Nahid Ul Islam About CV Research Projects Activities

Welcome!

Nahid

Nahid Ul Islam
Ph.D. in Computer Science
Arizona State University

Summary

I am a Ph.D. researcher in Computer Science at Arizona State University, working in close collaboration with Mayo Clinic on large-scale medical image analysis. My research specializes in Computer Vision and Deep Learning, with a focus on building foundation models that unify classification, detection, and segmentation within a single end-to-end framework — leveraging Vision Transformers (ViT), Swin Transformers, and large-scale multi-task pretraining to maximize the utility of expert annotations across diverse clinical imaging tasks.

I have published 6 conference and 2 journal papers in top venues with 150+ citations and h-index 7, and hold 3 granted and 5 pending U.S. patents reflecting real-world impact in AI for healthcare. I am a recipient of the prestigious President's Award for Innovation at Arizona State University (Dec'2024) and the SUN Award for invited research talk and subject matter expertise (Apr'2026).

Beyond academia, I completed two industry internships at Intel Corporation in Machine Learning and Computer Vision. I am proficient in Python, PyTorch, and TensorFlow, with hands-on experience in distributed training on HPC/GPU clusters, large-scale data pipelines, and end-to-end model development for production-scale AI systems.

Education

Ph.D. in Computer Science Arizona State University, USA (2018 – 2026)
Supervisor: Dr. Jianming Liang
Research Area: Computer Vision, Deep Learning, Medical Image Analysis
M.Sc. in Computer Science University of Texas at San Antonio, USA (2015 – 2017)
Supervisor: Dr. Qi Tian
Research Area: Computer Vision, Machine Learning, Image Processing
B.Sc. in Computer Science BRAC University, Bangladesh (2010 – 2014)
Supervisor: Rubel Biswas
Research Area: Image Processing

Experience

Graduate Research Assistant Arizona State University, Tempe, AZ (Aug 2018 – May 2026)
ASU-Mayo Clinic Joint Research Collaboration

  • Published 8 peer-reviewed papers (6 conference, 2 journal) on Computer Vision, Deep Learning and Medical Image Analysis.
  • Inventor of 3 granted and 5 pending U.S. patents in AI and Deep Learning, developed in collaboration with Mayo Clinic.
  • Built scalable AI/ML frameworks on High-Performance Computing (HPC) GPU infrastructure using distributed training, Linux/SLURM-based remote computing, workflow orchestration, and large-scale data pipelines for production-scale AI systems.
  • Developed a systematic benchmark comparing 12 CNN architectures, all major Vision Transformer and Swin Transformer variants, and 19 self-supervised pretrained models for pulmonary embolism diagnosis; proposed Embedding-based ViT (EViT), a two-stage framework that first learns 2D slice-level representations and then aggregates them for 3D exam-level diagnosis.
  • Designed Foundation X/X+, a multi-task foundation model for chest X-ray analysis integrating image classification, object detection (bounding box), and segmentation for disease analysis; conducted large-scale pretraining across 11 datasets and 20 tasks; introduced Lock-Release Pretraining to stabilize multi-task optimization and reduce task-specific overfitting; validated through extensive quantitative analysis demonstrating strong cross-task and cross-dataset generalization.
  • Developed Foundation CTPA/Chest CT-CTPA, a unified end-to-end slice-based 3D foundation model integrating classification, lesion detection, and segmentation within a single framework for coherent 2D slice-level and 3D volume-level learning; introduced Region-Guided ROI Alignment to transfer lesion detection cues to segmentation; demonstrated strong generalization to unseen downstream benchmarks and cross-domain transferability to broader CT imaging tasks.
Graduate Technical Intern Intel Corporation, Client Computing Group, Hillsboro, OR (May 2017 – Aug 2017)
Machine Learning and Computer Vision

  • Benchmarked R-CNN, Fast R-CNN, Faster R-CNN, and Mask R-CNN for real-time obstacle detection, classification, and collision prediction for moving objects.
  • Presented experimental findings and actionable recommendations to the team, contributing to pathfinding and technical readiness decisions.
Graduate Software Engineering Intern Intel Corporation, Client Computing Group, Hillsboro, OR (Feb 2017 – May 2017)
Machine Learning and Computer Vision

  • Researched deep learning approaches for human activity recognition from videos to enable smart home applications.
  • Developed an activity recognition system combining CNNs and optical flow for temporal modeling; conducted iterative experiments to improve model accuracy and robustness.

