Yi Qin (Eason)

I'm now currently a Ph.D. candidate supervised by Prof. Xiaomeng Li at ECE, the Hong Kong University of Science and Technology. I also had the opportunity to work with Prof. Hao Wang and Prof. Lu Mi. I am also collaborating with Guangdong Cardiovascular Institute and Prince of Wales Hospital on Echocardiography AI. Before joining HKUST, I obtained BEng at the South China University of Technology majoring in Automation Science and Engineering.

Email  /  CV  /  Scholar  /  Twitter  /  Linkedin

Research

My research interests lie at the intersection of machine intelligence and digital healthcare, with a particular interest on echocardiography and cardiology. Specific interested topics include:

• Diffusion-based Generative Model
• Trustworthy/Explainable ML (Energy-based Concept-based models)
• Foundation Model (FM) based Disease Diagnosis/Prognosis Prediction

We introduce Energy-Based Concept Bottleneck Models (ECBM) as a unified framework for concept-based prediction, concept correction, and fine-grained interpretations based on conditional probabilities.

To fully exploit the diffusion model's ability to guide the registration task, we present two modules in FSDiffReg: Feature-wise Diffusion-Guided Module (FDG) and Score-wise Diffusion-Guided Module (SDG).

We propose Multi-Agent Collaborative Expertise Extractor, a multi-agent system that builds EchoCardiography Expertise Database, the richest cardiac knowledge base from diverse sources. We also introduce Echocardiography Expertise-enhanced Visual Instruction Tuning, a lightweight tuning method that efficiently injects this expertise into models by training less than 1% of parameters.

EchoViewCLIP addresses these issues using a large dataset with 38 standard views and OOD samples. It introduces a Temporal-informed Multi-Instance Learning (TML) module for capturing key frames and a Negation Semantic-Enhanced (NSE) detector for OOD rejection. A quality assessment branch boosts reliability. The model achieves 96.1% accuracy, advancing fine-grained view recognition and robust OOD handling in echocardiography.

We introduce CvG-Diff, a fast, high-quality sparse-view CT reconstruction method that models artifacts deterministically and introduces two novel techniques—EPCT and SPDPS—to reduce errors and improve efficiency.

We introduce XMedGPT, a multi-modal medical AI assistant that enhances clinical usability by combining accurate diagnostics with visual-text explainability and uncertainty quantification, enabling transparent and trustworthy decision-making.

We introduce RCMed, a full-stack medical AI assistant that enhances multimodal accuracy through hierarchical vision-language grounding and a self-reinforcing correlation loop. Trained on 20M samples, it excels in 165 clinical tasks across 9 modalities, achieving state-of-the-art performance and strong generalization in real-world cancer diagnosis and cell segmentation.

We propose Energy-Based Conceptual Diffusion Models (ECDMs), a framework that unifies the concept-based generation, conditional interpretation, concept debugging, intervention, and imputation under the joint energy-based formulation.

CUDA improves the robustness of Concept Bottleneck Models under domain shifts by aligning concept representations with adversarial training, allowing flexible differences, and enabling concept inference without labels. It boosts interpretability and outperforms state-of-the-art CBM and domain adaptation methods.

Projects and Patents

Honors, Awards, and Services

This website is adapted from Jonathan Barron's source code.