrode array) within the globus pallidus interna, the latter visualized using a fast gray matter acquisition T1 inversion recovery (FGATIR) sequence. Stereotactic coordinates of the contact on the left are (−18.6, 5.1, 0.5), and the radial error from the intended trajectory is 1.5 mm.

Fig. 8.

Intraoperative imaging to determine proper lead placement. a Use of intraoperative CT to image lead position intraoperatively prior to closure. In this figure, imaging is being obtained prior to removal of the stylet from the DBS lead. b Intraoperative CT merged with preoperative MRI indicates the location of contact 2 (Medtronic 3,387 electrode array) within the globus pallidus interna, the latter visualized using a fast gray matter acquisition T1 inversion recovery (FGATIR) sequence. Stereotactic coordinates of the contact on the left are (−18.6, 5.1, 0.5), and the radial error from the intended trajectory is 1.5 mm.

Close modal

A brief review of the history of DBS surgery and the concomitant advances in intraoperative technologies informs an understanding of how “awake” may have evolved to “asleep” and the comparisons between the two. The surgical technique for DBS for movement disorders was developed in the late 1980 s and drew upon the steps used to perform a safe and effective radiofrequency thalamotomy [71]. At the time, MRI had not yet been incorporated as a routine part of stereotactic neurosurgical planning. Instead, CT and occasionally ventriculography were used to create a coordinate space, and a stereotactic atlas of the human brain was used to select a coordinate-based starting point. The targeted region was then mapped using test stimulation with or without MER to determine where to place the lead [72].

Subsequent advances in imaging, such as MRI, served to supplant the use of ventriculography at many institutions. High-resolution postoperative imaging allowed clinicians to compare functional outcomes following DBS with the actual location of the leads [73, 74]. This, along with advances that rendered targeted structures visible on MRI, led to “direct” anatomical targeting as an alternative to “indirect” atlas-based targeting. While a key and unresolved question remained whether the optimal DBS target is anatomical, physiological, or whether anatomy corresponds to physiology, the application of asleep DBS seemed inevitable [75, 76]. Studies evaluating functional outcomes following asleep DBS have resulted in growing confidence in using image-based targeting alone for DBS lead placement without confirmatory awake testing [47, 77‒80].

The expertise required to obtain good functional results with awake versus asleep surgery is different. Awake surgery is guided by a discerning neurological exam to assess clinical benefit and side effects during test stimulation and interpretation of electrophysiological data. With asleep DBS, there is greater reliance upon the navigation software, the quality of the MR images, and target selection by the surgeon. Sources of error that may require repositioning of the lead must be factored into both approaches and include intrinsic and human errors with the stereotactic frame, hidden errors in the navigation software, image distortion from the magnetic field, and brain shift resulting from pneumocephalus or changes in head position. The strategies to account for and minimize these errors vary between awake and asleep DBS [47, 81].

Ultimately, effective DBS surgery should include some means of confirming that the lead is appropriately placed at the time of surgery. This can be done with test stimulation, electrophysiological recordings, or intraoperative imaging.

MERs and Local Field Potentials

Regardless of the targeting method used, imaging and technical errors in DBS surgery introduce inaccuracies that can either cancel each other out or be additive [82]. In addition to the potential errors noted above, there are inevitable movements caused by force and compression mechanics with the introduction of cannulas and electrodes in brain parenchyma [83]. Real-world usage shows that there is significant discordance between anatomy as seen on the preoperative MRI and the neurophysiology mapping data, necessitating retargeting in approximately 20% of trajectories [84]. MERs and local field potentials (LFPs) offer neurophysiological validation of brain target location [85].

MER allows the recording of action potentials and the identification of specific neurons through the detection of neural signatures including spike firing rate, amplitude, morphology, and response to electrical stimulation. For instance, substantia nigra pars reticulata cells exhibit high frequency, low amplitude spikes that are inhibited with electrical stimulation, while STN cells have high amplitude, low-frequency spikes that do not exhibit inhibition to stimulation [86]. While spikes can help identify individual neuronal cell type, background voltage fluctuations reflect the activity of neuronal populations in the vicinity of the electrode tip (Fig. 9). The summation of multiple frequency bands (alpha, beta, theta, etc.) leads to a synthesis of the composite LFP time series (background, Fig. 9). Power changes in individual frequency bands are reflected in the time-series as small voltage fluctuations. Such information can help distinguish functional subdomains of targets to a millimetric scale. An example is increased beta band power in motor STN compared to its nonmotor ventral territory (overlay, Fig. 9) [87].

Fig. 9.

LFPs and Power spectra assist in target localization. Top panel shows a 3D rendering of the surgical trajectory of a microelectrode superimposed on a coronal MRI. a denotes the area of zona incerta. b, c denote dorsal and ventral STN border, respectively. d denotes substantia nigra. The lower panel shows an example of in vivo intraoperative MER of the surgical trajectory shown above. The overlay (bottom right) displays the power spectrum of LFPs obtained from dorsal and ventral STN (b, c). The asterisk on the spectra denotes an increased beta band power in dorsal STN (black) compared to ventral STN (red).

Fig. 9.

LFPs and Power spectra assist in target localization. Top panel shows a 3D rendering of the surgical trajectory of a microelectrode superimposed on a coronal MRI. a denotes the area of zona incerta. b, c denote dorsal and ventral STN border, respectively. d denotes substantia nigra. The lower panel shows an example of in vivo intraoperative MER of the surgical trajectory shown above. The overlay (bottom right) displays the power spectrum of LFPs obtained from dorsal and ventral STN (b, c). The asterisk on the spectra denotes an increased beta band power in dorsal STN (black) compared to ventral STN (red).

Close modal

LFPs can be acquired from both the high impedance microelectrodes and the low impedance DBS leads. LFPs can be used not only to improve intraoperative target localization but also potentially offer a physiological biomarker useful in choosing optimal contacts for stimulation and provide closed-loop systems with the needed information to optimize therapy [88‒93]. Additional advantages of physiological recordings include the unique opportunity to probe function of neurons and circuits in the human brain [94]. Disadvantages of intraoperative physiological mapping include the need for a skilled team that is trained to interpret the data, increased duration of surgery, and the possible increased number of passes through brain parenchyma. This may lead to increased risk of infection and bleeding [82, 95].

Several commercially available systems can provide in vivo physiological recordings. Advanced signal processing techniques and artificial intelligence algorithms are emerging to automate the interpretation of MER and LFP by detecting spectral signatures of the raw spiking data and automatically calculating nuclear subdomains and boundaries [96, 97]. The systems can analyze up to 5 electrodes simultaneously and suggest the best final implantation location based on real-time data. Early research also suggests that artificial intelligence can be used to monitor treatment effects [91, 98, 99] and optimize anesthesia [100‒102] to maximize patient comfort while minimizing sedation and providing a guide to closed-loop programming.

Stimulation

Intraoperative stimulation has been an integral part of stereotactic functional procedures since the very beginning of stereotaxis. Using a radiofrequency electrode, the target area was explored physiologically, and both positive and negative effects of stimulation were noted before performing a lesion. Intraoperative stimulation prior to creating a lesion was implemented for detection of capsular response, and high frequency stimulation in the ventrolateral thalamus was found to arrest tremor [103].

In DBS surgery, intraoperative macrostimulation through the contacts of the DBS lead for evaluation and verification of the electrode location in awake patients has been the gold standard of physiological mapping. The use of four contacts provides a higher resolution in mapping of the target and its surroundings compared to an radiofrequency electrode [104]. This also has the advantage of avoiding changes in lead position between stimulation and final electrode placement.

Recently, “directional” DBS electrodes have been introduced, providing the possibility of performing more detailed intraoperative mapping of the area around the electrode. The created “stimulation maps” are valuable tools for evaluation and orientation of the region [105]; however, this approach has certain limitations, especially in relation to alleviation of symptoms.

  • In some targets, such as the GPi, immediate clinical positive effects of intraoperative stimulation are rarely seen.

  • Intraoperative evaluation of gait and balance is not feasible.

  • Similar tremor arrest may be achieved from neighboring structures, such as the Vim of the thalamus and caudal zona incerta [106].

  • There is unpredictability of possible late emergence of side effects on chronic DBS that are not evident during the intraoperative stimulation, such as alteration of speech following STN DBS [107].

