Student project: neuromorphic spiking neural network for neural decoding

At IMEC (Thesis)

We will develop a neuromorphic hardware system based on a spiking neural network for decoding neural signals.


What you will do
The nervous system of the body uses "spikes" for communicating between neurons across the human body. For example, neurons in the brain (Central Nerve System, CNS) send action potentials with specific temporal patterns to the neurons in our limbs (Peripheral Nerve System, PNS) to control body movements via our spinal cord. However, the nerve connection in paralyzed patients might be damaged, e.g., due to a spinal cord injury. Neuroscientists develop neural decoders to translate specific temporal patterns of the spike signals from the CNS into commands that can control prosthetic limbs or perform muscle nerve stimulation. However, these kinds of neural decoders typically consume high power, which requires a significant power source or battery, so they are challenging to fit into a wearable form factor or even implantable.


Unlike conventional artificial neural networks, neuromorphic spiking neural networks can perform pattern recognition based on very sparse neural signals in an event-based manner. Its power consumption can be therefore extremely low, making it suitable for implantable applications.
In this project, the student will study spike sorting approaches with a spiking neural network. Spike sorting algorithms use the shapes of the waveforms collected with one or more electrodes to distinguish the activity of one or more neurons from background noise. The spiking neural network will be first developed in software, but it will need to be compatible with embedded requirements of IMEC's neuromorphic digital spiking neural networks architectures.


The tasks in this project include:

  • Literature study.
  • Neural decoding requirement study (spike sorting, decoding, and spiking neural networks for real-time biomedical signal analysis).
  • Spiking neural network software (Python, C/C++).
  • Hardware design and verification in Field Programmable Gate Arrays (FPGA). This task will exploit IMEC's already available neuromorphic digital architectures, modified and tweaked to fit the biomedical spike sorting task.

What we do for you
As the world-best research institute, we are the center of excellence in nano-electronics design for the Internet of Things and Healthcare applications. In this internship project, you will be working on a cutting-edge research project under the supervision of world-renowned researchers from multidisciplinary backgrounds.


Who you are

  • You are a Master student in Electrical Engineering or Artificial Intelligence background with good analytical skills.
  • You are available for a period of 9 months.
  • Knowledge of neural networks is preferred (experience with TensorFlow, PyTorch, or similar).
  • Strong interest in biomedical applications, especially in neural interfacing.
  • Willing to learn the basics of hardware description languages (VHDL). VHDL will be used to verify the spiking neural network functionalities in digital hardware. At IMEC, we have strong knowledge of digital hardware design, and students are supposed to learn its basics. Previous experience with VHDL or Verilog is a strong plus.
  • Familiar with real-time signal processing.
  • Good written and verbal English skills.
  • Entitled to do an internship in the Netherlands.
  • Motivated student eager to work independently and expand knowledge in the field.


Click on 'apply' to submit your application. You will then be redirected to e-recruiting.