Publications

Prototype-Guided Zero-Shot Medical Image Segmentation with Large Vision-Language Models Permalink

Published in Journal of Applied Sciences, 2025

This paper presents a prototype-guided zero-shot medical image segmentation framework leveraging large vision-language models for enhanced segmentation accuracy without requiring fine-tuning on domain-specific datasets.

Recommended citation: Pham, H., & Cheng, S. (2025). "Prototype-Guided Zero-Shot Medical Image Segmentation with Large Vision-Language Models." Applied Sciences, 15(21), 11441. https://doi.org/10.3390/app152111441

Non-Iterative Cluster Routing: Analysis and Implementation Strategies Permalink

Published in Journal of Applied Sciences, 2024

This paper introduces a non-iterative cluster routing technique for capsule networks, offering enhanced performance with fewer parameters compared to traditional iterative methods. The approach maintains the part–whole relationship and demonstrates robust generalization to novel viewpoints.

Recommended citation: Pham, H., & Cheng, S. (2024). "Non-Iterative Cluster Routing: Analysis and Implementation Strategies." Applied Sciences, 14(5), 1706. https://doi.org/10.3390/app14051706

Enhancing Semantic Segmentation through Reinforced Active Learning: Combating Dataset Imbalances and Bolstering Annotation Efficiency Permalink

Published in Journal of Electronic & Information Systems, 2024

This study introduces a reinforced active learning framework for semantic segmentation, integrating techniques such as Dueling Deep Q-Networks, Prioritized Experience Replay, and Noisy Networks. The approach addresses dataset imbalances and annotation efficiency, demonstrating improved segmentation performance with reduced labeling efforts.

Recommended citation: Han, D., Pham, H., & Cheng, S. (2024). "Enhancing Semantic Segmentation through Reinforced Active Learning: Combating Dataset Imbalances and Bolstering Annotation Efficiency." Journal of Electronic & Information Systems, 5(2), 45–60. https://doi.org/10.30564/jeis.v5i2.6063

Deep learning-based rectum segmentation on low-field prostate MRI to assist image-guided biopsy Permalink

Published in SPIE Conference, Medical Imaging: Image-Guided Procedures, Robotic Interventions, and Modeling, 2023

This study presents a deep learning approach for automatic rectum segmentation in low-field (58–74 mT) prostate MRI scans. Utilizing a U-Net architecture, the model achieved a Dice similarity coefficient of 0.89, demonstrating its potential to enhance the accuracy and efficiency of image-guided prostate biopsies in low-resource settings.

Recommended citation: Pham, H., Le, D. B. T., Sadinski, M., Narayanan, R., Nacev, A., & Zheng, B. (2023). "Deep learning-based rectum segmentation on low-field prostate MRI to assist image-guided biopsy." In Proceedings of SPIE, Vol. 12466, Medical Imaging 2023: Image-Guided Procedures, Robotic Interventions, and Modeling. https://doi.org/10.1117/12.2654511

Identifying an optimal machine learning generated image marker to predict survival of gastric cancer patients Permalink

Published in SPIE Conference, Medical Imaging: Computer-Aided Diagnosis, 2022

This study investigates the feasibility of identifying and applying a novel quantitative imaging marker, generated through machine learning techniques, to predict the survival of gastric cancer patients. The approach aims to enhance prognostic accuracy and assist in personalized treatment planning.

Recommended citation: Pham, H., Jones, M., Gai, T., Islam, W., Danala, G., & Zheng, B. (2022). "Identifying an optimal machine learning generated image marker to predict survival of gastric cancer patients." In Proceedings of SPIE, Vol. 12033, Medical Imaging 2022: Computer-Aided Diagnosis. https://doi.org/10.1117/12.2611788

Huong Pham