Projects

Department of Medicine / Centre for Neuroscience Studies

Ongoing Research Projects | 2025-Present

• Implementing AI algorithms for pre-surgical decision-making for neurological diseases and in particular for epilepsy.
• Chain-of-thought and PEFT fine-tuning of EEG-Language models for seizure detection.
• RAG (Retrieval-Augmented Generation) to search over neurologists' reports to retrieve the searched content.
• Analyzing the pipelines and monitoring systems in the hospital in order to develop programming code to make the processes more efficient.
• Improving the algorithms designed to precisely detect hippocampal sclerosis measurements and MRI arterial spin labeling asymmetry detection (Kingston General Hospital Project).

Human-Centered AI and Interactive Machines Lab / Ingenuity Labs Research Center

Research Projects | 2023-2025

• Developed AI and ML models for time-series forecasting and anomaly detection applications (Bank of Montreal Project).
• Identified critical research gaps and created multi-domain data systems using self-supervised learning techniques.
• Authored scientific manuscripts, protocols, and grant applications, collaborating with researchers to ensure clarity, accuracy, and impactful delivery.

Artificial Brain Research Lab

Research Assistant: Projects Overview | 2014 - 2023

• Applied self-supervised contrastive learning with label information for low-rank, sparse, and orthogonal data representations using Lagrangian multipliers.
• Developed an ECG prediction model for arrhythmia detection (including atrial fibrillation and myocardial infarction) using a resampling technique with orthogonality constraints (Kyungpook Hospital project).
• Built a document translation system for high-cursive ancient documents for the Korean National Library, integrating detection, recognition, and translation modules (Korean National Cultural Project).
• Implemented a low-shot image recognition model for blurry, noisy environments, focusing on edge-based representations for intra- and inter-features distinction for images at long stand-offs (LG project).
• Created a contactless, robust palm-print recognition model using a deep convolutional network (Human-Robot Interaction project).
• Developed a regularized, low-shot attention transfer model with t-normalization for classifying and translating ancient documents.

Science and Technology Center, Shiraz University of Technology

Research Assistant: Projects Overview | 2010 - 2013

• Performed clustering on medical datasets, including breast cancer and ECG data, by integrating K-Means with evolutionary optimization algorithms such as the Imperialist Competitive Algorithm (ICA).
• Implemented evolutionary algorithms (Imperialist Competitive Algorithm and Particle Swarm Optimization) in conjunction with conventional controllers.