The development of asleep DBS may be considered as an advance. Unfortunately, it also means that the opportunity for a younger generation of neurosurgeons to learn from stimulating the awake brain will be lost and that a technique which has, often through fortuitous serendipity [108], been the driving force behind the introduction and refinement of many brain targets, will pass into obscurity. Moreover, and oddly, this trend goes against a current interest in doing as many neurosurgical operations as possible under local anesthesia, and it ignores the potential risks of intubation and general anesthesia in patients who are often old and debilitated.

Confirmatory Imaging

Why do Air Force pilots, despite using computer-guided, robot-assisted bombs for “surgical precision” bombing of a target, or unleashing a cruise missile based on sophisticated real-time image-guided trajectory, need to fly again above the target to verify and document that it has been hit and to check for eventual “collateral damage”? This is because despite all sophistications used for targeting, trajectory, and pin-pointing, the only way to ascertain the targeting precision is to verify it visually. So should stereotactic neurosurgeons reason and act.

The fact is that obtaining an immediate postoperative – or intraoperative – stereotactic image was indeed the routine among our historical masters: Ernest Spiegel and Henry Wycis used to inject a droplet of the contrast agent pantopaque in the target area immediately after lesioning to verify the position of the lesion on stereotactic X-ray. Irving Cooper did the same when performing pallidotomies or thalamotomies; he used contrast to inflate a balloon at the target to visualize the location of the lesion on X-ray. Lars Leksell used to stereotactically place silver clips at the target area following pallidotomy and perform a stereotactic X-ray to verify the lesion location in relation to the planned target [109]. In all these procedures, the intraoperative radiological verification was in relation to the stereotactic frame’s coordinates, not in relation to the anatomical structures per se, since the brain parenchyma could not be visualized on X-ray. With the advent of CT and especially MRI, one could immediately verify on a stereotactic imaging the existence of possible pneumocephalus or early hematoma, and one could verify the location of the lesion or the DBS lead, not only in stereotactic space in relation to the frame coordinates, but also in relation to the desired anatomical brain structure as visualized on MRI [65, 110].

A stereotactic functional procedure always starts with imaging, and surgeons put great effort and time in ensuring a geometrically precise and purposeful imaging procedure to extract as much information as possible regarding the brain target of interest. However, this meticulousness in preoperative imaging is not met by a similar care when it comes to immediate post-procedural imaging, as evidenced by publications illustrated with postoperative MRI figures that give no clue whatsoever as to the precise location of the leads [111, 112]. In fact, a DBS surgery should not be considered complete and the frame should not be detached from the head before confirmation by stereotactic imaging of accurate lead placement. This is the essence of the concept of image-guided and image-verified functional stereotactic neurosurgery.

Unfortunately, an immediate postoperative stereotactic CT or MRI has not been, and still is not, a routine in many centers, probably due to logistics and/or the lack of time (or will) after a lengthy surgery. Instead, a non-stereotactic regular (often a T1w) MRI, or more often a non-stereotactic CT scan is done a day or days later, mainly to rule out bleeding. Then, image-fusion software with inherent and variable fusion errors [113] is used to evaluate the approximate position of the leads.

The benefit of proper stereotactic imaging during surgery or immediately after carries many advantages: to validate the accuracy of the procedure in stereotactic space; allow repositioning of the lead in the same surgical session if its location is not optimal; and allow an exact confirmation of the position of each of the electrode contacts in the target area (Fig. 10). It will permit, if desired, performing asleep surgery; it will facilitate a proper and exact evaluation of the relationship of structure to function; and it will provide the exact anatomical basis for a future diffusion tensor imaging MR aimed at evaluating the pathways and circuits modulated by stimulation. Finally, high-resolution postoperative imaging may enable serendipitous discoveries whenever unexpected but positive effects of stimulation happen, paving the way for new indications for existing brain targets, or new brain targets for existing or new indications [108, 114]. Evidence, based on expert opinion, that proper intraoperative-postoperative structural MRI in patients with DBS hardware is a desirable technical advance is shown by the efforts of dedicated teams to explore and ensure the compatibility and safety of MRI, even at higher magnetic field strengths than 1.5T, in patients with implanted DBS systems [115].

Fig. 10.

Structural MRI implemented for GPi DBS. Top panel, a postoperative stereotactic 1.5T proton density MRI, with 2 mm-thick contiguous axial scans from level of AC-PC to 6 mm below. Bottom panel, contacts 0, 1, 2, and 3 of a Medtronic DBS electrode with depiction of their placement in the postero-ventro-lateral GPi.

Fig. 10.

Structural MRI implemented for GPi DBS. Top panel, a postoperative stereotactic 1.5T proton density MRI, with 2 mm-thick contiguous axial scans from level of AC-PC to 6 mm below. Bottom panel, contacts 0, 1, 2, and 3 of a Medtronic DBS electrode with depiction of their placement in the postero-ventro-lateral GPi.

Close modal

At the end of the day, what does a neurosurgeon do when they are referred a patient who is not doing well on DBS? They request and examine an MRI of the brain – not a PET scan, not a functional MRI, not a tractography MRI – but a structural MRI with adequate sequences, to see where the electrodes are. And the literature bears testimony to the all too many revisions of DBS leads that have occurred simply because neurosurgeons, regardless of the frame or physiological means they used during implantation, have neglected to perform a proper intra- or post-operative imaging to verify accurate lead location [116‒118].

When it comes to “Advances in Technical Aspects of DBS Surgery,” one of the most important “technical” aspects today, as has always been the case since the birth of stereotaxis, is imaging. In our MRI era, the most important preoperative and intraoperative/immediate postoperative imaging remains a “fifty shades of grey” structural MR image using sequences adapted to the brain target that needs to be visualised. This is one of the major advances contributing to a “paradigm shift” in DBS surgery [119]. This is the basis and sine qua non for a proper interpretation of all other imaging modalities in image-based DBS research, such as fMRI, DTI, and other “advanced” renderings of the brain which are so popular today in the “scholarly” DBS literature.

Robotics

Robotic arms have been used in the functional neurosurgical operating room since the 1980 s, with the first description [120] of the PUMA (Programmable Universal Machine for Assembly), an apparatus aimed to hold a needle for brain biopsy. In 1987, Benabid et al. [121] described the first surgical version of the Neuromate, with an ultrasound-based registration mode, designed to perform robotic-guided brain biopsy, but also to implant DBS leads or to perform stereoelectroencephalography. The typical robotic arm used in the field of stereotactic surgery is a supervisory controlled system [122] that allows the surgeon to: (1) plan offline, using dedicated software, a trajectory on a set of CT or MRI; (2) download the surgical plan to the surgical robot; and (3) supervise the robot that executes the plan (Fig. 11).

Fig. 11.

Advancements in robotic surgery and surgical visualization. a The OR at the Grenoble University Hospital, 2007. The Neuromate Robot used from 1987 to 2012 to implant DBS leads (pictured), brain biopsy, and sEEG. A frame-based or ultrasound-based registration was primarily used. b The OR at the Grenoble University Hospital, 2020. The ROSA Robot is used since 2012 for the same aims but is now coupled with an intraoperative CT scan, and uses a frameless registration based on bone fiducials.

Fig. 11.

Advancements in robotic surgery and surgical visualization. a The OR at the Grenoble University Hospital, 2007. The Neuromate Robot used from 1987 to 2012 to implant DBS leads (pictured), brain biopsy, and sEEG. A frame-based or ultrasound-based registration was primarily used. b The OR at the Grenoble University Hospital, 2020. The ROSA Robot is used since 2012 for the same aims but is now coupled with an intraoperative CT scan, and uses a frameless registration based on bone fiducials.

Close modal

Robotic arms allow unlimited trajectories, sub-millimetric applicative precision and accuracy of lead insertion, and unlimited 3D navigation around the head [123, 124]. Usually, different modes of co-registration are possible, including surface registration, frame-based registration, and frameless (bone fiducial) registration. Furthermore, these software programs can co-register different sets of images in the same space with limited fusion errors. There is no need to enter or check any x, y, and z coordinates, as the robot software automatically computes the coordinates of the entry point and target into 3D space, thereby negating human error. The surgeon only needs to click on an entry point and target on a defined set of images. It is also possible to correct, if needed, the axis of any trajectory with very small increments (0.5 mm or less). It remains critical for the surgeon to verify the safety of the planned trajectory.

A robotic arm must be used with caution and appropriate skills. We advise the following principles for the use of stereotactic robotic arms:

  • The head must be firmly fixed, to avoid inaccuracy, and to prevent any head movement when a surgical tool is inserted into the brain. We use a stereotactic frame as a head holder and never use any “clamps” with nonrigid, head fixations.

