Stage 1

• Interim report (2022)
• Attended to Computational Psychiatry Course 2022 (Location: Zurich)

Stage 2

• Attended to Computational Psychiatry Conference 2023 (Location: Dublin, Ireland)
• Attended to 11th International Conference on E-Health and Bioengineering 2023 (Location: Bucharest, Romania)
• Publications:
  • Conference poster:
    • Conference: Computational Psychiatry Conference 2023 (Location: Dublin, Ireland)
    • Title: State observer-based Dynamic Causal Modelling for EEG application in Computational Psychiatry
    • Authors: Andrei Popescu, Catalin Buiu
  • Conference paper:
    • Conference: 11th International Conference on E-Health and Bioengineering 2023 (Location: Bucharest, Romania)
    • Title: Mean Membrane Potential Estimation for Neural Mass Models in EEG Recordings Using a Linear State Observer
    • Authors: Andrei Popescu, Catalin Buiu
  • Journal paper:
    • Journal: Journal of Control Engineering and Applied Informatics
    • Title: Robust control and state observer design for neural mass model applications using simulated EEG signals
    • Authors: Andrei Popescu, Catalin Buiu

Stage 3

• Final report (2023-2024)

The impact of the project

The anticipated impact of the scientific results will be presented from the perspective of modeling, classification, and simulation:

• Modeling and Classification: The mathematical and computational models obtained using the Dynamic Causal Modeling (DCM) method can validate those existing in the field of computational psychiatry. Furthermore, they may pinpoint the brain regions contributing to the decision-making process along with the level of impairment for a specific pathology.
• Simulation: The models can be utilized for the verification and validation of potential treatments. Moreover, they can be employed for estimation and prediction, specifically to replicate the decision-making process under certain conditions, otherwise impossible to test in a real-life scenario.

The most significant result is, undoubtedly, the combination of models obtained through the Dynamic Causal Modeling technique with the state estimation method. The primary advantage is that this method can yield information and details about a cortical column that are not directly measurable. In layman terms, it can be said that this method represent a “mathematical microscope”, meaning that merging together mathematical models with real-time EEG observations, one can reveal hidden information about the brain activity otherwise not available.