Research focus
Explore Masud Rabbani’s research on EEG, brain-computer interfaces, intentional eye-blink classification, nonlinear brain-signal analysis, and non-invasive cerebral sensing.
5 related publications · Research projectsResearch overview
My EEG and brain-computer interface research asks how physiological signals can become dependable inputs for real-time computing systems. The work connects wearable electroencephalography, intentional eye-blink patterns, nonlinear signal analysis, machine learning, and emerging agentic-AI architectures. Across these projects, the central challenge is not simply classifying a signal in a clean dataset; it is deciding which signal representations remain meaningful when recordings are noisy, people differ, and a system must respond quickly enough to support interaction.
One line of work treats eye blinks as intentional control signals rather than artifacts to discard. A real-time study of consecutive eye blinks compared statistical, time-domain, amplitude, and frequency-domain representations with conventional machine-learning and neural models. The result is a research pathway toward hands-free commands and more accessible interfaces. Related work studies attention in EEG through nonlinear and chaos-based features, while a broader survey organizes nonlinear techniques used to characterize complex brain dynamics.
A complementary direction explores whether low-frequency acoustic signals associated with cerebral blood flow could provide another non-invasive view of brain activity. That concept is deliberately framed as an early sensing hypothesis, not a clinical diagnostic system. Together, the EEG and acoustic studies define a broader program: combine carefully designed sensing with transparent signal processing, compare multiple modeling strategies, and validate every proposed interaction beyond a single controlled experiment.
My contribution
My contribution to this research area combines system development with empirical EEG analysis. In our 2025 Scientific Reports eye-blink BCI study, the published author-contribution statement records that I developed the algorithm, designed the hardware system, collected the data, and performed the data analysis. The work evaluated wearable eight-channel EEG as a real-time input for distinguishing no blink, one blink, and two consecutive blinks.
I also coauthored research that organizes nonlinear EEG-analysis methods, applies chaos-based features to attention data, and frames cerebral blood-flow sound sensing as an early research hypothesis. Across these studies, I focus on accessible physiological interaction and transparent comparison of signal representations while keeping prototype results separate from clinical claims.
Key findings
Each result links to its complete publication record, abstract, authors, affiliations, and publisher source.
A controlled study with 10 participants classified three blink states. The strongest conventional model achieved 89.0% accuracy; the YOLO model reported 98.67% recall, 95.39% precision, and 99.5% mAP50.
Read the supporting paperA survey synthesized 28 EEG studies across six nonlinear-analysis families: correlation dimension, Lyapunov exponents, recurrence analysis, surrogate-data analysis, nonlinear regression, and entropy-based measures.
Read the supporting paperA PhysioNet-based attention study reported mean time delay of 2.50 seconds, mean embedding dimension of 4.41, and mean correlation dimension of 2.23.
Read the supporting paperQuestions
How can voluntary blink sequences and other recognizable EEG events support fast, low-burden commands for people who may not be able to use a conventional interface?
Which time, frequency, statistical, and nonlinear features retain useful information across recording sessions, wearable devices, participants, and changing real-world conditions?
How can machine learning and agentic AI coordinate sensing, interpretation, feedback, and task-level decisions while keeping the human user in control?
Research method
Collect multichannel wearable EEG or exploratory acoustic signals with documented sensor placement, timing, and task conditions so the origin of each signal is clear.
Build time-domain, frequency-domain, amplitude, statistical, and nonlinear features that expose different aspects of physiological dynamics instead of relying on one representation.
Evaluate conventional classifiers, neural models, and emerging AI workflows with measures suited to the task, including precision, recall, latency, and generalization.
Connect validated signal events to understandable interface actions, then study usability, reliability, and failure modes in increasingly realistic settings.
Journal article · 2025
Masud Rabbani, Nafi Us Sabbir Sabith, Anubhav Parida, Iysa Iqbal, Sayed Mashroor Mamun, Rumi Ahmed Khan, Farhad Ahmed, Sheikh Iqbal Ahamed. “EEG based real time classification of consecutive two eye blinks for brain computer interface applications”. Scientific Reports, 15, 21007. (2025). https://doi.org/10.1038/s41598-025-07205-0.
Conference paper · 2026
Masud Rabbani, Rubaba Amyeen, Md Mazhar Hossain, Mostofa Rafid, Iysa Iqbal, Hansika Kolli, Sheikh Iqbal Ahamed. “Agentic AI for EEG-Based Brain–Computer Interfaces: A Review of Methods, Systems, and Applications”. 2026 IEEE 50th Annual Computers, Software, and Applications Conference (COMPSAC), 2844–2852. (2026). https://doi.org/10.1109/COMPSAC69091.2026.00428.
Conference paper · 2024
Masud Rabbani, Subarna Alam, Md. Raihan Mia, Anubhav Parida, Iysa Iqbal, Hansika Kolli, Parama Sridevi, Kazi Shafiul Alam, Paramita Basak Upama, Rumi Ahmed Khan, Sheikh Iqbal Ahamed. “Listening to the Brain: A Novel Approach to Understanding Cerebral Dynamics through Blood Flow Sounds”. 2024 IEEE 48th Annual Computers, Software, and Applications Conference (COMPSAC), 950–957. (2024). https://doi.org/10.1109/COMPSAC61105.2024.00131.
Conference paper · 2023
Masud Rabbani, Sayed Mashroor Mamun, Parama Sridevi, Iysa Iqbal, Anubhav Parida, Anushka Kolli, Hansika Kolli, M. Rubaiyat Hossain Mondal, Mohammad Aftab Rassel, Enayet Hossain, Farhad Ahmed, Sheikh Iqbal Ahamed. “A Chaos-based Non-linear Analysis Method for Detecting Human Attention Levels in EEG Signals”. 2023 IEEE 23rd International Conference on Bioinformatics and Bioengineering (BIBE), 201–204. (2023). https://doi.org/10.1109/BIBE60311.2023.00039.
Conference paper · 2023
Sheikh Iqbal Ahamed, Masud Rabbani, Richard J. Povinelli. “A Comprehensive Survey on Detection of Non-linear Analysis Techniques for EEG Signal”. 2023 IEEE International Conference on Digital Health (ICDH), 184–194. (2023). https://doi.org/10.1109/ICDH60066.2023.00034.
Responsible translation
The near-term opportunity is accessible interaction: a small vocabulary of dependable physiological commands may be more useful than a complex interface that works only under laboratory conditions. The research therefore emphasizes interpretable tasks, real-time feasibility, and explicit reporting of prototype limitations.
Clinical or assistive use requires larger and more diverse cohorts, repeat measurements, cross-device testing, and collaboration with end users and domain specialists. None of the systems described here should be read as a medical device or diagnostic claim. They are research contributions designed to make the next validation step clearer and more rigorous.
Continue exploring