  • We recommend intraoperative imaging at the beginning of the first trajectory to check for any possible deviation and the depth of probe insertion; the 3D images obtained during surgery can be easily and rapidly co-registered into surgical planning. The movement of the robot must be systematically supervised by the surgeon to avoid any collision.

  • Preventative maintenance of the robot must be regularly performed by the manufacturer.

Electrode Implantation Technique

Important advances in DBS lead and lead placement methodology have been made in recent years. Technologies associated with lead placement including frame-based, frameless, and robotic methods have seen advances, as covered in previous sections. Here, we will discuss significant innovations in DBS lead technology, with a focus on efficacy and complication avoidance. Lead directionality, as noted above, has significant potential benefits for increasing the therapeutic window in patients with PD or ET who are having DBS systems implanted [125‒128].

Once a strategy is put in place to utilize directionality, it is important to confirm lead rotation radiographically and secure the lead in such a manner as to not have migration or malrotation. Fluoroscopy, or other forms of intraoperative imaging, can be utilized to confirm the rotational orientation of a lead. Flat-panel CT has been demonstrated to be accurate in determining the rotational orientation of DBS leads and can allow for standardized templates to be developed that are specific to different lead designs [129]. It should be maintained that the therapeutic agency of DBS – even directional DBS – relies on accurate placement of electrodes, and directionality will not compensate for a poorly placed lead [130].

Upon confirmation of orientation, the DBS lead requires secure fixation. Animal studies have demonstrated postimplantation rotation of DBS leads, in particular when implanted with torque [131]. Hence, an assortment of fixation techniques has been used over the years, including proprietary caps, titanium plates, and various cements, with preventing lead migration in mind. Given the need to maintain proper rotational orientation, it becomes more important to properly secure leads according to the manufacturer’s recommendations. Even optimally designed, placed, and secured leads are only as good as their connection to the pulse generator. Typically, an intermediate connection via an extension lead is used. Extension leads allow for flexibility in pulse generator placement, while ideally shielding the primary lead from any undue manipulation. While retromastoid tunneling to an anterior chest wall pulse generator is typical, there may be certain situations where abdominal or posterior pulse generator placement may be preferred. In these scenarios, different strategies will need to be pursued in the selection of extension lead and in tunneling technique. There are variations in length and gauge of these leads in addition to differences in connectivity and elasticity. Table 2 summarizes aspects of directionality, fixation, and connectivity in commonly used, commercially available DBS leads.

Table 2.

Comparison of DBS leads and functionality

 Comparison of DBS leads and functionality
 Comparison of DBS leads and functionality

Implantable Pulse Generators

When DBS was first introduced for clinical use, IPGs for DBS were only available as a non-rechargeable option and had to be replaced within three to 5 years postimplantation [132]. Rechargeable (RC) IPGs offer a number of advantages over non-rechargeable devices including longer IPG life span, resulting in fewer IPG replacement surgeries along with significant cost reduction [133]. Nowadays, the choice of using non-rechargeable or RC IPGs becomes important not only for surgeon and neurologist but also for patients and caregivers. While RC IPGs have the benefit of reducing the number of needed IPG replacement surgeries, patients must be reliable and capable of appropriately recharging the battery on a regular basis.

Currently, the choice of IPGs has been diversified by the entry of new manufacturers into the DBS market. Prominent features of IPGs from each vendor are summarized in Table 3 [127].

Table 3.

Features of currently available implantable pulse generators. Adapted from [127]

 Features of currently available implantable pulse generators. Adapted from [127]
 Features of currently available implantable pulse generators. Adapted from [127]

Innovations in IPG technology include not only RC systems with various controlling patterns but also advances in shaping and miniaturization, which improve clinical outcomes and patient comfort [127]. For patients who are independent and socially active, RC IPGs may be more beneficial. However, for those who are not independent and require social support, the recharging process may be somewhat complicated and difficult. Another important aspect of IPGs is their volume and shape. Especially for the pediatric or the slim patient, these factors can produce problems such as protrusion, infection, and pain. For example, Choi et al. [134] reported that some patients found the pain and cosmetic outcome from IPG implantation to be worse than expected.

RC IPG technology still needs improvement to reduce charging time and frequency. Perhaps an ideal option would be an IPG that could be placed on the skull surface, or even in the burr hole, and allow for a minimally invasive, awake surgical procedure. As well, both non-rechargeable and RC IPGs should possess the decoding technology to provide optimal flexible stimulation to adjust as the patient’s needs warrant [6]. Future innovations of IPG technology will greatly improve the patient’s symptoms. Infection must be avoided during IPG implantation, as the infection rate is reported to be around 2–5% in recent reports [135‒138]. As infection necessitates removal of the IPG, measures must be taken to mitigate this risk both peri- and post-operatively. First, studies suggest that spreading vancomycin powder in the IPG insertion site may reduce rates of infection [139]. More recently, and akin to cardiac implantable electronic devices, absorbable antibacterial envelopes that surround the IPG have shown promise in preventing infection [140]. In patients having nonrechargeable IPGs replaced, some literature supports that subsequent IPG replacement procedures carry similar infection risk [67], while others suggest that removal of a portion of the avascular subcutaneous pocket may restore regional blood flow and decrease the risk of infection [141].

Imaging and Anatomical Modeling

After the surgical procedure, a crucial step that completes the technological framework of DBS is to reconstruct electrode placement. This is essential to confirm successful localization in the target structure and to guide stimulation settings.

Electrode localization can be confirmed using X-ray images (which have the downside of projecting data to single planes), registrations between postoperative CT and preoperative data (with the advantage of high signal-to-noise ratio for the electrodes), or acquiring postoperative MRI (with the advantage of showing both electrodes and target structures in the same image). In the first two approaches, nonlinear effects of brain shift due to pneumocephalus need to be considered, so that even if intraoperative CT or MRI is acquired at the conclusion of surgery, a delayed image should be acquired as well.

These approaches should be considered gold standards for clinical evaluations in the individual patient. However, if the aim is to compare information across patients (or even across cohorts and DBS centers), a model-based reconstruction of electrode placement is needed. With retrospective modelling based on stimulation location, we refer to a spatial definition of the target structure, such as the STN (Fig. 12). The involved 3D representation can further aid in DBS programming.

Fig. 12.

A combination of advanced approaches and precise datasets may lead to useful and meaningful DBS models. The figure shows a non-exhaustive list of approaches that were developed during the last 5 years to refine image registration, electrode localization, and biophysical modeling. Example datasets that augment model-based DBS reconstructions are shown. For instance, expert-defined streamlines based on animal data (pathway atlas), group-derived optimal stimulation sites (bradykinesia sweet spot), histological stack data, or multimodal atlas datasets can be helpful to inform the spatial properties of the model.

Fig. 12.

A combination of advanced approaches and precise datasets may lead to useful and meaningful DBS models. The figure shows a non-exhaustive list of approaches that were developed during the last 5 years to refine image registration, electrode localization, and biophysical modeling. Example datasets that augment model-based DBS reconstructions are shown. For instance, expert-defined streamlines based on animal data (pathway atlas), group-derived optimal stimulation sites (bradykinesia sweet spot), histological stack data, or multimodal atlas datasets can be helpful to inform the spatial properties of the model.

Close modal

Beyond the ability to compare results across patients, a model-based concept has other advantages. First, it allows group-level data analysis [142] and the creation of symptom-specific probabilistic “sweet spots” [12, 143, 144]. Models allow for the creation of algorithms for automated stimulation parameter tuning [145] and calculating brain connectivity profiles associated with optimal outcomes [146]. Finally, model-based representations allow for combinations with 3D atlas data derived from histology [56], precise expert-defined brain atlases of complex regions [140], large normative connectome databases [12, 147], insights derived from brain lesion data [148], and datasets informed by combinations of histology, animal research, and expert knowledge [149].

However, as with any type of model, this approach comes with shortcomings, since precise image registrations are needed and instead of the raw data, a model of it is shown. While creating a 3D model of electrode placement is easy, creating one that is actually meaningful is surprisingly complex [150]. The reason for this is threefold: first, stereotactic targets are small (∼10 × 4 × 6 mm for the STN) and imaging resolution often suboptimal (∼0.5 × 0.5 × 2 mm for a typical T2-sequence). Second, DBS targets are surrounded by small but functionally significant grey and white matter structures [149, 151, 152]. Third, millimeters matter; spatial shifts of electrodes by a mere 2 mm may have significant effects on clinical outcomes [55].

