Developing a low-cost, highly configurable research platform

Neurostimulation is increasingly used to treat conditions such as Parkinson’s, epilepsy and chronic pain, either to replace or complement pharmaceutical approaches. New research into the use of neurostimulation to manage mood disorders, chronic drug-resistant inflammatory diseases and cognitive diseases such as Alzheimer’s is also gaining ground globally.

Currently, the cost and time to develop bespoke neuromodulation platforms in the early stages can present barriers to innovation. Our neurotechnology experts developed a platform to support neurostimulation research and development by dramatically reducing the cost and time to iterate new concepts. The platform architecture allows innovators to rapidly iterate, test and refine their design, significantly reducing early-stage investment.

Neurotech | Electronics and software

We developed this system to be highly adaptable, allowing us to customise it to meet the specific needs of our clients who are working on neuromodulation therapies. The miniaturised system is designed to be low cost and easy to implement, enabling a broad range of research use cases. By leveraging modern components and established design practices, we can significantly reduce the time and cost for our clients to introduce ever more advanced stimulation and sensing capabilities, without compromising on performance.
Stathis Louridas, Director of Systems, Team Consulting

Man looking at PCB CAD on computer screen

The current challenges in developing neuromodulation systems

New neuromodulation systems remain costly and time-consuming to develop due to their need for ultra-low power consumption and miniaturised electronic designs, often using custom ASICs to achieve the desired performance. This cost and complexity of development can introduce a barrier to innovation, where agility is required to investigate a concept and build confidence prior to investing in clinical use devices.

Our electronics and software experts set out to develop a platform capable of supporting the research and development of the next generation of stimulation devices across a broad range of clinical use cases, including instrumentation to enable closed loop stimulation and monitoring. To maximise the flexibility of the design to meet future research requirements, we designed the platform using high-performance, commercial components. This helped to avoid the need for costly ASIC development, allowing us to achieve an adaptable research platform in a short timeframe.

The result of 6 weeks of effort was a high-performance platform that combines low-power electronics with wireless configuration and data streaming, designed to enable innovation in closed loop neuromodulation and novel stimulation algorithms.

Identifying platform requirements and proven development techniques

The aim was to create a practical, adaptable research platform that could support a wide range of experimental use cases and evolve quickly as requirements change. We began by surveying current academic and clinical literature on neurostimulation and implantable neuromodulation devices. This enabled us to determine a set of research-driven requirements covering form factor, stimulation flexibility, sensing capability and system configurability. From this we identified the key platform requirements for the first iteration of the platform.

Platform requirements

Stimulation capability

Objectives: Support stimulation patterns commonly used in the treatment of chronic pain, movement disorders, and epilepsy. Ensure a broad capability to address multiple clinical research areas:

Deep Brain Stimulation (DBS)
Vagus Nerve Stimulation (VNS)
Responsive Nerve Stimulation (RNS)
Peripheral Nerve Stimulation (PNS)
Spinal Cord Stimulation (SCS)
Muscle Stimulation

Requirement: Support constant current stimulation with a high bandwidth. Be compatible with the most common existing stimulation patterns:

Monophasic
Biphasic (symmetric and asymmetric)
Sinusoidal

Sensing capability

Objective
Capable of real-time measurements to support monitoring and development of closed loop systems.

Requirements
Biopotential sensing circuit for implanted probes which meets the low noise requirements of EEG measurements

Research potential

Objective: Quickly and at a low cost adapt to emerging stimulation patterns and new technologies, and enable testing and development of new algorithms in-situ.

Requirements:

Secure Over the Air (OTA) software upgrade capability to expand firmware capability in-situ
Full configuration of stimulation profile via BLE
Use off-the-shelf components to reduce cost of development and enable low-volume manufacture

Clinical use-case

Objective: Allow for onward development as an implantable system and as a body-worn platform. Enable use for single point in time measurements, powered externally when implanted for muscle stimulation.

Requirements:

Miniaturised form factor suitable for a broad range of applications
Flexible connectivity for a variety of electrode options
Support for Near Field Communications (NFC) and measurements when powered exclusively through NFC
Architecture to consider safety and reliability

Connectivity

Objective: Meet configuration and monitoring requirements of implantable medical devices and enable multiple connectivity protocols to expand its use cases.

Requirements: The device should have BLE and NFC communication capability to allow for configuration and data streaming and to enable multiple clinical use cases.

Person holding smartphone in front of computer screens showing code, with a PCB research set up on the desk

A flexible architecture

Drawing on these requirements, we were able to develop a platform architecture based on BLE-capable implantable hardware and a mobile phone. The mobile phone could be used to communicate wirelessly to configure the device, view data in real-time and carry out OTA updates.

The implantable hardware needed to be power efficient and capable of arbitrary waveform generation and biopotential sensing. More specifically, the waveform generation was designed to be as configurable as possible to suit a wide variety of applications, from muscle stimulation to neuromodulation for proven established therapies, as well as experimental therapies that researchers may be theorising about now and in the future. Similarly, by adding biopotential sensing, we created an all-in-one platform that can both deliver modulation signals as well as measure them. This is both beneficial for researching a variety of waveforms as well as developing a closed loop system. The hardware was designed to work in tandem with bespoke software, which had to be similarly robust, configurable and expandable, providing users with flexibility, adaptability and customisation.

Hardware architecture design

The hardware architecture was designed to deliver versatile, high-quality constant-current neurostimulation, coupled with low noise biopotential sensing, in an ultra-compact form factor.

