Stereotactic planning and placement of a deep brain stimulation electrode into the subthalamic nucleus for Parkinson's disease
Deep brain stimulation (DBS) begins with converting a patient's anatomy into a precise, reproducible coordinate space. A rigid stereotactic frame — or, increasingly, a frameless robotic platform — is fixed to the skull, and high-resolution MRI fused with a stereotactic CT is used to compute x, y, z coordinates for the target nucleus relative to fixed internal landmarks.
DBS is an established, reversible, adjustable neuromodulation therapy for movement disorders refractory to optimized medical management, and its indications continue to expand:
• Parkinson's disease — motor fluctuations, dyskinesia, and medication-refractory tremor in patients with preserved levodopa responsiveness; the dominant target is the subthalamic nucleus (STN), with the globus pallidus internus (GPi) as an alternative • Essential tremor — disabling action tremor unresponsive to propranolol/primidone; target is the ventral intermediate nucleus (VIM) of the thalamus • Dystonia — generalized or segmental dystonia, including DYT1-positive cases; target is typically GPi • Emerging indications — treatment-resistant epilepsy (anterior nucleus of thalamus), obsessive-compulsive disorder, and treatment-resistant depression under investigational protocols
Patient selection requires a multidisciplinary team: neurologist, neurosurgeon, neuropsychologist, and psychiatrist evaluate diagnosis certainty, levodopa responsiveness, cognitive status, and realistic expectations before surgery is offered.
Stereotactic surgery relies on a Cartesian coordinate system anchored to reproducible intracranial landmarks:
• AC-PC line — the line connecting the anterior commissure (AC) and posterior commissure (PC), visible on midsagittal MRI; it defines the anteroposterior axis and the origin (midcommissural point, MCP) for target coordinates • Frame-based systems — a Leksell or Cosman-Roberts-Wells (CRW) frame is rigidly pinned to the outer table of the skull under local anesthesia, providing a fixed external reference grid that remains valid throughout the procedure • Frameless / robotic systems — skull-mounted or bone-fiducial-based platforms (e.g., ROSA, Neuromate) achieve comparable accuracy with improved patient comfort and faster room turnover • Image fusion — a volumetric stereotactic MRI (for soft-tissue nucleus visualization) is co-registered with a stereotactic CT (which is free of the geometric distortion inherent to MRI) to compute the final target coordinates
Indirect targeting uses standardized atlas offsets from the AC-PC midpoint (e.g., STN ≈ 11–13 mm lateral, 2–3 mm posterior, 3–5 mm inferior to MCP), while direct targeting visualizes the STN itself on high-field or susceptibility-weighted MRI. Modern practice combines both, then refines the plan with intraoperative microelectrode recording.
A target coordinate alone is not a surgical plan. Planning software overlays the proposed needle path onto the patient's fused imaging, from a frontal burr hole entry point down through cortex and white matter to the target, while a risk map of nearby vessels, sulci, and the ventricular system is used to steer the trajectory into a safe corridor.
The ideal entry point balances a short, straight trajectory with maximal clearance from critical structures:
• Located anterior to the coronal suture, roughly 2.5–3.5 cm from the midline, to achieve an oblique angle that traverses the middle frontal gyrus • Avoids traversing a sulcus directly — sulcal vessels are numerous and difficult to visualize on standard imaging • Avoids the frontal horn of the lateral ventricle, since a ventricular breach causes CSF egress, brain shift, and loss of targeting accuracy ("brain sag") • Trajectory angle is chosen so a single pass can reach the STN, red nucleus, and substantia nigra in sequence — useful electrophysiological landmarks during MER
Modern planning platforms (e.g., Surgical Theater, Brainlab Elements, StealthStation) automatically segment vasculature from a contrast-enhanced or susceptibility-weighted sequence and score candidate trajectories:
• Vessel proximity — trajectories are penalized if they pass within ~2 mm of any segmented vessel ≥1 mm diameter; hemorrhage along the electrode track is the most serious acute surgical complication of DBS • Ventricular margin — a minimum clearance (commonly >3 mm) from the lateral ventricle wall reduces the risk of CSF loss and pneumocephalus-related brain shift • Sulcal and gyral crossing — the path should cross the fewest possible sulci and avoid eloquent cortex • Multiple parallel trajectories (a "Ben-gun" microdrive array, typically 3–5 channels spaced 2 mm apart) may be planned simultaneously so several candidate tracks can be sampled electrophysiologically in one pass
Symptomatic intracranial hemorrhage occurs in roughly 1–3% of DBS lead placements and is the leading driver of permanent morbidity from the procedure — which is why trajectory risk-scoring against a segmented vascular map is now standard of care at most centers.
Imaging defines an anatomical target, but the STN is only 3–5 mm wide and its exact borders vary between patients and can shift with brain pulsation or micro-movement. Microelectrode recording (MER) advances a high-impedance recording electrode in small steps along the planned trajectory, listening to extracellular action potentials that reveal — in real time — exactly which nucleus the tip is in.
