Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
Machine Learning, Virtual Reality and Wearable devices to understand emotional and cognitive contributions to pain perception
This project aims to evaluate how emotional and cognitive contributions, introduced using different virtual reality scenarios, change the subjective pain perception. We aim to conduct a study comprising Transcutaneous Electrical Nerve Stimulation (TENS) and Virtual Reality (VR). TENS will be used to induce short painful stimuli while VR will be used to modulate the cognitive and emotional inputs by presenting different inputs (e.g., fire/electricity vs water stimuli on the virtual limb). Several physiological measures (EEG and multisensory wearable) will be collected together with the subjective pain perception. Machine Learning will be applied to physiological data, to disentangle the physical reaction to pain with respect to psychological components.
Keywords: Pain; Machine Learning; Signal Processing; Virtual Reality; EEG;
Chronic pain is a highly prevalent medical condition and affects around 20% of the general adult population. Pain is a subjective and not fully understood experience typically measured through patient’s self-report. Pain has also emotional and cognitive components that impact on its subjective experience. Nowadays there is no reliable and universal model of pain prediction that considers all its multifaceted aspects. It is necessary to understand if the perceived pain is more physical or psychological to propose personalized therapies that are more efficient in the long term, tackling the opioid crisis and decreasing the cost on healthcare and society.
Chronic pain is a highly prevalent medical condition and affects around 20% of the general adult population. Pain is a subjective and not fully understood experience typically measured through patient’s self-report. Pain has also emotional and cognitive components that impact on its subjective experience. Nowadays there is no reliable and universal model of pain prediction that considers all its multifaceted aspects. It is necessary to understand if the perceived pain is more physical or psychological to propose personalized therapies that are more efficient in the long term, tackling the opioid crisis and decreasing the cost on healthcare and society.
The student will be guided in understanding the principal cause of chronic pain, its effects and meaning in terms of reduction of quality of life and consequences on the healthcare system. They will also be introduced to state of the art of pain perception and AI with scientific literature readings. In the project, starting from models already developed for a previously collected offline dataset, the aim is to discover how different emotional and cognitive inputs alters the subjective perception of pain.
The major goals for the student will be:
1) Finalization of a virtual reality environment to induce different emotional and cognitive inputs to alter the pain perception.
2) Design of a Machine Learning pipeline that will include: Signal filtering and processing, extraction of the most relevant features, development of new machine learning models to find the most meaningful biomarkers in the EEG and wearable physiological signal
3) Testing on Healthy subjects and data collection
4) Analysis and evaluation of the dynamic of subjective bias in the different conditions
**Requested skills**: Knowledge of Python, knowledge of AI and Machine Learning algorithms, Signal Processing and Statistical Analysis, knowledge of basic physiological signals (e.g., EEG). Good Programming skills. Highly motivated.
**Extra skills**: Knowledge of Unity, neuroscience of Pain, Deep Learning, Explainable AI, Computational neuroscience
The student will be guided in understanding the principal cause of chronic pain, its effects and meaning in terms of reduction of quality of life and consequences on the healthcare system. They will also be introduced to state of the art of pain perception and AI with scientific literature readings. In the project, starting from models already developed for a previously collected offline dataset, the aim is to discover how different emotional and cognitive inputs alters the subjective perception of pain. The major goals for the student will be:
1) Finalization of a virtual reality environment to induce different emotional and cognitive inputs to alter the pain perception.
2) Design of a Machine Learning pipeline that will include: Signal filtering and processing, extraction of the most relevant features, development of new machine learning models to find the most meaningful biomarkers in the EEG and wearable physiological signal
3) Testing on Healthy subjects and data collection
4) Analysis and evaluation of the dynamic of subjective bias in the different conditions
**Requested skills**: Knowledge of Python, knowledge of AI and Machine Learning algorithms, Signal Processing and Statistical Analysis, knowledge of basic physiological signals (e.g., EEG). Good Programming skills. Highly motivated.
**Extra skills**: Knowledge of Unity, neuroscience of Pain, Deep Learning, Explainable AI, Computational neuroscience
Noemi Gozzi, PhD Student at the Neuroengineering laboratory, Email: noemi.gozzi@hest.ethz.ch
Valerio Aurucci, PhD Student at the Neuroengineering laboratory, Email: giuseppevalerio.aurucci@hest.ethz.ch
Dr. Stanisa Raspopovic, Assistant Professor Neuroengineering laboratory, Head ETH Zurich, Switzerland Email: stanisa.raspopovic@hest.ethz.ch
Noemi Gozzi, PhD Student at the Neuroengineering laboratory, Email: noemi.gozzi@hest.ethz.ch
Valerio Aurucci, PhD Student at the Neuroengineering laboratory, Email: giuseppevalerio.aurucci@hest.ethz.ch
Dr. Stanisa Raspopovic, Assistant Professor Neuroengineering laboratory, Head ETH Zurich, Switzerland Email: stanisa.raspopovic@hest.ethz.ch