Tasnuba Islam

AI • Clinical Data • Medical Imaging Researcher

Research Intern, Kamaleswaran Lab, Duke University

Graduate Teaching Assistant, Department of Electrical and Computer Engineering, North South University

I am a researcher working on the development of reliable AI methods for healthcare. My work focuses on structured clinical data pipelines, foundational models for healthcare data, synthetic EHR generation using generative models, medical image segmentation, and speech emotion recognition for low-resource languages.

I have mentored and guided 400+ students across Mathematics and Engineering.

Email GitHub Scholar LinkedIn

News & Updates

Recent accomplishments and ongoing work

FEB

2026

Presented Work at PRFC 2026 (Geneva)

Contributed to a collaborative research project from the Kamaleswaran Lab, presented at the Pediatric Research for the Future Children (PRFC) Global Forum in Geneva. The work focused on learning patient state representations from sparse pediatric ED data to support early sepsis triage.

Research

FEB

2026

Joined as Graduate Teaching Assistant (ECE)

Joined the Department of Electrical and Computer Engineering as a Graduate Teaching Assistant, supporting coursework, mentoring students, and assisting with academic instruction.

Teaching

SEP

2025

Abstract Accepted at PRFC 2026 — A Kamaleswaran Lab Collaboration

Our project on offline reinforcement learning for pediatric sepsis triage has been accepted as an oral presentation.

Publication

SEP

2025

Started Working on LightSeg and BanglaFusion

Developing hybrid CNN–Transformer architectures for low-resource SER and interpretable fetal ultrasound segmentation. Preparing for publication.

Research

JUL

2025

Joined the Kamaleswaran Lab at Duke University

Began working as a Research Intern focusing on structured clinical datasets, synthetic EHR generation using GANs, and pediatric sepsis studies.

Research

APR

2025

Graduated from North South University

Completed a Bachelor of Science in Computer Science and Engineering with distinction.

Milestone

APR

2025

Joined CloudlyIO as Software Engineer Intern

Worked on backend development, ML pipelines, and engineering automation. Completed internship in June 2025.

Industry

APR

2025

Presented LightSeg at Capstone Showcase

Presented “LightSeg: An Interpretable and Lightweight Framework for Multiclass Fetal Ultrasound Segmentation” at the Capstone Project Showcase hosted by North South University.

Showcase

JAN

2025

Started as Undergraduate Teaching Assistant (ECE)

Assisted in Circuits and Engineering Economics. Mentored 120+ students, graded scripts, conducted office hours, and supported faculty until August 2025.

Teaching

JAN

2024

Undergraduate Teaching Assistant — Math & Physics

Assisted in Calculus and Analytic Geometry, Advanced Physics, and Probability and Statistics. Mentored 280+ students and supported faculty through December 2024.

Teaching