“Live as if you were to die tomorrow. Learn as if you were to live forever.”
- Mahatma Gandhi
Thesis and Projects:
Biomedical Engineering, Bio signal Processing, BCI, Brain Signal processing, Neural Network(BP)
Undergraduate Thesis:
Physiological signal Analysis using ECG, EMG, EEG and Blood Pressure Measurement
Supervision: Professor Dr. Md. Mostafizur Rahman
- Thesis work involved collecting, processing and analyzing physiological signals for affective computing purposes.
- Detection of different human cognitive state by analyzing EEG using neural network.
- Detection of human physical and mental stresses due to overuse of computer by analyzing physiological signals.
We collected and analyzed EEG signals by using 3 electrode system for detecting four cognitive states of human subject for brain-computer interface purpose. Cognitive states are:
- Resting State (RS)
- Thought (TH)
- Memory (MR)
- Emotion (EM)
All the signals were collected by using BIOPACK and for better detection of the cognitive state a three layer back propagation neural network was used. Result of the thesis work shows that using Gamma band with Alpha, Theta and Delta increase classification performance and our system of collecting and processing EEG signals with limited feature number, generates higher detection accuracy than other systems and more importantly classification performance remains approximately invariant with the number of NN hidden layer units.
Fig: Electrodes connected in to the occipital region of the brain.
Accepted Conference: EICT-2013
IEEE Link: http://ieeexplore.ieee.org/xpl/articleDetails.jsp?reload=true&arnumber=6777878
Human performance analysis(stress detection):
We also worked with the detection of stress level of computer user in office like working environment by analyzing the physiological signals. We used ECG, EMG, EOG and EEG signals for the detection of induced stress level. The goal of our work was to determine the factors contribute most for the induced stress in computer mediated tasks. We were mainly worked with three factors,
1. Stress due to intense eye work (due to continuous screen monitoring, reading online or offline articles etc)
2. Stress due to Mental strain (work related to intense mental activity)
3. Stress due to muscle strain.
We believe detecting the stress level will make the computer more intelligent so that it can operate by taking users affective state in to account.
We also worked with the detection of stress level of computer user in office like working environment by analyzing the physiological signals. We used ECG, EMG, EOG and EEG signals for the detection of induced stress level. The goal of our work was to determine the factors contribute most for the induced stress in computer mediated tasks. We were mainly worked with three factors,
1. Stress due to intense eye work (due to continuous screen monitoring, reading online or offline articles etc)
2. Stress due to Mental strain (work related to intense mental activity)
3. Stress due to muscle strain.
We believe detecting the stress level will make the computer more intelligent so that it can operate by taking users affective state in to account.
Fig: ECG Data Collection in the Laboratory and raw ECG and Heartbit signal
Fig: EMG Data Collection in the Laboratory and raw EMG and Rms EMG signal
Fig: EOG data collected at the Laboratory for the left to Right movement of the eye
Accepted Conference: ICCIT-2013 & ICCIT-2014
- IEEE Link: http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=6997300
- IEEE Link: http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=7073120
Projects:
1. Line follower robotic system
2. Design of a High Gain Micro strip Patch Antenna for WLAN/Bluetooth Application at 2450 MHz Frequency
3. Noise cancellation in a communication channel by using modified LMS algorithm
1. Line follower robotic system
- Designed and constructed a robotic system that can follow a certain color line
2. Design of a High Gain Micro strip Patch Antenna for WLAN/Bluetooth Application at 2450 MHz Frequency
- Designed a micro strip patch antenna using 4nec2 at 2459 MHz frequency and
- Practically implemented the designed antenna using PCB
3. Noise cancellation in a communication channel by using modified LMS algorithm
- Implemented modified LMS algorithm by using MATLAB to observe the noise cancellation in a communication channel