Personal tools

Deep neural networks for seizure detection

From iis-projects

Revision as of 14:51, 22 June 2020 by Xiaywang (talk | contribs)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search
Non-EEG Seizure.jpg

Description

Epilepsy is one of the most prevalent chronic neurological disorders. One-third of patients with epilepsy continue to suffer from seizures despite pharmacological therapy. For these patients with drug-resistant epilepsy, efficient algorithms for seizure detection are needed in particular during pre-surgical monitoring. Many efforts have been pursued in this direction with the fabrication of many ASIC and the development of advanced machine/deep-learning to optimize both the energy efficiency for years-operating devices and the accuracy in the epilepsy detection. In terms of time-series analysis, a big variety of deep-learning approaches are arising for efficient processing such as InceptionTime [1], MultiScale-CNN [2], and Temporal Convolutional Networks (TCN) [3].

The thesis would be a 6-month full-time project with the following steps to accomplish:

1 - Development in a high-level programming language (python) of different deep learning algorithms for time series classification. In particular, the initial targets will be the InceptionTime, the MultiScale-CNN, the Temporal Convolutional Networks, and a bidirectional LSTM.

2 - Benchmarking of these algorithms on a large - scale dataset collected by the Bern Inelspital about epileptic patients [4].

3 - Comparison with state of the art methods (Local Binary pattern + Hyperdimensional computing [5], Short-time Fourier transform + Convolutional Neural Networks and classical machine learning methods).

4 - Characterization of the algorithm on different computing platforms, from the high-level number of operations to the number of cycles and energy consumption on embedded devices (e.g. GAP8, a Greenwaves multi-core chip).


Status: Available

Looking for one student for Master's thesis.

Supervision: Michael Hersche, Xiaying Wang, Alessio Burrello

Prerequisites

  • Machine Learning
  • Python & C Programming


Character

20% Theory
80% Programming

Professor

Luca Benini

↑ top


Literature

  • H. I. Fawaz et al., InceptionTime: Finding AlexNet for Time Series Classification [1]
  • Z. Cui et al., Multi-Scale Convolutional Neural Networks for Time Series Classification [2]
  • C. Lea et al., Temporal Convolutional Networks: A Unified Approach to Action Segmentation, [3]
  • iEEG-SWEZ data base [4]
  • A. Burello et al., Laelaps: An Energy-Efficient Seizure Detection Algorithm from Long-term Human iEEG Recordings without False Alarms [5]


Practical Details

↑ top