Decoding of Intraneural Recordings in transfemoral amputees using Neuromorphic Computing (in collaboration with The University of Chicago)

Direct interfaces to peripheral nerves provide accurate sensory and motor information on the nervous system. This information can be exploited to decode the motor intent and restore motor functions to amputees. Recent efforts have been made to decipher the motor commands from electroneurographic (ENG) signals recorded using invasive interfaces implanted in the peripheral nerves [1]. However, these signals are very noisy and difficult to interpret, and the standard approach relies on conventional preprocessing and classifiers (e.g SVM, LDA).

In this project, transverse intrafascicular multi-channel electrodes (TIMEs) [2] are implanted in two transfemoral amputees to record ENG signals while the participants performed 3 different tasks: flexion and extension of the knee, flexion and extension of the ankle, and flexion and extension of the fingers.

In this project, the student will contribute to the overall research goal with the following tasks:

  • - Define a new preprocessing to mitigate the noise in the signal and maximize the signal quality.

  • - Perform signal processing using conventional methods (e.g. PCA, SVM, linear classifier) to define the ground truth for the spiking implementation

  • - Implement a spiking neural network (in simulation or Dynapse chip)

  • - Implement a learning algorithm for SNN to discriminate between the movements

    • - - Review of existing literature on supervised spiking-based learning rules

    • - - Implement the learning rule in simulation (using Brian2 or Pytorch) and compare the performance to conventional SVM classification.

    • - - Map the simulated network on neuromorphic hardware (i.e DYNAPSE).

In this project, we will investigate a unique unpublished dataset, and we will combine knowledge on neurorobotics, signal processing, and neuromorphic computing.


[1] Cracchiolo, M., Valle, G., Petrini, F., Strauss, I., Granata, G., Stieglitz, T., ... & Micera, S. (2020). Decoding of grasping tasks from intraneural recordings in a trans-radial amputee. Journal of neural engineering, 17(2), 026034.

[2] I. Strauss et al. “Characterization of multi-channel intraneural stimulation in transradial amputees”. In: Scientific Reports9.1 (2019)


  • - Spiking Simulator, Brain2

  • - A mixed-signal neuromorphic chip, the DYNAPSEx

  • - Anonymized ENG dataset from 3 patients in multiple sessions and repetitions

  • - Code for extracting the Ripple LLC data

  • - Manual of the neural recorder and. of the front-end and the neural recordings


Semester project or master thesis


Programming skills in Python, 
Interest in neuronal systems (biological or artificial).


INI: Elisa Donati elisa (at), and Farah Baracat
UChicago: Giacomo Valle

© 2023 Institut für Neuroinformatik