Student project: Resource Constraint Neural Signal Interpretation for Close-Loop Applications

*Important for non-EU students: You'll need to be registered at a Dutch university to meet immigration requirements.

This project aims to explore data analysis methods suitable for resource constraint hardware setups, employed to extract relevant information for enabling close-loop neuromodulation and BCI applications.

What you will do

Recent neuromodulation and brain-computer interface approaches emphasize the importance of devising bi-directional neural interfaces facilitating neural readout and stimulation towards close-loop applications. To improve performance of these approaches, deploying close-loop methodology within an implant or a wearable device is a necessity. Implementing close-loop in such small form-factor and low-power systems often imposes limitation in terms of signal processing and data interpretation pipelines. Hence, resource constraint methods are required to pre-process and clean neural data, compress it and/or extract relevant low- and high-level features, interpret those features towards understanding the neural status and/or the impact of stimulation, and use it to provide control and/or adapt the stimulation paradigms. Often, the execution of all these steps needs to be fast (millisecond level) such that the control commands or stimulation delivery can be done within the required time window. Application specific requirements might assist in reducing the complexity of the data processing pipeline and make the deployment of such close-loop solutions feasible, e.g., processing only short-duration data segments of evoked compound action potentials (eCAPs) to determine if stimulation amplitude needs to be adapted to activate desired neural fibers.

At imec, we have developed a new neuromodulation system capable of stimulating neural tissue in vivo and capturing neural response, hence facilitating closed-loop operation. This system supports low channel count (up to 64 channels) and has been in use to explore novel stimulation paradigms in simple animal models such as earthworms, but also in large animal models such as pigs. Furthermore, data analytics and software infrastructures have been developed, facilitating fast analysis and near real-time closed-loop operation. This project aims to investigate resource constrained data analysis pipelines required to facilitate close-loop neuromodulation / BCI applications. Several closed-loop use case scenarios will be explored, covering peripheral nerve interfaces capturing eCAPs and brain interfaces capturing local field potentials (LFPs) or Electrocorticography (ECoG) recordings towards extracting relevant neural activation parameters. The resource constrains will be defined by imec stimulation and sensing platform capabilities as well as application driven latencies in terms of required stimulation adaptation timings. Implementation of the proposed data pipeline on imec close-loop resource constrain platform will be explored towards realizing a demo setup.

The candidate will be involved in exploring suitable application use cases and requirements for close-loop data analysis implementation. The main contribution is expected in defining and implementing the data processing pipeline for the selected use cases and exploring its implementation within the resource constrained hardware setup. Execution of the project should lead to a library of data processing modules and a demo setup.

Student tasks will include:

• Literature review.
• Get acquainted with available stimulator and sensing system.
• Get acquainted with control-loop software framework.
• Get acquainted with current in-vivo experimental setup.
• Define several use cases and data processing pipeline specifications for those in the resource constraint settings.
• Develop a data processing module required for extracting relevant neurophysiological information.
• Explore integration of data processing modules within a resource constraint system available at imec.
• Implement a demo within the available imec platform for one of the selected use case.
• Report and documentation, depending on results, write a paper.

What we do for you

• You will be working on state-of-the-art technology and tools for low channel count stimulation and sensing of neural tissue, that can be deployed to improve efficacy of neuromodulation treatments in the healthcare domain or for BCI control.
• You will be working in an inspiring high-tech environment, located within the Holst Centre in Eindhoven, and part of the larger IMEC organization, world-leader in R&D on nanotechnology and electronics.
• You will receive support from experienced researchers having diverse background relevant for the execution of the project.
• You will be a member of our multi-disciplinary team of researchers, engineers and innovators, and will be offered an opportunity to contribute to our ambitious aims in making real impact on actual healthcare needs.

Who you are

• Excellent MSc student in Computer Science, Biomedical Engineering, Electrical Engineering, or equivalent.
• You are entitled to do an internship in The Netherlands (have EU nationality and/or currently study at Dutch University).
• Available for 9 months or longer.
• Have good signal processing skills, in particular time-series data analysis.
• Have good programming skills in python/Matlab.
• Having experience with microprocessor programming is a plus.
• Know-how in control systems is a plus.
• Eager to take ownership for your student project.
• Have a structured way of working.
• Have good command of spoken and written English.

Interested

Does this position sound like an interesting next step in your career at imec? Don’t hesitate to submit your application by clicking on ‘APPLY’.
Should you have more questions about the job, you can contact jobs@imec.nl.