Thesis

Ph.D. Thesis Arizona State University
Title: Maximizing the Utility of Expert Annotations for Medical Image Analysis
– Research on developing foundation models for medical image analysis via large-scale pretraining across multiple datasets and tasks. Leveraging diverse annotations spanning classification, localization, and segmentation within a single end-to-end framework, while designing strategies that ensure synergy across tasks. Aiming to maximize the utilization of expert annotations to build scalable and generalizable models that enhance cross-task learning and improve clinical reliability.
M.Sc. Thesis University of Texas at San Antonio
Title: Human Face Detection Followed by Gender and Age Estimation Using Patch-Based DCT and Histogram of Oriented Gradients
– Worked on Face Detection, Gender Classification and Age Estimation classification algorithm enhancement using Discrete Cosine Transformation, Histogram of Oriented Features as Feature Extraction and Convolutional Neural Network and Support Vector Machine as Machine Learning techniques.
B.Sc. Thesis BRAC University
Title: Automated Parking Lot Management for Bengali License Plates Using Hough Transformation and Image Segmentation
– Worked on Automated parking lot management system for Bengali Language, by using Connected Component labeling, Bounding Box, Canny Edge Detection & Hough Transformation for license plate localization and used OCR for pattern recognition following by database to store as well as manipulating the system.

Peer-refereed Journal Publications

2026 1. Islam, N. U., Ma, D., Gotway, M. B., & Liang, J. Foundation Chest CT-CTPA: Maximizing the Utility of Heterogeneous Expert Annotations via a Unified Slice-Based 2D-3D Foundation Model. Medical Image Analysis. [to be submitted].
2026 2. Islam, N. U., Ma, D., Pang, J., Gotway, M. B., & Liang, J. Foundation X⁺: Enhanced Knowledge Sharing for Integrated Classification, Localization, and Segmentation with Lock-Release Pretraining and Region-Guided ROI Alignment in Chest X-ray Analysis. Medical Image Analysis. [under review].
2024 3. Islam, N. U., Zhou, Z., Gehlot, S., Gotway, M. B., & Liang, J. Seeking an optimal approach for Computer-aided Diagnosis of Pulmonary Embolism. Medical Image Analysis, 91, 102988.
2024 4. Guo, Z., Islam, N. U., Gotway, M. B., & Liang, J. Stepwise incremental pretraining for integrating discriminative, restorative, and adversarial learning. Medical Image Analysis, 103159.

Peer-refereed Conference Full Publications

2025 2. Islam, N. U., Ma, D., Pang, J., Senthil Velan, S., Gotway, M. B., & Liang, J. Foundation X: Integrating Classification, Localization, and Segmentation through Lock-Release Pretraining Strategy for Chest X-ray Analysis. WACV.
2022 3. Guo, Z., Islam, N. U., Gotway, M. B., & Liang, J. Discriminative, restorative, and adversarial learning: Stepwise incremental pretraining. MICCAI Workshop on Domain Adaptation and Representation Transfer.
2022 4. Pang, J., Haghighi, F., Ma, D., Islam, N. U., Hosseinzadeh Taher, M. R., Gotway, M. B., & Liang, J. POPAR: Patch Order Prediction and Appearance Recovery for Self-supervised Medical Image Analysis. MICCAI Workshop.
2022 5. Ma, D., Hosseinzadeh Taher, M. R., Pang, J., Islam, N. U., Haghighi, F., Gotway, M. B., & Liang, J. Benchmarking and boosting transformers for medical image classification. MICCAI Workshop.
2021 6. Islam, N. U., Gehlot, S., Zhou, Z., Gotway, M. B., & Liang, J. Seeking an optimal approach for computer-aided pulmonary embolism detection. MLMI 2021 (MICCAI Workshop).

Publicly Released Software

1. Foundation X/X+: An end-to-end model integrating classification, localization, and segmentation tasks [GitHub]
2. Foundation CTPA/Chest CT-CTPA: A slice-based 3D foundation model for chest CT imaging [GitHub]
3. Stepwise Incremental Pretraining: Achieving discriminative, restorative, and adversarial learning [GitHub]
4. Benchmarking Transformers: Benchmarking and boosting transformers for medical image classification [GitHub]
5. POPAR: Restoring patch order and appearance for self-supervised medical image analysis [GitHub]
6. Evaluating and optimizing deep learning methods for computer-aided diagnosis of pulmonary embolism [GitHub]

