arXiv subjects

Minh-Van Nguyen

Publications and source records attributed to Minh-Van Nguyen.

9 recordsLinked to original sources

arXiv2026

Rad-JEPA 3D: Radiology Joint-Embedding Predictive Model for 3D Computed Tomography

Self-supervised pretraining is central to 3D medical image analysis, where unlabeled CT volumes are abundant but expert annotations are scarce. Yet existing volumetric encoders often fail to preserve the coarse spatial and geometric structure that downstream reasoning depends on, limiting their performance on organ disentanglement, abnormality detection, and spatial understanding when paired with language models. We introduce Rad-JEPA 3D, a joint-embedding predictive framework that learns volumetric CT representations by predicting the latent features of a complete scan from a masked view. At its core is a hybrid H-Mamba encoder that fuses a Mamba state-space branch, which models inter-slice continuity through sequential scanning, with a grouped-query attention branch, which captures cross-plane spatial context, combined through a lightweight per-token router. To improve the quality of intermediate representations, we further propose Hidden States Orthogonal Regularization (HSOR), which aligns student-teacher hidden states and reduces feature redundancy throughout the encoder. This layer-wise regularization produces more consistent and discriminative volumetric representations, leading to improved performance on organ recognition and spatial reasoning tasks. Pretrained on approximately 120,000 CT scans, Rad-JEPA 3D attains state-of-the-art results despite its compact size: with only 4.0B total parameters, it achieves competitive results with state-of-the-art on closed-ended VQA and the best average spatial-reasoning score on the Spatial-Med benchmark. Ablation studies confirm that the hybrid block and HSOR contribute complementary gains, and that the induced spatial structure can substitute for raw language-model scale on volumetric reasoning tasks.

cs.CV↗

arXiv2026

PRS-Med: Position Reasoning Segmentation in Medical Imaging

Prompt-based medical image segmentation has rapidly emerged, yet existing methods rely on explicit prompts like bounding boxes and struggle to reason about the spatial relationships essential for clinical diagnosis. While general-domain models attempt complex coordinate regression, these approaches often lack the structured reliability required for medical applications. In this work, we introduce PRS-Med, a unified framework that adopts an elegant, clinical-first approach to position reasoning segmentation. By utilizing a medical vision-language model integrated with a segmentation decoder, PRS-Med mimics the structured "search patterns" used by radiologists to identify pathologies within specific anatomical zones. To support this robust reasoning, we present the Medical Position Reasoning Segmentation (PosMed) dataset, comprising 116,000 expert-validated, spatially grounded question-answer pairs across six imaging modalities. Unlike previous brittle attempts at spatial reasoning, PosMed leverages a scalable, deterministic pipeline validated by board-certified radiologists to ensure clinical accuracy. Extensive experiments demonstrate that our zone-based reasoning not only improves segmentation accuracy (mean Dice improvements up to +31.2\%) but also provides a high-confidence interpretability layer that outperforms state-of-the-art complex reasoning models. By prioritizing functional reliability over unnecessary technical complexity, PRS-Med offers a practical and scalable baseline for the next generation of intelligent medical assistants.

cs.CV↗

arXiv2024

Sing-On-Your-Beat: Simple Text-Controllable Accompaniment Generations

Singing is one of the most cherished forms of human entertainment. However, creating a beautiful song requires an accompaniment that complements the vocals and aligns well with the song instruments and genre. With advancements in deep learning, previous research has focused on generating suitable accompaniments but often lacks precise alignment with the desired instrumentation and genre. To address this, we propose a straightforward method that enables control over the accompaniment through text prompts, allowing the generation of music that complements the vocals and aligns with the song instrumental and genre requirements. Through extensive experiments, we successfully generate 10-second accompaniments using vocal input and text control.

cs.SD↗

arXiv2024

KDAS: Knowledge Distillation via Attention Supervision Framework for Polyp Segmentation

Polyp segmentation, a contentious issue in medical imaging, has seen numerous proposed methods aimed at improving the quality of segmented masks. While current state-of-the-art techniques yield impressive results, the size and computational cost of these models create challenges for practical industry applications. To address this challenge, we present KDAS, a Knowledge Distillation framework that incorporates attention supervision, and our proposed Symmetrical Guiding Module. This framework is designed to facilitate a compact student model with fewer parameters, allowing it to learn the strengths of the teacher model and mitigate the inconsistency between teacher features and student features, a common challenge in Knowledge Distillation, via the Symmetrical Guiding Module. Through extensive experiments, our compact models demonstrate their strength by achieving competitive results with state-of-the-art methods, offering a promising approach to creating compact models with high accuracy for polyp segmentation and in the medical imaging field. The implementation is available on https://github.com/huyquoctrinh/KDAS.

