Kaixu Chen (陈 凯旭)

Kaixu Chen (陈 凯旭)

he/him

I am a researcher at the University of Tsukuba, specializing in the application of AI, DL, and ML to video and image processing.

Medical Image Analysis & AI Research

Medical Research Overview

My medical research focuses on developing video-based intelligent systems for clinical motion analysis and diagnosis, with an emphasis on human gait and posture assessment. By integrating computer vision, deep learning, and multi-view geometry, I aim to extract clinically meaningful motion representations from ordinary video data and translate them into reliable diagnostic insights.

A central challenge in medical video analysis is the gap between data-driven models and clinical interpretability. To address this, my work emphasizes phase-aware and attention-guided modeling, enabling neural networks to focus on diagnostically relevant body regions and motion phases (e.g., stance/swing cycles in gait). This design not only improves classification performance but also enhances model transparency, which is critical for clinical adoption.

Technically, my research explores multimodal fusion frameworks that combine RGB video, optical flow, skeletal keypoints, and domain priors into unified learning architectures. I further investigate multi-view 3D pose reconstruction and temporal refinement techniques to achieve robust motion estimation under unconstrained camera setups, making the approach applicable to real-world clinical environments.

Ultimately, my goal is to build interpretable, scalable, and non-invasive medical AI systems that support clinicians in screening, diagnosis, and longitudinal monitoring of movement-related disorders, thereby contributing to improved patient outcomes and more accessible healthcare technologies.

Research Lists

  1. Chen K, Asada T, Ienaga N, et al. Two-stage video-based convolutional neural networks for adult spinal deformity classification[J]. Frontiers in Neuroscience, 2023, 17: 1278584. Link
  2. Chen K, Xu J, Asada T, et al. PhaseMix: A Periodic Motion Fusion Method for Adult Spinal Deformity Classification[J]. IEEE Access, 2024. Link
  3. Tsukumo T, Chen K, Asada T, et al. Spinal Disease Classification Using Deep Learning on Dual-View Videos[C]//Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual International Conference. 2025, 2025: 1-5. Link