Patents

Granted 1. Nahid Ul Islam, Shiv Gehlot, Zongwei Zhou, Jianming Liang. U.S. Patent No. US12236592B2, “Systems, methods, and apparatuses for systematically determining an optimal approach for the computer-aided diagnosis of a pulmonary embolism” issued Feb 25, 2025.
2. Zuwei Guo, Nahid Ul Islam, Jianming Liang. U.S. Patent No. 12,572,813, “Systems, methods, and apparatuses for implementing discriminative, restorative, and adversarial (DiRA) learning using stepwise incremental pre-training for medical image analysis” issued March 10, 2026.
3. DongAo Ma, Jiaxuan Pang, Nahid Ul Islam, Mohammad Reza Hosseinzadeh Taher, Fatemeh Haghighi, Jianming Liang. U.S. Patent No. US12394186B2, “Systems, methods, and apparatuses for implementing self-supervised domain-adaptive pre-training via a transformer for use with medical image classification” issued Aug 19, 2025.
Pending 1. Nahid Ul Islam, DongAo Ma, Jiaxuan Pang, Shivasakthi Senthil Velan, Jianming Liang. U.S. Provisional Patent Application No. 63/744,744, “Integrating Classification, Localization, and Segmentation through Lock-Release Pretraining Strategy for Chest X-ray Analysis,” filed Jan 13, 2025 (patent pending).
2. Madhumitha Saravan, Nahid Ul Islam, Jiaxuan Pang, Jianming Liang. U.S. Provisional Patent Application No. 63/766,286, “Benchmarking and Boosting of 3D Segmentation Models” filed Mar 3, 2025 (patent pending).
3. Zuwei Guo, Nahid Ul Islam, Jianming Liang. U.S. Patent Application No. US20250029372A1, “Systems, methods, and apparatuses for implementing stepwise incremental pre-training for integrating discriminative, restorative, and adversarial learning into an AI model” filed July 17, 2024 (patent pending).
4. Zuwei Guo, Nahid Ul Islam, Jianming Liang. U.S. Patent Application No. US20240078434A1, “Systems, methods, and apparatuses for implementing discriminative, restorative, and adversarial (DiRA) learning using stepwise incremental pre-training for medical image analysis” filed Sept 1, 2023 (patent pending).
5. Jiaxuan Pang, Fatemeh Haghighi, DongAo Ma, Nahid Ul Islam, Mohammad Reza Hosseinzadeh Taher, Jianming Liang. U.S. Patent Application No. US20240078666A1, “Systems, methods, and apparatuses for implementing patch order prediction and appearance recovery (popar) based image processing for self-supervised learning medical image analysis” filed Sept 1, 2023 (patent pending).

Talks and Presentations

Technical Skills

Programming Languages: Python, C/C++, Java, MATLAB
ML Frameworks & Libraries: PyTorch, TensorFlow, Scikit-Learn, OpenCV, Matplotlib, NumPy, Weights & Biases, Git/GitHub
Core AI & Machine Learning:
Deep Learning, Foundation Models, Large-scale Pretraining, Transfer Learning, Fine-tuning, Multitask/Multimodal Learning, Supervised/Self-Supervised Learning, CNNs, Large Language Models (LLMs), Transformer Architectures (Vision/Swin)
Data & Experimentation: Dataset Curation, Annotation Integration, Experiment Design, Benchmarking, Distributed Training, HPC / Linux Cluster Workflows
Model Analysis & Interpretability: Grad-CAM, t-SNE Embedding Analysis, Model Monitoring, Quantitative / Qualitative Evaluation
Operating System: Windows, Linux

Timeline

May, 26' Successfully defended Ph.D. Dissertation and graduated from Arizona State University.
Apr, 26' Earned SUN Award, Arizona State University, for invited research talk and providing subject matter expert feedback.
Dec, 24' Received the prestigious President’s Award for Innovation, Arizona State University — Project: “Annotation-efficient Deep Learning for Computer-aided Diagnosis in Medical Imaging.”
May, 24' Certificate of GPSA Travel Grant Reviewer, Arizona State University.
May, 23' Certificate of GPSA Travel Grant Reviewer, Arizona State University.
Aug, 22' Travel Grant, MICCAI 2022, awarded by the GPSA (Arizona State University).
Apr, 22' Travel Grant, CVPR 2022, awarded by the GPSA (Arizona State University).
Aug, 18' Started Ph.D in Computer Science at Arizona State University.
May, 18' Received prestigious CIDSE Doctoral Fellowship from the School of Computing, Informatics, and Decision Systems Engineering at Arizona State University.
Mar, 18' Got admission into Ph.D program in Computer Science at Arizona State University.
Dec, 17' Completed M.Sc. in Computer Science from University of Texas at San Antonio, USA.
Nov, 17' Sucessfully defended Masters Thesis at University of Texas at San Antonio.
May, 17' Joined Intel Corporation for the summer'17 semester as Graduate Technical Intern.
Feb, 17' Joined Intel Corporation for the spring'17 semester as Graduate Software Engineering Intern.
Aug, 15' Started M.Sc. in Computer Science at University of Texas at San Antonio, USA.
Nov, 14' Completed B.Sc. in Computer Science from BRAC University, Bangladesh.