Instead of focusing on one therapy or waveform, the design prioritises precise amplitude control, accurate timing and a wide frequency range. This gives researchers the flexibility to test diverse stimulation approaches which previously may not have been possible, while ensuring reliable signal quality.

neurotransmitter_ar16x9_no-txt (1080p).00_00_00_00

Designing in flexible neurostimulation capabilities

We designed the platform with flexibility in mind, allowing it to support both established neuromodulation protocols and more exploratory research without the need for costly hardware redesign.

The stimulation architecture supports:

Constant-current output suitable for biological tissue
Monophasic stimulation
Biphasic stimulation (symmetric and asymmetric)
Configurable pulse widths and inter-phase delays
Adjustable stimulation frequency and session duration
Sinusoidal waveforms for exploratory and research use

Adaptable sensing and system integration

The platform integrates biopotential sensing hardware alongside stimulation, enabling researchers to measure signals while simultaneously testing novel neuromodulation patterns as well as providing the capability to develop closed loop modulation algorithms. The architecture was designed to minimise the interference between these functions, supporting the clean acquisition of biosignals during or around stimulation events.

The overall hardware was designed to be extensible, leveraging standard interfaces and off-the-shelf components to allow additional sensors or capabilities to be integrated with minimal change. Meanwhile, integrated wireless connectivity was added to support configuration, control and data streaming, removing the need for physical access and simplifying experimental workflows for rapid iteration.

Embedded software for repeatable data capture

The embedded firmware was engineered for precision, responsiveness and safety-first reliability. A modular, event-driven architecture enforces deterministic timing and predictable behaviour, helping ensure that any research data captured is repeatable, interpretable and free from software-induced variability.

Despite the rapid timeline, we followed established best practices for real-time systems and safety aware design when developing the software. This not only set a good foundation for onward development as a medical device in the future, it also created a more realistic design and helped to ensure the platform would remain reliable throughout its use in any research activities.

Real-time safety supervision

The platform incorporates key learnings from our expertise in medical systems development, including a dedicated, high-priority, safety subsystem to continuously monitor system health and subsystem activity. If operation deviates from the expected envelope, the firmware initiates a hardware-driven emergency stop, immediately halting stimulation and latching a fault state.

Safety logic runs at high cadence, is isolated from non-critical workloads, and acts as the final authority within the system.

Deterministic stimulation engine

To ensure consistent, repeatable stimulation sessions without reliance on software timing loops or thread scheduling, the stimulation delivery is handled by a deterministic waveform engine designed for microsecond-level accuracy:

Waveforms are validated and precomputed prior to delivery
Timing is governed by real-time counters and autonomous event routing
Buffered, hardware-triggered transfers eliminate jitter
The CPU remains focused on supervision and coordination

High-fidelity biosensing pipeline

A dedicated biosensing subsystem runs in parallel with stimulation, providing the app with clean, time-aligned biosignal data suitable for real-time visualisation and analysis:

Biopotential data is captured using event-driven, DMA-backed transfers
Artefact blanking and stimulation-aligned timing are applied
Data is packaged efficiently for wireless streaming
Clear separation is maintained between acquisition, processing and transmission

Status, reporting and system cohesion

To give users a clear, real-time view of device behaviour throughout experiments, a lightweight status service is used to aggregate system telemetry, including state, fault flags and runtime health, and publish updates wirelessly.

Well-defined subsystem boundaries, explicit contracts and disciplined thread prioritisation ensure the firmware remains predictable, testable and straightforward to extend the platform functionality in future developments.

Android test application development

To complement the embedded platform, we developed an Android application that provides the user with control, configuration and visualisation of both stimulation and biosensing data.

The application communicates with the device over Bluetooth Low Energy (BLE) using GATT services, providing a simple, structured interface aligned with the firmware’s command and telemetry model.

Test application features

Stimulation control

The app allows users to configure and deploy stimulation profiles, including:

Selection of monophasic, biphasic, or sinusoidal waveforms
Configuration of pulse durations, delays and frequency
Definition of stimulation session length
Start and stop control with real-time feedback

Configuration changes are validated by the device before execution, ensuring that stimulation is delivered as intended.

Biosensing visualisation

The application can initiate biosensing and receive streamed biopotential data in real time. Incoming data is graphed live, allowing users to observe biosignals and stimulation effects during experiments.

Built on time alignment and artefact indicators provided by the firmware, the application presents a clear and interpretable view of recorded signals to assist with research.

System feedback

The app receives periodic status updates from the research platform, including operational state and fault conditions. This ensures users always understand the current state of the system and can respond quickly during experimental work.

Continued improvement

We rapidly designed and developed a high-performance neurostimulation and instrumentation platform that can be quickly adapted to a wide range of research applications. Leveraging off-the-shelf components and established design patterns, the research platform allows new features and capabilities to be introduced more quickly and cost-effectively than most commercial platforms using fully custom ASICs.

Neuromodulation techniques are constantly evolving, with innovative stimulation technologies and clinical use-cases being a key focus. This work demonstrates that a coherent neuromodulation platform can be developed in a compressed timeline while adding forward thinking functionality to enable the development of the next generation of neuromodulation devices.
Thomas Watts, Senior Consultant - Advanced MedTech, Team Consulting

Explore

How could we help you?

Let’s start working together on your next challenge.

Get in touch