To tackle these challenges, a variety of tools have been introduced that aim at improving image registration and electrode localizations. For instance, modern pipelines [55] involve empirically validated multispectral registration algorithms [153], manual refinement tools for nonlinear registrations [154], brain shift correction strategies [55], phantom-validated electrode localization [155], and orientation detection algorithms [156]. Together with precise atlas datasets defined from a combination of histology, and structural and diffusion MRI [68], derived 3D models have become capable of predicting clinical outcomes across cohorts and centers [143, 147, 157‒161].

While this shows that models have become useful – on a group level – it does not automatically mean they are meaningful, especially on a single patient level. In other words, predictive models are not necessarily mechanistic models. Further indirect validation for accuracy of the aforementioned models has come from electrophysiological work [162, 163]. For instance, Rappel et al. [164] compared spiking activity data from 933 trajectories with electrode localizations and reported agreement not only with the STN model but even with its functional subzones. Recently, Kehnemouyi et al. [165] demonstrated that the degree of stimulation of the STN derived from such 3D models accounted for a decrease in beta burst duration. Weaver et al. [166] used 3D models that involved DBS, direct cortical stimulation, and tractography to shed light on human hyperdirect pathway functionality. While such confirmation studies largely augment the value of computerized 3D DBS models, further work to confirm and optimize them is still needed. These models may increase our understanding of human neurophysiology and help create better models as part of a virtuous cycle (e.g., using a known sweet spot, adjusting surgical technique, and creating a new model with the new information).

Our understanding and implementation of DBS as a therapeutic procedure are rapidly expanding, with its great potential to modulate neurocircuitry at the tissue level in a minimally invasive, but highly effective, manner. The techniques of DBS system implantation have been greatly improved by developments and innovation in stereotactic technology, surgical techniques, and imaging. Innovations in DBS technology and surgery promise to make DBS surgery more accurate and effective, with potentially better outcomes for a wider spectrum of patients; however, such advancements should not become too complicated and cumbersome such that their dissemination and widespread use are hindered. Additional multidisciplinary work is needed to translate these technological advances into patient benefit.

AMishra is partially supported by the Neurosurgery Research and Education Foundation (2022 Medical Student Summer Research Fellowship). PB is a consultant for Abbott and Boston Scientific, and shareholder in Mithridaticum AB. AB serves as a consultant for Abbott. SC serves as a consultant for Medtronic, Boston Scientific, and Zimmer Biomet. MH received lecturing fees from Boston Scientific. AH was supported by the (Deutsche Forschungsgemeinschaft, 424778381 – TRR 295), Deutsches Zentrum für Luftund Raumfahrt (DynaSti grant within the EU Joint Programme-Neurodegenerative Disease Research, JPND), the National Institutes of Health (R01 13478451, 1R01NS127892-01 & 2R01 MH113929) as well as the New Venture Fund (FFOR Seed Grant). AML is a consultant to Abbott, Boston Scientific, Insightec, and Medtronic, and is a Scientific Director at Functional Neuromodulation. JSN performs limited consulting for Abbott, FHC, and Boston Scientific. FP is a consultant for Boston Scientific, Abbott, and Medtronic. PAS receives research support from Medtronic. JKK is a consultant for Medtronic, Boston Scientific, Aleva, and Inomed. All other authors declare no conflicts of interest.

The work on this manuscript was supported by an unconditional grant of the World Society for Stereotactic and Functional Neurosurgery. The working process was coordinated with the Research and Education Committees of the World Society for Stereotactic and Functional Neurosurgery.

M.S., A.M., A.M., A.H., A.B., P.B., S.C., O.F., A.L., J.N., F.P., P.S., J.K., M.H., and J.C. prepared and drafted the manuscript, critically edited the manuscript, and approved the manuscript for submission.

