Closed-Loop Adaptive DBS Current Control 2026
Closed-Loop Adaptive DBS Current Control
Closed-loop adaptive DBS systems replace static stimulation with feedback-driven current adjustment using LFP biomarkers and adaptive algorithms. This dataset spans filings from 2010 to 2026 across commercial and academic assignees.
From Fixed Stimulation to Feedback-Driven Current Control
Closed-loop adaptive DBS systems dynamically adjust stimulation parameters — current amplitude, frequency, pulse width, and waveform shape — in response to measured physiological signals. The core architecture spans three functional stages: sensing (LFP recording, evoked potentials, neurochemical sensing, kinematic sensors), signal processing and classification, and stimulation parameter adjustment.
Among retrieved results, the dataset spans filings and publications from 2010 to 2026, revealing a three-phase maturation arc. The foundational phase (2010–2014) established Q-learning and neurochemical feedback architectures. The development phase (2015–2020) saw brain network model-based approaches and surging literature output from the DBS Think Tank community.
The clinical convergence phase (2021–2026) brought evoked potential-based adaptive DBS from Boston Scientific, accelerating Chinese academic filings from Fudan University and Tianjin University, and multidisciplinary wearable-DBS integration approaches. The most recent filings incorporate directional multi-contact current feedback and kinematic sensor fusion as alternative feedback channels.
In this dataset, US-jurisdiction filings dominate the commercial assignee portfolio. Boston Scientific Neuromodulation holds the most commercially current active patent position with a 2026-dated US grant, while Chinese academic filings in retrieved records represent the fastest-growing segment by recency, with Fudan University and Tianjin University as primary filers.
Technology Cluster Distribution and Filing Timeline
Across retrieved patent records, four distinct technology clusters define the closed-loop adaptive DBS landscape. Filing activity has shifted from foundational RL-based control (2010–2014) toward biomarker-driven and network-model architectures (2016–2023) and, most recently, directional current feedback and wearable sensor fusion (2024–2026).
Patent Filings by Technology Cluster — Closed-Loop Adaptive DBS (Dataset Snapshot)
In this dataset, biomarker-driven threshold and phase-responsive control accounts for the largest cluster with the most commercially active assignees, followed by reinforcement learning and model-based adaptive control, brain network modeling, and load-adaptive hardware.
↗ Click bars to exploreClosed-Loop DBS Patent Filing Activity by Phase — Dataset Snapshot
In this dataset, filing activity accelerated markedly in the 2021–2026 clinical convergence phase, with the most recent records (2024–2026) driven by Chinese academic institutions and a pending US commercial filing, reflecting a geographic shift in originating jurisdiction.
↗ Click bars to exploreClinical Target Areas for Closed-Loop Adaptive DBS Systems
Across retrieved records, closed-loop adaptive DBS architectures have been developed and validated across at least five distinct clinical application domains, each associated with specific feedback biomarkers, stimulation targets, and hardware requirements.
Parkinson’s Disease and Movement Disorders
Beta-band LFP suppression in the subthalamic nucleus is the primary feedback signal in this dataset. Literature records document beta-driven closed-loop DBS effects on motor behavior (2019), adaptive PID/neural network control of GPi beta oscillations (2021), and dual-threshold control policies customized to individual therapeutic windows. Boston Scientific’s evoked potential patents specifically reference Parkinson’s disease network activation as a control target.
Neural BiomarkerEssential Tremor Closed-Loop DBS
Multiple records address essential tremor (ET) as a distinct application domain. The fully implanted adaptive DBS study (2020) recruited ET patients with chronically implanted electrocorticography strips, and the University of Colorado ANN-based DBS patent (2022) covers both Parkinson’s disease and ET. Fudan University’s adaptive closed-loop DBS method (2022, CN active) is also directed at multi-state tremor control.
In-situ Neural RecordingNeuropsychiatric Disorders: OCD, Tourette, Depression
The DBS Think Tank proceedings (4th–7th annual meetings, 2016–2020) and literature on algorithmic closed-loop DBS reference obsessive-compulsive disorder, Tourette syndrome, and depression as active application areas. STN beta and high-frequency oscillation coupling has been identified as a biomarker candidate for OCD in closed-loop control contexts.
Neuropsychiatric DBSSleep Disorders and Obesity DBS
The University of Colorado ANN patent (2022, US active) addresses sleep-stage-adaptive DBS to improve sleep in Parkinson’s disease patients using a feedforward neural network trained on subthalamic nucleus LFPs. Oregon Health & Science University filed a patent (2018, US active) on DBS electrode stimulation controlled by brown adipose tissue temperature feedback for obesity treatment, targeting energy-efficient closed-loop current control.
Metabolic and Sleep DBSLeading Patent Assignees in Closed-Loop Adaptive DBS — Dataset Snapshot
In this dataset, Boston Scientific Neuromodulation Corporation holds the most commercially current active filing position with 3 active patents including a 2026-dated US grant, while Battelle Memorial Institute accounts for 3 active filings in retrieved records covering brain network model-based closed-loop DBS. EPFL represents the most prolific academic assignee in retrieved records with 5 active or pending patents on adaptive closed-loop neuromodulation.
Top Assignees by Filing Count — Closed-Loop Adaptive DBS (Dataset Snapshot, Retrieved Records)
↗ Click bars to exploreBoston Scientific Neuromodulation Corp.