A tungsten or platinum-iridium microelectrode with an exposed tip of only a few microns (impedance ~0.5–1.0 MΩ) is advanced by a precision microdrive in 0.5–1.0 mm steps along the planned trajectory, starting several millimeters above the anticipated target:
• Signal is amplified, filtered (typically 300 Hz–5 kHz for spikes), and played through an audio speaker — experienced neurophysiologists can identify nucleus transitions by ear • At each depth, background noise, spike amplitude, firing rate, and firing pattern (regular, irregular, bursting) are recorded and logged against depth • Kinesthetic testing — passive limb movement while recording — identifies "movement-responsive" cells characteristic of the sensorimotor STN, confirming not just which nucleus but which functional territory within it
As the electrode descends from thalamus toward the substantia nigra, it crosses a predictable sequence of neurophysiological zones:
• Thalamus / zona incerta — sparse, low-frequency (~10–15 Hz), relatively quiet background • Dorsal STN border — an abrupt increase in background noise and a step up in firing rate as the electrode enters the nucleus • STN body — dense, irregular, high-frequency discharge (25–45 Hz) with bursting patterns; many cells respond to passive joint movement (kinesthetic cells), and beta-band (13–30 Hz) oscillatory activity correlates with parkinsonian rigidity/bradykinesia • Ventral STN border / substantia nigra pars reticulata (SNr) — an abrupt transition to more regular, higher-frequency (60–90 Hz) tonic discharge marks the exit from STN
Mapping these transitions across 3–5 parallel tracks builds a 3-D electrophysiological "atlas" of the target for that individual patient, which is used to select the single best track for permanent lead placement.
The STN sensorimotor territory targeted for Parkinson's disease is only a few millimeters across, nested within a nucleus that itself is 3–5 mm in its longest dimension — MER routinely refines the final lead position by 1–3 mm relative to the pure imaging-based plan.
Once MER confirms the target, a temporary test lead (or the permanent lead itself) delivers pulsed current while the awake patient is examined live. The surgical team increases stimulation amplitude step-wise, watching tremor and rigidity improve while simultaneously screening for side effects — searching for the widest possible gap between the voltage that helps and the voltage that hurts.
Test stimulation uses the same three programmable parameters as chronic therapy — amplitude (voltage or current), pulse width, and frequency:
• High-frequency stimulation (typically 130–185 Hz) is thought to work by a combination of local depolarization block, synaptic depression, and disruption of pathological oscillatory synchrony (notably beta-band activity) in the basal ganglia-thalamocortical motor loop, rather than simple "lesioning" • At each candidate contact and depth, amplitude is raised gradually from 0 V while the patient performs tasks such as finger tapping, arm rigidity checks (cogwheeling), and rest-tremor observation • Onset voltage for benefit and onset voltage for the first side effect are both recorded, defining a therapeutic window for that contact
Side effects during test stimulation are highly informative — each corresponds to spread of current into a specific neighboring structure, and effectively confirms lead position:
• Internal capsule (posterolateral to STN) — tonic muscle contraction of the contralateral face/arm/leg, appearing at low, reproducible voltage as current spreads to corticospinal fibers • Oculomotor tract / medial STN — conjugate eye deviation or diplopia • Medial lemniscus (posterior to STN) — paresthesias in the contralateral limb • Substantia nigra / limbic circuits — mood changes, acute dysarthria, or postural instability if current spreads too far ventrally or medially
A good target and trajectory typically show benefit onset around 1–2 V with side effects not appearing until well beyond that (often 3–4 V or more) — a wide therapeutic window. A narrow window (side effects appearing just above the benefit threshold) suggests the lead should be repositioned before permanent placement.
Directional (segmented) leads allow current to be steered away from a side-effect-generating structure without moving the electrode at all — widening a marginal therapeutic window by shaping the electric field toward the STN and away from the internal capsule or lemniscus.
Once the optimal track and depth are confirmed, the temporary electrode is exchanged for a permanent quadripolar (or eight-contact directional) DBS lead, anchored at the skull. Days to weeks later, the lead is connected via a subcutaneously tunneled extension wire to an implantable pulse generator (IPG) placed under the skin of the chest — completing a closed, programmable neurostimulation system.
The permanent DBS lead is a flexible, insulated wire roughly 1.27 mm in diameter carrying four (conventional) or eight (directional/segmented) platinum-iridium contacts spanning the dorsoventral extent of the STN:
• Conventional leads have four full-ring contacts; directional leads split one or more rings into 2–3 independent segments arranged radially, allowing the electric field to be shaped toward the target and away from side-effect structures • The lead is advanced to the depth confirmed by MER and test stimulation, then secured to the skull with a burr-hole cap or anchoring device to prevent migration • Intraoperative fluoroscopy or O-arm CT confirms final lead position before the scalp is closed; a post-operative CT fused to the pre-operative MRI provides the definitive documented lead location
In the same or a staged second operation, an extension wire is tunneled subcutaneously from the scalp incision, behind the ear, down the neck, and over the clavicle to a subcutaneous or subfascial pocket — most commonly in the infraclavicular chest, occasionally the abdomen:
• The extension connects to the implantable pulse generator (IPG), a battery-powered computer roughly the size of a small stopwatch that generates the programmed pulse train • Bilateral procedures use either two IPGs or a single dual-channel device • The system is entirely internal — no wires or components cross the skin — minimizing infection risk during years of use
Two IPG battery technologies are in clinical use:
• Primary-cell (non-rechargeable) IPGs last roughly 3–5 years depending on stimulation settings, then require a simple outpatient replacement surgery • Rechargeable IPGs use an internal lithium-ion cell recharged transcutaneously via an external wand roughly weekly, extending device life to 15 years or more and reducing the number of replacement surgeries over a patient's lifetime
Postoperatively, the system is programmed non-invasively with a wireless clinician tablet: each contact's amplitude, pulse width, and frequency can be adjusted independently, and directional leads allow "current steering" — activating individual segments to sculpt the stimulation field toward the sensorimotor STN while avoiding the capsule and lemniscus identified during test stimulation. Programming is refined over weeks to months as the initial micro-lesion effect (a transient benefit from the insertion trauma itself) wears off and chronic stimulation settings are optimized.
The completed system is a closed loop only in the broad sense of therapy delivery — most implanted DBS systems today are open-loop (fixed settings adjusted at clinic visits), though newer sensing-enabled IPGs can record local field potentials (e.g., beta-band power) from the same lead and begin to adapt stimulation to real-time brain signals.