eess.IV↗

arXiv2023

ALGNet: Attention Light Graph Memory Network for Medical Recommendation System

Medication recommendation is a vital task for improving patient care and reducing adverse events. However, existing methods often fail to capture the complex and dynamic relationships among patient medical records, drug efficacy and safety, and drug-drug interactions (DDI). In this paper, we propose ALGNet, a novel model that leverages light graph convolutional networks (LGCN) and augmentation memory networks (AMN) to enhance medication recommendation. LGCN can efficiently encode the patient records and the DDI graph into low-dimensional embeddings, while AMN can augment the patient representation with external knowledge from a memory module. We evaluate our model on the MIMIC-III dataset and show that it outperforms several baselines in terms of recommendation accuracy and DDI avoidance. We also conduct an ablation study to analyze the effects of different components of our model. Our results demonstrate that ALGNet can achieve superior performance with less computation and more interpretability. The implementation of this paper can be found at: https://github.com/huyquoctrinh/ALGNet.

cs.AI↗

arXiv2023

M^2UNet: MetaFormer Multi-scale Upsampling Network for Polyp Segmentation

Polyp segmentation has recently garnered significant attention, and multiple methods have been formulated to achieve commendable outcomes. However, these techniques often confront difficulty when working with the complex polyp foreground and their surrounding regions because of the nature of convolution operation. Besides, most existing methods forget to exploit the potential information from multiple decoder stages. To address this challenge, we suggest combining MetaFormer, introduced as a baseline for integrating CNN and Transformer, with UNet framework and incorporating our Multi-scale Upsampling block (MU). This simple module makes it possible to combine multi-level information by exploring multiple receptive field paths of the shallow decoder stage and then adding with the higher stage to aggregate better feature representation, which is essential in medical image segmentation. Taken all together, we propose MetaFormer Multi-scale Upsampling Network (M$^2$UNet) for the polyp segmentation task. Extensive experiments on five benchmark datasets demonstrate that our method achieved competitive performance compared with several previous methods.

eess.IV↗

arXiv2022

Res-Dense Net for 3D Covid Chest CT-scan classification

One of the most contentious areas of research in Medical Image Preprocessing is 3D CT-scan. With the rapid spread of COVID-19, the function of CT-scan in properly and swiftly diagnosing the disease has become critical. It has a positive impact on infection prevention. There are many tasks to diagnose the illness through CT-scan images, include COVID-19. In this paper, we propose a method that using a Stacking Deep Neural Network to detect the Covid 19 through the series of 3D CT-scans images . In our method, we experiment with two backbones are DenseNet 121 and ResNet 101. This method achieves a competitive performance on some evaluation metrics

eess.IV↗

arXiv2022

EfficientNet for Brain-Lesion classification

In the development of technology, there are increasing cases of brain disease, there are more treatments proposed and achieved a positive result. However, with Brain-Lesion, the early diagnoses can improve the possibility for successful treatment and can help patients recuperate better. From this reason, Brain-Lesion is one of the controversial topics in medical images analysis nowadays. With the improvement of the architecture, there is a variety of methods that are proposed and achieve competitive scores. In this paper, we proposed a technique that uses efficient-net for 3D images, especially the Efficient-net B0 for Brain-Lesion classification task solution, and achieve the competitive score. Moreover, we also proposed the method to use Multiscale-EfficientNet to classify the slices of the MRI data

eess.IV↗

arXiv2021

Refined Deep Neural Network and U-Net for Polyps Segmentation

The Medico: Multimedia Task 2020 focuses on developing an efficient and accurate computer-aided diagnosis system for automatic segmentation [3]. We participate in task 1, Polyps segmentation task, which is to develop algorithms for segmenting polyps on a comprehensive dataset. In this task, we propose methods combining Residual module, Inception module, Adaptive Convolutional neural network with U-Net model, and PraNet for semantic segmentation of various types of polyps in endoscopic images. We select 5 runs with different architecture and parameters in our methods. Our methods show potential results in accuracy and efficiency through multiple experiments, and our team is in the Top 3 best results with a Jaccard index of 0.765.

eess.IV↗