Gent | More than two weeks ago
In this PhD we will research the neural activity associated with the interoceptive pathway as indexed by Heartbeat Evoked Potentials to design an AI/ML-driven objective outcome parameter for the cortical effects of neuromodulation treatments as an AI/ML-driven biomarker for depression and seasonal affective disorder in clinical neuroscience investigations.
Depression and seasonal affective disorder (SAD) are associated with altered interoceptive processing, and Heartbeat Evoked Potentials (HEPs) reflect these changes. By studying HEPs, we can gain insights into how depression and SAD affects the brain’s processing of bodily signals and how neuromodulation treatments like transcranial current stimulation (tCS), transcutaneous auricular vagus nerve stimulation (taVNS) and vagus nerve stimulation (VNS) may help alleviate depressive symptoms.
tCS is a non-invasive brain stimulation technique that applies low-intensity electrical currents to the scalp to modulate neuronal activity. taVNS is a non-invasive neuromodulation technique that stimulates the auricular branch of the vagus nerve, located in the outer ear. Imec is currently developing miniature implants for peripheral nervous system stimulation (VNS) as the vagus nerve is a major component of the parasympathetic nervous system that plays a role in emotional regulation by affecting areas of the brain involved in mood control, such as the limbic system and prefrontal cortex. Stimulating the vagus nerve is believed to modulate neurotransmitter levels (such as serotonin and norepinephrine) and improve connectivity and activity in these brain regions, helping to alleviate depressive symptoms.
Building further upon first results on taVNS influencing HEPs [1], and upon the joint AAA predictive healthcare research on AI/ML for neuromodulation, in this PhD, we want to research the potential of neuromodulation treatments for depression and SAD. This research will focus on personalized treatment protocols while addressing the high inter-patient physiological variability that typically complicates the effectiveness of neuromodulation therapies. To do so, we will leverage the (medical) lessons learned from both taVNS and tCS research to explore the potential of personalized AI/ML in optimizing these neuromodulation techniques for the individual patient. The inter-patient psychological variability will require the design of advanced (hybrid) AI/ML models capable of adapting treatment strategies in real-time based on patient-specific biomarkers, physiological signals, and neural feedback. This will allow us to design personalized, trustworthy and advanced AI/ML models that can ultimately inform and enhance novel invasive VNS methods, facilitating innovation in the field.
This research idea aligns with the roadmap of imec NL on brain & peripheral nerve stimulation, and builds further upon IDLab’s joint research with imec NL on AI/ML for neuromodulation, MEPs (motor evoked potentials) and the design of digital biomarkers for a.o. stress & depression within the AAA Preventive Healthcare.
We will research the neural activity associated with the interoceptive pathway as indexed by HEPs to design an AI/ML-driven objective outcome parameter for the cortical effects of neuromodulation treatments as an AI/ML-driven biomarker for depression and SAD in clinical neuroscience investigations. This biomarker is the first step towards designing closed-loop VNS treatments using imec’s implant for peripheral nervous system stimulation for depression and/or bioelectronic methods that can provide light-based treatment that can be placed on the skin and safely and effectively replace the lack of sun rays during fall and winter periods in high latitude areas for SAD depression treatment.
[1] Tasha Poppa, Lars Benschop, Paula Horczak, Marie-Anne Vanderhasselt, Evelien Carrette, Antoine Bechara, Chris Baeken, Kristl Vonck, Auricular transcutaneous vagus nerve stimulation modulates the heart-evoked potential, Brain Stimulation, 15 (1): 260-269, 2022
Involved teams:
PIs:
We also collaborate with Kristl Vonck, Evelien Carette and Sofie Carette from Neurology, UZGhent, whose medical expertise will be crucial in ensuring the integration of clinical insights and guiding our research in areas where neurological input is essential.
Required background: AI/ML computer science background with biomedical / electrical engineering neurotechnology experience or aspiration
Type of work: 10% biomedical & statistical analysis of (ta)VNS and HEPs, 10% literature study, 20% VNS simulation, 60% data analysis and AI/ML modeling/evaluation
Supervisor: Sofie Van Hoecke
Co-supervisor: Vojkan Mihajlovic
Daily advisor: Sofie Van Hoecke
The reference code for this position is 2025-136. Mention this reference code on your application form.