1.
Lozano AM, Lipsman N. Probing and regulating dysfunctional circuits using deep brain stimulation. Neuron. 2013;77(3):406–24.
2.
Chou KL, Grube S, Patil PG. Deep brain stimulation: a new life for people with Parkinson’s, dystonic, and essential tremor. New York: Demos Medical Publishing; 2011.
3.
Taghva A, Corrigan JD, Rezai AR. Obesity and brain addiction circuitry: implications for deep brain stimulation. Neurosurgery. 2012;71(2):224–38.
4.
Vedam-Mai V, Deisseroth K, Giordano J, Lazaro-Munoz G, Chiong W, Suthana N, et al. Proceedings of the eighth annual deep brain stimulation think tank: advances in optogenetics, ethical issues affecting DBS research, neuromodulatory approaches for depression, adaptive neurostimulation, and emerging DBS technologies. Front Hum Neurosci. 2021;15:644593.
5.
Lozano AM, Lipsman N, Bergman H, Brown P, Chabardes S, Chang JW, et al. Deep brain stimulation: current challenges and future directions. Nat Rev Neurol. 2019;15(3):148–60.
6.
Krauss JK, Lipsman N, Aziz T, Boutet A, Brown P, Chang JW, et al. Technology of deep brain stimulation: current status and future directions. Nat Rev Neurol. 2021;17(2):75–87.
7.
Spiegel EA, Wycis HT, Marks M, Lee AJ. Stereotaxic apparatus for operations on the human brain. Science. 1947;106(2754):349–50.
8.
Tasker RR. Simple localization for stereoencephalotomy using the “portable” central beam of the image intensifier. Confin Neurol. 1965;26(3):209–12.
9.
Talairach J, De Aljuriaguerra J, David M. A stereotaxic study of the deep encephalic structures in man; technic; physiopathologic and therapeutic significance. Presse Med. 1952;60(28):605–9.
10.
Mehrkens JH, Bötzel K, Steude U, Zeitler K, Schnitzler A, Sturm V, et al. Long-term efficacy and safety of chronic globus pallidus internus stimulation in different types of primary dystonia. Stereotact Funct Neurosurg. 2009;87(1):8–17.
11.
Foltynie T, Zrinzo L, Martinez-Torres I, Tripoliti E, Petersen E, Holl E, et al. MRI-guided STN DBS in Parkinson’s disease without microelectrode recording: efficacy and safety. J Neurol Neurosurg Psychiatry. 2011;82(4):358–63.
12.
Boutet A, Gramer R, Steele CJ, Elias GJB, Germann J, Maciel R, et al. Neuroimaging technological advancements for targeting in functional neurosurgery. Curr Neurol Neurosci Rep. 2019;19(7):42.
13.
Duchin Y, Shamir RR, Patriat R, Kim J, Vitek JL, Sapiro G, et al. Patient-specific anatomical model for deep brain stimulation based on 7 Tesla MRI. PLoS ONE. 2018;13(8):e0201469.
14.
Pallavaram S, Yu H, Spooner J, D’Haese PF, Bodenheimer B, Konrad PE, et al. Intersurgeon variability in the selection of anterior and posterior commissures and its potential effects on target localization. Stereotact Funct Neurosurg. 2008;86(2):113–9.
15.
Bot M, Schuurman PR, Odekerken VJJ, Verhagen R, Contarino FM, De Bie RMA, et al. Deep brain stimulation for Parkinson’s disease: defining the optimal location within the subthalamic nucleus. J Neurol Neurosurg Psychiatry. 2018;89(5):493–8.
16.
Li Z, Zhang JG, Ye Y, Li X. Review on factors affecting targeting accuracy of deep brain stimulation electrode implantation between 2001 and 2015. Stereotact Funct Neurosurg. 2016;94(6):351–62.
17.
Bender B, Mänz C, Korn A, Nägele T, Klose U. Optimized 3D magnetization-prepared rapid acquisition of gradient echo: identification of thalamus substructures at 3T. AJNR Am J Neuroradiol. 2011;32(11):2110–5.
18.
Sudhyadhom A, Haq IU, Foote KD, Okun MS, Bova FJ. A high resolution and high contrast MRI for differentiation of subcortical structures for DBS targeting: the Fast Gray Matter Acquisition T1 Inversion Recovery (FGATIR). NeuroImage. 2009;47(Suppl 2):T44–52.
19.
Beaumont J, Saint-Jalmes H, Acosta O, Kober T, Tanner M, Ferré JC, et al. Multi T1-weighted contrast MRI with fluid and white matter suppression at 1.5 T. Magn Reson Imaging. 2019;63:217–25.
20.
Vassal F, Coste J, Derost P, Mendes V, Gabrillargues J, Nuti C, et al. Direct stereotactic targeting of the ventrointermediate nucleus of the thalamus based on anatomic 1.5-T MRI mapping with a white matter attenuated inversion recovery (WAIR) sequence. Brain Stimul. 2012;5(4):625–33.
21.
Rasouli J, Ramdhani R, Panov FE, Dimov A, Zhang Y, Cho C, et al. Utilization of quantitative susceptibility mapping for direct targeting of the subthalamic nucleus during deep brain stimulation surgery. Oper Neurosurg. 2018;14(4):412–9.
22.
Ide S, Kakeda S, Ueda I, Watanabe K, Murakami Y, Moriya J, et al. Internal structures of the globus pallidus in patients with Parkinson’s disease: evaluation with quantitative susceptibility mapping (QSM). Eur Radiol. 2015;25(3):710–8.
23.
Vertinsky AT, Coenen VA, Lang DJ, Kolind S, Honey CR, Li D, et al. Localization of the subthalamic nucleus: optimization with susceptibility-weighted phase MR imaging. AJNR Am J Neuroradiol. 2009;30(9):1717–24.
24.
Liu T, Eskreis-Winkler S, Schweitzer AD, Chen W, Kaplitt MG, Tsiouris AJ, et al. Improved subthalamic nucleus depiction with quantitative susceptibility mapping. Radiology. 2013;269(1):216–23.
25.
Tasserie J, Lozano AM. Editorial. 7T MRI for neuronavigation: toward better visualization during functional surgery. J Neurosurg. 2022;137(5):1262–3.
26.
Chandran AS, Bynevelt M, Lind CRP. Magnetic resonance imaging of the subthalamic nucleus for deep brain stimulation. J Neurosurg. 2016;124(1):96–105.
27.
Zonenshayn M, Rezai AR, Mogilner AY, Beric A, Sterio D, Kelly PJ. Comparison of anatomic and neurophysiological methods for subthalamic nucleus targeting. Neurosurgery. 2000;47(2):282–92.
28.
Kraff O, Quick HH. 7T: physics, safety, and potential clinical applications. J Magn Reson Imaging. 2017;46(6):1573–89.
29.
Springer E, Dymerska B, Cardoso PL, Robinson SD, Weisstanner C, Wiest R, et al. Comparison of routine brain imaging at 3 T and 7 T. Invest Radiol. 2016;51(8):469–82.
30.
Abosch A, Yacoub E, Ugurbil K, Harel N. An assessment of current brain targets for deep brain stimulation surgery with susceptibility-weighted imaging at 7 tesla. Neurosurgery. 2010;67(6):1745–56.
31.
Forstmann BU, Isaacs BR, Temel Y. Ultra high field MRI-guided deep brain stimulation. Trends Biotechnol. 2017;35(10):904–7.
32.
Dammann P, Kraff O, Wrede KH, Özkan N, Orzada S, Mueller OM, et al. Evaluation of hardware-related geometrical distortion in structural MRI at 7 tesla for image-guided applications in neurosurgery. Acad Radiol. 2011;18:910–6.
33.
Leksell L. The stereotaxic method and radiosurgery of the brain. Acta Chir Scand. 1951;102(4):316–9.
34.
Solberg TD, Siddon RL, Kavanagh B. Historical development of stereotactic ablative radiotherapy. Berlin, Heidelberg: Springer; 2012.
35.
Konrad PE, Neimat JS, Yu H, Kao CC, Remple MS, D’Haese PF, et al. Customized, miniature rapid-prototype stereotactic frames for use in deep brain stimulator surgery: initial clinical methodology and experience from 263 patients from 2002 to 2008. Stereotact Funct Neurosurg. 2011;89(1):34–41.
36.
D’Haese PF, Pallavaram S, Konrad PE, Neimat J, Fitzpatrick JM, Dawant BM. Clinical accuracy of a customized stereotactic platform for deep brain stimulation after accounting for brain shift. Stereotact Funct Neurosurg. 2010;88(2):81–7.
37.
Bjartmarz H, Rehncrona S. Comparison of accuracy and precision between frame-based and frameless stereotactic navigation for deep brain stimulation electrode implantation. Stereotact Funct Neurosurg. 2007;85(5):235–42.
38.
Ball TJ, John KD, Donovan AM, Neimat JS. Deep brain stimulation lead implantation using a customized rapidly manufactured stereotactic fixture with submillimetric euclidean accuracy. Stereotact Funct Neurosurg. 2020;98(4):248–55.
39.
Bradac O, Steklacova A, Nebrenska K, Vrana J, de Lacy P, Benes V. Accuracy of VarioGuide frameless stereotactic system against frame-based stereotaxy: prospective, randomized, single-center study. World Neurosurg. 2017;104:831–40.
40.
Eleopra R, Rinaldo S, Devigili G, Mondani M, D’Auria S, Golfrè Andreasi N, et al. Globus pallidus internus deep brain stimulation using frame-based vs. Frameless stereotaxy in dystonia: a single-center experience. Front Neurol. 2021;12:643757.
