This repository introduces a region-based deep learning framework for identifying individual neural signatures from electroencephalography (EEG) recorded during natural reading tasks.

Mini-project using the PyTorch framework

UNet-HCRF Integration for EEG Pattern Recognition

Biomedical Signal Processing & Medical Images - projects realized by a team of students from Politecnico di Milano.

This application is primarily used to identify cognitive disabilities in people of any age. It does this by taking a screening test and estimating a user's likelihood of having a disability such as dyslexia or dyscalculia. It also offers various ways to monitor a patient's progress and create a graph as a result.

PyTorch implementation of the Subject-Conditioned Layer from “Mitigating Subject Dependency in EEG Decoding with Subject-Specific Low-Rank Adapters” enabling subject-specific LoRA adaptation for EEG decoding.

Machine-learning pipeline for predicting sedation states during propofol sedation based on EEG-derived graph-theoretical measures.

EEG-based machine learning classification of Alzheimer's Disease and Frontotemporal Dementia using frequency-domain analysis.

Transformers for hand movement classification using EEG data

Biomedical Robotics 2022 Practical Activity based on EEG Data

EEG-based deep learning classification with Grad-CAM explainability. Developed as a team project.

Code for a Random Forest Model and Bidirectional Long Short Term network. See README file for details on overarching thesis.

Using Optimization algorithms for Collective Feature Selection for Accuracy Boosting in Alzheimer Detection with KNN, LDA and Decision Tree Classifiers! (Accelerated With Parallel Programming ToolBox!)

A web-based toolkit to design complicated deep learning workflows for EEG signal processing and classification

Resources for the paper titled "Evaluating Latent Space Robustness and Uncertainty of EEG-ML Models under Realistic Distribution Shifts". Accepted at NeurIPS 2022.

This project showcases the development of an EEG-controlled 4DOF Robotic Arm using the STM32f103c8t6 microcontroller (Blue Pill). The arm is designed to perform basic movements controlled by EEG signals, making it suitable for assistive technologies and prosthetic applications.

Unified framework combining Riemannian Wave Classification (RWC) and Geometric Wave Learning (GWL). Treats supervised learning as quantum resonance (Breit-Wigner/Lorentzian) on label-driven discrete Ricci-flow manifolds, using holographic radial frequency (HRF) kernels and polychromatic spectral ensembles.

This project aims to better understand how ketamine affects consciousness. It investigates markers of ego dissolution and dose-response relationships. Several subsequent studies will probably be published based on this project.

Github Repository for the paper "Different Algorithms (Might) Uncover Different Patterns: A Brain-Age Prediction Case Study" - BIBM 2023