Boston Scientific Neuromodulation holds 3 active patents in this dataset on evoked potential-based adaptive DBS, spanning a 2023 US grant, a 2023 WO application, and a 2026-dated US active grant — the most recent commercial filing in retrieved records. These patents cover using evoked potentials to model network activation and maintain it within predetermined therapeutic ranges via a control algorithm. All three family members are currently active.
United StatesBattelle Memorial Institute
Battelle Memorial Institute holds 3 active patents in this dataset covering brain network model-based closed-loop DBS, including a 2016 WO international priority filing and two 2020 US active grants. These patents describe a software-based brain network model that estimates unmeasured neural signals and continuously updates stimulation parameters by comparing estimated to actual brain recordings. All three family members are currently active.
United StatesFour Frontier Directions in Closed-Loop DBS (2023–2026)
Based on the most recent filings (2023–2026) in this dataset, four directions are apparent: evoked potential-driven network activation control, multi-dimensional directional contact current feedback, kinematic and behavioral sensor fusion, and multidisciplinary design optimization for wearable-DBS integration.
Evoked Potential-Driven Network Activation Control
Boston Scientific’s active 2026 US patent on adaptive DBS based on neural signals with dynamics, and its WO/US 2023 counterparts, represent the leading commercial direction in retrieved records. These patents use evoked potentials to model whole-network activation and maintain it within patient-specific therapeutic ranges, moving beyond simple beta threshold detection toward network-level current titration. This is among the most commercially advanced patent families in the dataset.
Kinematic Sensor Fusion Bypassing Implanted Recording
Fudan University’s pending 2025 CN filing on closed-loop DBS control introduces wearable sensors placed on fingertips, wrist, and legs to capture kinematic and dynamic signals during Parkinson’s disease motor paradigms as the feedback signal, bypassing the need for implanted recording electrodes. This represents a significant hardware simplification pathway for closed-loop DBS that reduces implant complexity while maintaining feedback-driven current adjustment.
Open-Loop vs. Closed-Loop Adaptive DBS: Key Dimensions
Click any row to explore further.
| Dimension | Open-Loop DBS | Closed-Loop Adaptive DBS |
|---|---|---|
| Stimulation Mode | Fixed, continuously-on parameters | Dynamically adjusted based on feedback signals |
| Feedback Mechanism | None — no physiological sensing | LFP, evoked potentials, neurochemical sensors, kinematic sensors |
| Control Algorithm | Manual clinician programming | PID, Q-learning, neural networks, Bayesian optimization, brain network models |
| Adaptation to Impedance | Not addressed in real time | Load-adaptive feedback module detects real-time electrical status and stabilizes output (National Chiao Tung University, 2012) |
| Electrode Architecture | Single or multi-contact, fixed current distribution | Multi-contact directional current steering with per-contact amplitude, frequency, and spatial feedback (Tianjin University, 2024) |
| Clinical Evidence Base | Decades of clinical use across approved indications | Growing literature including fully implanted adaptive DBS in ET patients (2020) and beta-driven closed-loop studies (2019, 2021) |
| Commercial IP Status | Established; core patents largely expired | Active commercial IP from Boston Scientific (2026 US grant), Medtronic (2022–2023 US active), Battelle (2020 US active) |
| Neuropsychiatric Applications | Limited biomarker specificity | STN beta/high-frequency coupling biomarker candidates for OCD, Tourette, depression (DBS Think Tank 2016–2020) |
Frequently Asked Questions: Closed-Loop Adaptive DBS Patents
According to retrieved records, beta-band local field potential (LFP) suppression in the subthalamic nucleus (STN) is the primary feedback signal used for Parkinson’s disease. Literature records document beta-driven closed-loop DBS effects on motor behavior (2019) and adaptive PID/neural network control of GPi beta oscillations (2021).
Boston Scientific Neuromodulation Corporation holds a 2026-dated US active grant on adaptive deep brain stimulation based on neural signals with dynamics, making it the most commercially current active filing in this dataset, alongside companion 2023 US and WO patents.
No. According to the dataset, all three Sorin CRM SAS family members covering optimal DBS therapy with Q-learning (WO 2010, US 2012, EP 2012, US 2014) are now fully inactive, indicating lapsed commercial protection. The reinforcement learning-based parameter optimization design space is therefore open to new entrants without licensing obligations to the original filer.
Medtronic’s two US active patents (2022 and 2023) cover systems that can autonomously reconfigure their own closed-loop control system architecture — not just stimulation parameters — in response to internal performance metrics or external subsystem inputs. This is described as a meta-adaptive layer on top of standard biomarker-driven control, relevant to long-term chronic implant reliability.
Fudan University’s pending 2025 CN filing uses wearable sensors placed on fingertips, wrist, and legs to capture kinematic and dynamic signals during Parkinson’s disease motor paradigms as the closed-loop feedback signal. This bypasses the need for implanted recording electrodes for the feedback channel, representing a hardware simplification compared to LFP-based approaches.
According to the strategic implications section, load-adaptive current stabilization at the electrode-tissue interface — originally addressed by National Chiao Tung University (now lapsed) — remains an underserved area in recent patent filings relative to the algorithmic control layer. Hardware-focused teams are identified as having a white-space opportunity in current-regulation ASIC design for chronic implants subject to impedance drift from glial scarring.
Data and insights on this page are based on a limited patent and literature dataset and are for reference only. Figures may not represent the complete technology landscape.