41.
Khan FR, Henderson JM. Deep brain stimulation surgical techniques. Handb Clin Neurol. 2013;116:27–37.
42.
Henderson JM, Holloway KL, Gaede SE, Rosenow JM. The application accuracy of a skull-mounted trajectory guide system for image-guided functional neurosurgery. Comput Aided Surg. 2004;9(4):155–60.
43.
Starr PA, Martin AJ, Ostrem JL, Talke P, Levesque N, Larson PS. Subthalamic nucleus deep brain stimulator placement using high-field interventional magnetic resonance imaging and a skull-mounted aiming device: technique and application accuracy. J Neurosurg. 2010;112(3):479–90.
44.
Starr PA, Markun LC, Larson PS, Volz MM, Martin AJ, Ostrem JL. Interventional MRI-guided deep brain stimulation in pediatric dystonia: first experience with the clearpoint system. J Neurosurg Pediatr. 2014;14(4):400–8.
45.
Larson PS, Starr PA, Bates G, Tansey L, Richardson RM, Martin AJ. An optimized system for interventional magnetic resonance imaging-guided stereotactic surgery: preliminary evaluation of targeting accuracy. Neurosurgery. 2012;70(1 Suppl Operative):95–103.
46.
Larson P, Starr PA, Ostrem JL, Galifianakis N, San Luciano Palenzuela M, Martin A. Application accuracy of a second generation interventional MRI stereotactic platform: initial experience in 101 DBS electrode implantations. Neurosurgery. 2013;60(Suppl 1):187.
47.
Ostrem JL, Ziman N, Galifianakis NB, Starr PA, Luciano MS, Katz M, et al. Clinical outcomes using clearpoint interventional MRI for deep brain stimulation lead placement in Parkinson’s disease. J Neurosurg. 2016;124(4):908–16.
48.
Coffey RJ. Printed stereotactic atlases, review. In: Lozano AM, Gildenberg PL, Tasker RR, editors. Textb. Stereotact. Funct. Neurosurg. Berlin: Springer Verlag; 2009. p. 347–71.
49.
Schaltenbrand G, Bailey W. Atlas of stereotaxy of the human brain. Stuttgart: Georg Thieme Verlag; 1959.
50.
Schaltenbrand G, Wahren W. Atlas of stereotaxy of the human brain. Stuttgart: Georg Thieme Verlag; 1977.
51.
Nowinski WL. Evolution of human brain atlases in terms of content, applications, functionality, and availability. Neuroinformatics. 2021;19(1):1–22.
52.
Nowinski WL. The cerefy brain atlases: continuous enhancement of the electronic talairach-tournoux brain atlas. Neuroinformatics. 2005;3(4):293–300.
53.
Horn A, Blankenburg F. Toward a standardized structural-functional group connectome in MNI space. NeuroImage. 2016;124(Pt A):310–22.
54.
Lemaire JJ, De Salles A, Coll G, El Ouadih Y, Chaix R, Coste J, et al. MRI atlas of the human deep brain. Front Neurol. 2019;10:851.
55.
Horn A, Li N, Dembek TA, Kappel A, Boulay C, Ewert S, et al. Lead-DBS v2: towards a comprehensive pipeline for deep brain stimulation imaging. NeuroImage. 2019;184:293–316.
56.
Ilinsky I, Horn A, Paul-Gilloteaux P, Gressens P, Verney C, Kultas-Ilinsky K. Human motor thalamus reconstructed in 3D from continuous sagittal sections with identified subcortical afferent territories. eNeuro. 2018;5(3):ENEURO.0060–18.2018.
57.
Sedrak M, Gorgulho A, Frew A, Behnke E, Desalles A, Pouratian N. Diffusion tensor imaging and colored fractional anisotropy mapping of the ventralis intermedius nucleus of the thalamus. Neurosurgery. 2011;69(5):1124–9.
58.
Nowacki A, Debove I, Rossi F, Schlaeppi JA, Petermann K, Wiest R, et al. Targeting the posterior subthalamic area for essential tremor: proposal for MRI-based anatomical landmarks. J Neurosurg. 2018;131(3):820–7.
59.
Luján JL, Noecker AM, Butson CR, Cooper SE, Walter BL, Vitek JL, et al. Automated 3-dimensional brain atlas fitting to microelectrode recordings from deep brain stimulation surgeries. Stereotact Funct Neurosurg. 2009;87(4):229–40.
60.
Zhang Y, Wei H, Cronin MJ, He N, Yan F, Liu C. Longitudinal atlas for normative human brain development and aging over the lifespan using quantitative susceptibility mapping. NeuroImage. 2018;171:176–89.
61.
Sivaswamy J, Thottupattu AJ, Mehta R, Sheelakumari R, Kesavadas C. Construction of Indian human brain atlas. Neurol India. 2019;67(1):229–34.
62.
Ashkan K, Blomstedt P, Zrinzo L, Tisch S, Yousry T, Limousin-Dowsey P, et al. Variability of the subthalamic nucleus: the case for direct MRI guided targeting. Br J Neurosurg. 2007;21(2):197–200.
63.
Horn A, Kühn AA. Lead-DBS: a toolbox for deep brain stimulation electrode localizations and visualizations. NeuroImage. 2015;107:127–35.
64.
Rezai AR, Kopell BH, Gross RE, Vitek JL, Sharan AD, Limousin P, et al. Deep brain stimulation for Parkinson’s disease: surgical issues. Mov Disord. 2006;21(Suppl 14):S197–218.
65.
Hyam JA, Akram H, Foltynie T, Limousin P, Hariz M, Zrinzo L. What you see is what you get: lead location within deep brain structures is accurately depicted by stereotactic magnetic resonance imaging. Neurosurgery. 2015;11(Suppl 3):412–9.
66.
Bezchlibnyk YB, Sharma VD, Naik KB, Isbaine F, Gale JT, Cheng J, et al. Clinical outcomes of globus pallidus deep brain stimulation for Parkinson disease: a comparison of intraoperative MRI- and MER-guided lead placement. J Neurosurg. 2020;134(3):1072–82.
67.
Matias CM, Frizon LA, Nagel SJ, Lobel DA, Machado AG. Deep brain stimulation outcomes in patients implanted under general anesthesia with frame-based stereotaxy and intraoperative MRI. J Neurosurg. 2018;129(6):1572–8.
68.
Ewert S, Plettig P, Li N, Chakravarty MM, Collins DL, Herrington TM, et al. Toward defining deep brain stimulation targets in MNI space: a subcortical atlas based on multimodal MRI, histology and structural connectivity. NeuroImage. 2018;170:271–82.
69.
Anderson DN, Osting B, Vorwerk J, Dorval AD, Butson CR. Optimized programming algorithm for cylindrical and directional deep brain stimulation electrodes. J Neural Eng. 2018;15(2):026005.
70.
Mirzadeh Z, Ponce FA. DBS with versus without MER: clinical equipoise or malpractice? Mov Disord. 2015;30(3):439–41.
71.
Aziz TZ, Hariz M. To sleep or not to sleep during deep brain stimulation surgery for Parkinson disease? Neurology. 2017;89(19):1938–9.
72.
Hariz MI, Bergenheim AT. A comparative study on ventriculographic and computerized tomography-guided determinations of brain targets in functional stereotaxis. J Neurosurg. 1990;73(4):565–71.
73.
Wodarg F, Herzog J, Reese R, Falk D, Pinsker MO, Steigerwald F, et al. Stimulation site within the MRI-defined STN predicts postoperative motor outcome. Mov Disord. 2012;27(7):874–9.
74.
Starr PA, Turner RS, Rau G, Lindsey N, Heath S, Volz M, et al. Microelectrode-guided implantation of deep brain stimulators into the globus pallidus internus for dystonia: techniques, electrode locations, and outcomes. J Neurosurg. 2006;104(4):488–501.
75.
Montgomery EB. Validation of CT-MRI fusion for intraoperative assessment of stereotactic accuracy in DBS surgery. Mov Disord. 2015;30(3):439.
76.
Lozano AM. Deep brain stimulation for Parkinson disease. J Neurosurg. 2010;112(3):477–8.
77.
Chen T, Mirzadeh Z, Ponce FA. Asleep” deep brain stimulation surgery: a critical review of the literature. World Neurosurg. 2017;105:191–8.
78.
Chen T, Mirzadeh Z, Chapple KM, Lambert M, Shill HA, Moguel-Cobos G, et al. Clinical outcomes following awake and asleep deep brain stimulation for Parkinson disease. J Neurosurg. 2019;65:1–12.
79.
Ho AL, Ali R, Connolly ID, Henderson JM, Dhall R, Stein SC, et al. Awake versus asleep deep brain stimulation for Parkinson’s disease: a critical comparison and meta-analysis. J Neurol Neurosurg Psychiatry. 2018;89(7):687–91.
80.
Liu Z, He S, Li L. General anesthesia versus local anesthesia for deep brain stimulation in Parkinson’s disease: a meta-analysis. Stereotact Funct Neurosurg. 2019;97(5–6):381–90.
81.
Ko AL, Magown P, Ozpinar A, Hamzaoglu V, Burchiel KJ. Asleep deep brain stimulation reduces incidence of intracranial Air during electrode implantation. Stereotact Funct Neurosurg. 2018;96(2):83–90.
82.
Binder DK, Rau G, Starr PA. Hemorrhagic complications of microelectrode-guided deep brain stimulation. Stereotact Funct Neurosurg. 2003;80(1–4):28–31.
83.
