Nooshin Bahador

Welcome to my profile! I’m a Computer Science Engineer (PhD) with extensive experience in data analysis, especially in the healthcare industry.

You can find more of my work and projects on my GitHub.

For more details on my publications, please visit my Google Scholar profile.

Feel free to connect with me on LinkedIn.

Background¶

With over 10 years of experience in data analysis, I specialize in:

• Automating medical data workflows

• Transforming clinical data

• Building medical monitoring systems

I have extensive knowledge in handling both structured and unstructured healthcare data, focused on improving data quality and developing systems for clinical decision support.

Project¶

The main objective of this project is to develop a Seizure Monitoring System that can detect and analyze epileptic activity. Epileptic seizures are caused by sudden bursts of electrical activity in the brain, often resulting from imbalances in neurotransmitter levels. These seizures can be triggered by various factors, and understanding their onset and progression is essential for accurate diagnosis and effective treatment.

This project focuses on the detection of specific biomarker known as “chirp” in electrographic data. These biomarkers are consistently present during epileptic episodes and exhibit distinct frequency modulation patterns in the electrographic discharges. The identification of these patterns can serve as reliable indicators for advancing seizure detection technologies and improving patient outcomes.

The main objective of this project is to develop a Seizure Monitoring System that can detect and analyze epileptic activity. Epileptic seizures are caused by sudden bursts of electrical activity in the brain, often resulting from imbalances in neurotransmitter levels. These seizures can be triggered by various factors, and understanding their onset and progression is essential for accurate diagnosis and effective treatment.

This project focuses on the detection of specific biomarkers known as “chirp” in electrographic data. These biomarkers are consistently present during epileptic episodes and exhibit distinct frequency modulation patterns in the electrographic discharges. The identification of these patterns can serve as reliable indicators for advancing seizure detection technologies and improving patient outcomes.

For a detailed explanation of the “chirp” pattern and its significance in seizure detection, please refer to my published research.

100,000 Labeled Chirp Spectrogram Images – Download on Hugging Face!¶

Funded by Canadian Neuroanalytics Scholars Program

Citation Bahador, N., & Lankarany, M. (2025). Chirp localization via fine-tuned transformer model: A proof-of-concept study. arXiv preprint arXiv:2503.22713. [PDF]

Sample Generated Spectrogram¶

Sample Generated Label¶

Fine-Tuning Vision Transformer (ViT) with LoRA for Spectrogram Regression¶

Neuro Dominance Tracker¶

A Software for Identification and Characterization of Dominant Rhythm in Neural Time Series (e.g., EEG, LFP)

Visit the Website¶

Citation¶

If you use this software in your research, please cite it as follows:

Bahador N, Sengupta S, Saha J, Lankarany M, Zhang L, Skinner F. A software for identification and characterization of theta rhythms in the hippocampus. bioRxiv. Published online 26 March 2025. doi:10.1101/2025.03.25.645280.

Step-by-Step Instructions for Analyzing Data using MATLAB App¶

1. Open MATLAB App¶

• Launch the MATLAB App and wait for the App Designer window to appear.

• Click on the “Run” button to open the user interface.

2. Upload Your File¶

• Place the “fieldtrip-master” folder and the “abfload.m” script in the same directory as the NeuroDominanceTracker GUI executable.

• Upload your file.

• Enter the sampling frequency (e.g., 5000 Hz).

3. Perform Analysis¶

• Click on the “Perform Analysis” button.

• The app will analyze the recording in one-minute windows.

• Messages will appear in the MATLAB command window for each window analyzed, indicating if a dominant mid-frequency is observed.

• Wait for all windows to be analyzed (e.g., 118 windows for a sample two-hour recording).

4. View Results¶

• Once analysis is complete, a table listing all identified episodes and their characteristics will be displayed.

• You can extract this table as an Excel sheet by clicking the appropriate button.

5. Visualize Identified Events¶

• Select an event and enter its corresponding minute (e.g., 109).

• Click the button to visualize the selected event.

6. Time-Frequency Analysis¶

• A spectrogram of the one-minute segment starting from the selected minute will be plotted.

• The normalized power of mid, low, and high-frequency bands over time will be shown.

• Dominant mid-frequency segments will be highlighted in green.

• The power spectrum of the first dominant mid-frequency segment will be plotted.

7. Save Features¶

• Click the button to save the features as an Excel file.

Sample EEG Data¶

The sample EDF (European Data Format) files used in this project were obtained from the following publication:

Brown, L. A., Hasan, S., Foster, R. G., & Peirson, S. N. (2016). The raw EEG data, 4 files (EEG_A to D), in European data format (.edf) [Data set]. Zenodo. Available at: Brown et al. (2016) Direct access to the data can be found here: Zenodo Record.

FieldTrip Toolbox¶

This project leverages the FieldTrip toolbox to support different file formats, including FIF, EDF, and BDF.

Reference: Oostenveld, R., Fries, P., Maris, E., & Schoffelen, J.-M. (2011). FieldTrip: Open Source Software for Advanced Analysis of MEG, EEG, and Invasive Electrophysiological Data. Computational Intelligence and Neuroscience, 2011, 156869. You can access the FieldTrip toolbox here: https://www.fieldtriptoolbox.org/.