Principal Investigator: 

​​Dr James Pope

Lecturer in Data Science

School of Engineering Mathematics and Technology

University of Bristol

Project:

1. Determine Tone Bias in Open Source Skin Cancer Datasets

• Identify the skin tone bias in skin cancer 

• Implement explainable AI to understand the model biases. 

Funded By:

• Jean Golding Institute, University of Bristol.

Responsibilities:

• I have taken 9 hours of Lab class and conducted a Lecture.

• Taught how to operate electronic apparatus with safety.

• Learned how to build the circuit efficiently on the trainer board and simulated it in software.

•  Instruct how to use Electronic apparatus and how to design electronic circuits efficiently.

• Update the lab manuals to meet accreditation standards.

• Supervise the course projects.

• Held office hours regularly, 11 hours per week, and by appointment.

Taught Courses:

• Electrical Circuits - I Lab [EEE141L]

• Electrical Circuits - II Lab [EEE241L]

• Digital Logic Design Lab [CSE231L/EEE211L]

• Computer Architecture and Organization Lab [CSE332L/EEE436L]

• Introduction to Communications Systems Lab [EEE332L]

• Electrical Machines Lab [EEE336L]

• Verilog HDL: Modeling, Simulation, and Synthesis Lab [EEE413L]

Supervisor: 

Dr Tanzilur Rahman

Assistant Professor

Dept. of Electrical and Computer Engineering

North South University

Responsibilities:

• Perform original research.

• Use various sources for background study, including related database publications.

• Work under tight deadlines.

• Collaborate on other projects as needed.

• Maintain a Research Notebook (Code, Data, so on).

• Responsible for proposals and report writing.

Project:

1. Continuous and Real-Time Monitoring of ECG

   • Design an IoT-enabled device for acquiring ECG signals.

   • Develop an algorithm to extract vital signs.

   • Develop the model to classify abnormalities of ECG signals.

2. Measure Glucose level using a Smartphone Camera

   • Collect data with ground-level truth.

   • Develop the Machine Learning model for measuring glucose levels using a smartphone with the lowest standard error of prediction (SEP) at 17.02 mg/dL.

3. Transmission electron microscopy (TEM) Virus Image Classification

   • Developed a Deep Learning model for 14 virus classifications.

   • Used transfer learning using pre-trained model InceptionV3.

   • Achieved 99.27% accuracy on test data.

4. Smart Traveler Assistant App

   • Develop an algorithm to remove the noise of sensor data.

   • Analyze sensor data to develop the Smart Traveler Assistant App.

5. Analyze Green Tea effects

   • Develop a system to acquire ECG and PPG signals using NI-DAQ USB-6001.

   • Verify signals to PAT analysis.

   • Analyze signals to extract vital signs.

Funded By:

1. Worked on Bangladesh Govt. ICT Ministry fund.

2. Worked on North South University research fund NSU-CTRG.