Khan MF, Mewes K, Gross RE, Škrinjar O. Assessment of brain shift related to deep brain stimulation surgery. Stereotact Funct Neurosurg. 2008;86(1):44–53.
84.
Lozano CS, Ranjan M, Boutet A, Xu DS, Kucharczyk W, Fasano A, et al. Imaging alone versus microelectrode recording–guided targeting of the STN in patients with Parkinson’s disease. J Neurosurg. 2019;130(6):1847–52.
85.
Shamir RR, Duchin Y, Kim J, Patriat R, Marmor O, Bergman H, et al. Microelectrode recordings validate the clinical visualization of subthalamic-nucleus based on 7T magnetic resonance imaging and machine learning for deep brain stimulation surgery. Neurosurgery. 2019;84(3):749–57.
86.
Hutchison WD, Allan RJ, Opitz H, Levy R, Dostrovsky JO, Lang AE, et al. Neurophysiological identification of the subthalamic nucleus in surgery for Parkinson’s disease. Ann Neurol. 1998;44(4):622–8.
87.
Zaidel A, Spivak A, Grieb B, Bergman H, Israel Z. Subthalamic span of beta oscillations predicts deep brain stimulation efficacy for patients with Parkinson’s disease. Brain. 2010;133(Pt 7):2007–21.
88.
Flouty O, Yamamoto K, Green A, Aziz T. Commentary: operative technique and lessons learned from surgical implantation of the NeuroPace responsive Neurostimulation® system in 57 consecutive patients. Oper Neurosurg. 2021;20(2):E110–111.
89.
Fountas KN, Smith JR, Murro AM, Politsky J, Park YD, Jenkins PD. Implantation of a closed-loop stimulation in the management of medically refractory focal epilepsy: a technical note. Stereotact Funct Neurosurg. 2005;83(4):153–8.
90.
Krucoff MO, Wozny TA, Lee AT, Rao VR, Chang EF. Operative technique and lessons learned from surgical implantation of the neuropace responsive neurostimulation® system in 57 consecutive patients. Oper Neurosurg. 2021;20(2):E98–109.
91.
Tan H, Debarros J, He S, Pogosyan A, Aziz TZ, Huang Y, et al. Decoding voluntary movements and postural tremor based on thalamic LFPs as a basis for closed-loop stimulation for essential tremor. Brain Stimul. 2019;12(4):858–67.
92.
Senova S, Chaillet A, Lozano AM. Fornical closed-loop stimulation for alzheimer’s disease. Trends Neurosci. 2018;41(7):418–28.
93.
Shah SA, Tinkhauser G, Chen CC, Little S, Brown P. Parkinsonian tremor detection from subthalamic nucleus local field potentials for closed-loop deep brain stimulation. Annu Int Conf IEEE Eng Med Biol Soc. 2018. 2320–4.
94.
Lozano AM, Snyder BJ, Hamani C, Hutchison WD, Dostrovsky JO. Basal ganglia physiology and deep brain stimulation. Mov Disord. 2010;25(Suppl 1):S71–75.
95.
Hariz MI. Safety and risk of microelectrode recording in surgery for movement disorders. Stereotact Funct Neurosurg. 2002;78(3–4):146–57.
96.
Valsky D, Blackwell KT, Tamir I, Eitan R, Bergman H, Israel Z. Real-time machine learning classification of pallidal borders during deep brain stimulation surgery. J Neural Eng. 2020;17(1):016021.
97.
Thompson JA, Oukal S, Bergman H, Ojemann S, Hebb AO, Hanrahan S, et al. Semi-automated application for estimating subthalamic nucleus boundaries and optimal target selection for deep brain stimulation implantation surgery. J Neurosurg. 2018;130(4):1–10.
98.
Yao L, Brown P, Shoaran M. Resting tremor detection in Parkinson’s disease with machine learning and kalman filtering. Proceedings of the 2018 IEEE Biomed Circuits Syst Conf Biocas. Piscataway: IEEE; 2018.
99.
Gao C, Sun H, Wang T, Tang M, Bohnen NI, Müller MLTM, et al. Model-based and model-free machine learning techniques for diagnostic prediction and classification of clinical outcomes in Parkinson’s disease. Sci Rep. 2018;8(1):7129.
100.
Allen R, Smith D. Neuro-fuzzy closed-loop control of depth of anaesthesia. Artif Intell Med. 2001;21(1–3):185–91.
101.
Alexander JC, Romito BT, Çobanoǧlu MC. The present and future role of artificial intelligence and machine learning in anesthesiology. Int Anesthesiol Clin. 2020;58(4):7–16.
102.
Connor CW. Artificial intelligence and machine learning in anesthesiology. Anesthesiology. 2019;131(6):1346–59.
103.
Schaltenbrand G, Walker EA. Stereotaxic surgery for tremor. Ster. Hum. Brain. New York: Thieme-Stratton; 1982. p. 515–21.
104.
Shah A, Vogel D, Alonso F, Lemaire JJ, Pison D, Coste J, et al. Stimulation maps: visualization of results of quantitative intraoperative testing for deep brain stimulation surgery. Med Biol Eng Comput. 2020;58(4):771–84.
105.
Fytagoridis A, Åström M, Samuelsson J, Blomstedt P. Deep brain stimulation of the caudal zona incerta: tremor control in relation to the location of stimulation fields. Stereotact Funct Neurosurg. 2016;94(6):363–70.
106.
Blomstedt P, Sandvik U, Fytagoridis A, Tisch S. The posterior subthalamic area in the treatment of movement disorders: past, present, and future. Neurosurgery. 2009;64(6):1029–38.
107.
Tripoliti E, Zrinzo L, Martinez-Torres I, Frost E, Pinto S, Foltynie T, et al. Effects of subthalamic stimulation on speech of consecutive patients with Parkinson disease. Neurology. 2011;76(1):80–6.
108.
Hamani C, McAndrews MP, Cohn M, Oh M, Zumsteg D, Shapiro CM, et al. Memory enhancement induced by hypothalamic/fornix deep brain stimulation. Ann Neurol. 2008;63(1):119–23.
109.
Laitinen LV. Leksell’s unpublished pallidotomies of 1958-1962. Stereotact Funct Neurosurg. 2000;74:1–10.
110.
Hariz MI, Krack P, Melvill R, Jorgensen JV, Hamel W, Hirabayashi H, et al. A quick and universal method for stereotactic visualization of the subthalamic nucleus before and after implantation of deep brain stimulation electrodes. Stereotact Funct Neurosurg. 2003;80(1–4):96–101.
111.
Halpern CH, Torres N, Hurtig HI, Wolf JA, Stephen J, Oh MY, et al. Expanding applications of deep brain stimulation: a potential therapeutic role in obesity and addiction management. Acta Neurochir. 2011;153(12):2293–306.
112.
Schiff ND, Giacino JT, Kalmar K, Victor JD, Baker K, Gerber M, et al. Behavioural improvements with thalamic stimulation after severe traumatic brain injury. Nature. 2007;448(7153):600–3.
113.
O’Gorman RL, Jarosz JM, Samuel M, Clough C, Selway RP, Ashkan K. CT/MR image fusion in the postoperative assessment of electrodes implanted for deep brain stimulation. Stereotact Funct Neurosurg. 2009;87(4):205–10.
114.
Coenen VA, Honey CR, Hurwitz T, Rahman AA, McMaster J, Bürgel U, et al. Medial forebrain bundle stimulation as a pathophysiological mechanism for hypomania in subthalamic nucleus deep brain stimulation for Parkinson’s disease. Neurosurgery. 2009;64(6):1106–14.
115.
Sammartino F, Krishna V, Sankar T, Fisico J, Kalia SK, Hodaie M, et al. 3-Tesla MRI in patients with fully implanted deep brain stimulation devices: a preliminary study in 10 patients. J Neurosurg. 2017;127(4):892–8.
116.
Rolston JD, Englot DJ, Starr PA, Larson PS. An unexpectedly high rate of revisions and removals in deep brain stimulation surgery: analysis of multiple databases. Parkinsonism Relat Disord. 2016;33:72–7.
117.
Ellis TM, Foote KD, Fernandez HH, Sudhyadhom A, Rodriguez RL, Zeilman P, et al. Reoperation for suboptimal outcomes after deep brain stimulation surgery. Neurosurgery. 2008;63(4):754–60.
118.
Tagliati M, Jankovic J, Pagan F, Susatia F, Isaias IU, Okun MS, et al. Safety of MRI in patients with implanted deep brain stimulation devices. NeuroImage. 2009;47(Suppl 2):T53–57.
119.
Zrinzo L, Hariz M, Hyam JA, Foltynie T, Limousin P. A paradigm shift toward MRI-guided and MRI-verified DBS surgery. J Neurosurg. 2016;124(4):1135–7.
120.
Kwoh YS, Hou J, Jonckheere EA, Hayati S. A robot with improved absolute positioning accuracy for CT guided stereotactic brain surgery. IEEE Trans Biomed Eng. 1988;35(2):153–60.
121.
Benabid AL, Cinquin P, Lavalle S, Le Bas JF, Demongeot J, de Rougemont J. Computer-driven robot for stereotactic surgery connected to CT scan and magnetic resonance imaging: technological design and preliminary results. Appl Neurophysiol. 1987;50(1–6):153–4.
122.
Nathoo N, Cavuşoğlu MC, Vogelbaum MA, Barnett GH. In touch with robotics: neurosurgery for the future. Neurosurgery. 2005;56(3):421–33.
123.
Liu L, Mariani SG, De Schlichting E, Grand S, Lefranc M, Seigneuret E, et al. Frameless ROSA® robot-assisted lead implantation for deep brain stimulation: technique and accuracy. Oper Neurosurg. 2020;19(1):57–64.
