MissFarwa Abbas

Biomedical and Statistical Signal Processing

Development of signal processing methodologies for the analysis of multimodal biomedical data such as EEG, sEMG, MEG, and MRI. This includes statistical signal modeling, source localization, and feature extraction techniques aimed at understanding neural dynamics and identifying clinically relevant biomarkers in neurological and movement disorders.

Inverse Problems and Imaging
Inverse problems arising in biomedical imaging and neurophysiology, including the reconstruction of latent neural sources and physiological states from indirect, noisy, and high-dimensional measurements. This includes theoretical and computational approaches to ill-posed inverse problems in neuroimaging and related modalities.

Deep Learning for Image and Signal Reconstruction
Strong interest in deep learning based approaches for image and signal reconstruction, particularly hybrid model-based and data-driven methods. This includes neural architectures for accelerated and robust reconstruction in biomedical imaging, as well as learning priors and representations that improve generalization and interpretability in inverse problems.

Convex Optimization and Structured Signal Representations
Convex optimization methods for learning low-rank, sparse, and structured representations of signals and images. This spans optimization algorithms, regularization techniques, and theoretical guarantees for recovering meaningful structure in high-dimensional biomedical data, with applications to imaging, neural signal analysis, and connectivity modeling.

Computational Modeling of Neural Systems
Mathematical and computational modeling of neural population dynamics using differential equations and stochastic systems, with the goal of linking physiological mechanisms to observable signals and imaging data.

Neuromodulation and Data-Driven Neuroscience
Analysis of non-invasive brain stimulation data (e.g., TMS) to study neuroplasticity and working memory, using signal processing, biophysical modeling, and machine learning techniques to characterize brain-state dependent effects, individual variability and pharmacological effects on neuroplasticity.