124.
Von Langsdorff D, Paquis P, Fontaine D. In vivo measurement of the frame-based application accuracy of the Neuromate neurosurgical robot. J Neurosurg. 2015;122(1):191–4.
125.
Schnitzler A, Mir P, Brodsky M, Verhagen L, Groppa S, Alvarez R, et al. Directional versus omnidirectional deep brain stimulation: results of a multi-cente prospective blinded crossover study. Parkinsonism Relat Disord. 2020;79:e16.
126.
Abbott. Data on file. PROGRESS largest study memo SJM-INF-0419-0314. n.d.
127.
Paff M, Loh A, Sarica C, Lozano AM, Fasano A. Update on current technologies for deep brain stimulation in Parkinson’s disease. J Mov Disord. 2020;13(3):185–98.
128.
Mishra A, Ramdhani RA. Directional deep brain stimulation in the treatment of Parkinson’s disease. Neurology. 2022;18(1):64.
129.
Hunsche S, Neudorfer C, Majdoub FE, Maarouf M, Sauner D. Determining the rotational orientation of directional deep brain stimulation leads employing flat-panel computed tomography. Oper Neurosurg. 2019;16(4):465–70.
130.
Mishra A, Unadkat P, McBriar JD, Schulder M, Ramdhani RA. An institutional experience of directional deep brain stimulation and review of the literature. Neuromodulation. 2023. In press.
131.
Rau A, Urbach H, Coenen VA, Egger K, Reinacher PC. Deep brain stimulation electrodes may rotate after implantation—an animal study. Neurosurg Rev. 2021;44(4):2349–53.
132.
Stewart CDM, Eljamel S. Prediction of implantable pulse generator longevity in deep brain stimulation: limitations and possible solutions in clinical practice. Stereotact Funct Neurosurg. 2011;89(5):299–304.
133.
Anheim M, Fraix V, Chabardès S, Krack P, Benabid AL, Pollak P. Lifetime of Itrel II pulse generators for subthalamic nucleus stimulation in Parkinson’s disease. Mov Disord. 2007;22(16):2436–9.
134.
Choi H, Gaiha R, Moeschler SM, Bendel MA, McCormick ZL, Teramoto M, et al. Factors associated with implantable pulse generator site pain: a multicenter cross-sectional study. Neuromodulation. 2021;24(8):1351–6.
135.
Bernstein JE, Kashyap S, Ray K, Ananda A. Infections in deep brain stimulator surgery. Cureus. 2019;11(8):e5440.
136.
Helmers A-K, Lübbing I, Birkenfeld F, Witt K, Synowitz M, Mehdorn HM, et al. Complications of impulse generator exchange surgery for deep brain stimulation: a single-center, retrospective study. World Neurosurg. 2018;113:e108–12.
137.
Hardaway FA, Raslan AM, Burchiel KJ. Deep brain stimulation-related infections: analysis of rates, timing, and seasonality. Neurosurgery. 2018;83(3):540–7.
138.
Abode-Iyamah KO, Chiang H-Y, Woodroffe RW, Park B, Jareczek FJ, Nagahama Y, et al. Deep brain stimulation hardware–related infections: 10-year experience at a single institution. J Neurosurg. 2019;130(2):629–38.
139.
Pepper J, Meliak L, Akram H, Hyam J, Milabo C, Candelario J, et al. Changing of the guard: reducing infection when replacing neural pacemakers. J Neurosurg. 2017;126(4):1165–72.
140.
Hagedorn JM, Canzanello N, Bendel MA, P Pittelkow T, J Lamer T. Antibacterial envelope use for the prevention of surgical site infection in spinal cord stimulator implantation surgery: a retrospective review of 52 cases. J Pain Res. 2021;14:2249–54.
141.
Thrane JF, Sunde NA, Bergholt B, Rosendal F. Increasing infection rate in multiple implanted pulse generator changes in movement disorder patients treated with deep brain stimulation. Stereotact Funct Neurosurg. 2014;92(6):360–4.
142.
Treu S, Strange B, Oxenford S, Neumann WJ, Kühn A, Li N, et al. Deep brain stimulation: imaging on a group level. NeuroImage. 2020;219:117018.
143.
Dembek TA, Roediger J, Horn A, Reker P, Oehrn C, Dafsari HS, et al. Probabilistic sweet spots predict motor outcome for deep brain stimulation in Parkinson disease. Ann Neurol. 2019;86(4):527–38.
144.
Boutet A, Germann J, Gwun D, Loh A, Elias GJB, Neudorfer C, et al. Sign-specific stimulation ‘hot’ and ‘cold’ spots in Parkinson’s disease validated with machine learning. Brain Commun. 2021;3(2):fcab027.
145.
Vorwerk J, Brock AA, Anderson DN, Rolston JD, Butson CR. A retrospective evaluation of automated optimization of deep brain stimulation parameters. J Neural Eng. 2019;16(6):064002064002.
146.
Horn A, Fox MD. Opportunities of connectomic neuromodulation. NeuroImage. 2020;221:117180.
147.
Li N, Baldermann JC, Kibleur A, Treu S, Akram H, Elias GJB, et al. A unified connectomic target for deep brain stimulation in obsessive-compulsive disorder. Nat Commun. 2020;11(1):3364.
148.
Joutsa J, Shih LC, Horn A, Reich MM, Wu O, Rost NS, et al. Identifying therapeutic targets from spontaneous beneficial brain lesions. Ann Neurol. 2018;84(1):153–7.
149.
Petersen MV, Mlakar J, Haber SN, Parent M, Smith Y, Strick PL, et al. Holographic reconstruction of axonal pathways in the human brain. Neuron. 2019;104(6):1056–64.e3.
150.
Horn A. The impact of modern-day neuroimaging on the field of deep brain stimulation. Curr Opin Neurol. 2019;32(4):511–20.
151.
Horn A, Ewert S, Alho EJL, Axer M, Heinsen H, Fonoff ET, et al. Teaching NeuroImages: in vivo visualization of Edinger comb and Wilson pencils. Neurology. 2019;92(14):e1663–4.
152.
Alho EJL, Alho ATDL, Horn A, Martin Md GM, Edlow BL, Fischl B, et al. The ansa subthalamica: a neglected fiber tract. Mov Disord. 2020;35(1):75–80.
153.
Ewert S, Horn A, Finkel F, Li N, Kühn AA, Herrington TM. Optimization and comparative evaluation of nonlinear deformation algorithms for atlas-based segmentation of DBS target nuclei. NeuroImage. 2019;184:586–98.
154.
Edlow BL, Mareyam A, Horn A, Polimeni JR, Witzel T, Tisdall MD, et al. 7 Tesla MRI of the ex vivo human brain at 100 micron resolution. Sci Data. 2019;6(1):244.
155.
Husch A, V Petersen M, Gemmar P, Goncalves J, Hertel F. PaCER: a fully automated method for electrode trajectory and contact reconstruction in deep brain stimulation. Neuroimage Clin. 2018;17:80–9.
156.
Dembek TA, Hoevels M, Hellerbach A, Horn A, Petry-Schmelzer JN, Borggrefe J, et al. Directional DBS leads show large deviations from their intended implantation orientation. Parkinsonism Relat Disord. 2019;67:117–21.
157.
Neudorfer C, Germann J, Elias GJB, Gramer R, Boutet A, Lozano AM. A high-resolution in vivo magnetic resonance imaging atlas of the human hypothalamic region. Sci Data. 2020;7:305–14.
158.
Elias GJB, Boutet A, Joel SE, Germann J, Gwun D, Neudorfer C, et al. Probabilistic mapping of deep brain stimulation: insights from 15 Years of therapy. Ann Neurol. 2021;89(3):426–43.
159.
Horn A, Reich M, Vorwerk J, Li N, Wenzel G, Fang Q, et al. Connectivity Predicts deep brain stimulation outcome in Parkinson disease. Ann Neurol. 2017;82(1):67–78.
160.
Al-Fatly B, Ewert S, Kübler D, Kroneberg D, Horn A, Kühn AA. Connectivity profile of thalamic deep brain stimulation to effectively treat essential tremor. Brain. 2019;142(10):3086–98.
161.
Irmen F, Horn A, Mosley P, Perry A, Petry-Schmelzer JN, Dafsari HS, et al. Left prefrontal connectivity links subthalamic stimulation with depressive symptoms. Ann Neurol. 2020;87(6):962–75.
162.
Neumann WJ, Schroll H, De Almeida Marcelino AL, Horn A, Ewert S, Irmen F, et al. Functional segregation of basal ganglia pathways in Parkinson’s disease. Brain. 2018;141(9):2655–69.
163.
Nowacki A, Nguyen TAK, Tinkhauser G, Petermann K, Debove I, Wiest R, et al. Accuracy of different three-dimensional subcortical human brain atlases for DBS –lead localisation. Neuroimage Clin. 2018;20:868–74.
164.
Rappel P, Grosberg S, Arkadir D, Linetsky E, Abu Snineh M, Bick AS, et al. Theta-alpha oscillations characterize emotional subregion in the human ventral subthalamic nucleus. Mov Disord. 2020;35(2):337–43.
165.
Kehnemouyi YM, Wilkins KB, Anidi CM, Anderson RW, Afzal MF, Bronte-Stewart HM. Modulation of beta bursts in subthalamic sensorimotor circuits predicts improvement in bradykinesia. Brain. 2021;144(2):473–86.
166.
Weaver KE, Caldwell DJ, Cronin JA, Kuo CH, Kogan M, Houston B, et al. Concurrent deep brain stimulation reduces the direct cortical stimulation necessary for motor output. Mov Disord. 2020;35(12):2348–53.