Bowen Hu

University of Chinese Academy of Sciences, Shenzhen Institutes of Advanced Technology

Papers

Total Citations

H-Index

About

Bowen Hu is a researcher at the forefront of medical imaging and 3D computer vision, with a focused expertise in reconstructing complex anatomical structures from limited data. His primary research areas include 3D shape reconstruction, point cloud generation, and graph-based deep learning, specifically applied to neurosurgical planning and brain surgery navigation. Hu’s major contribution lies in developing novel generative models that can infer complete, high-fidelity 3D brain shapes from sparse or incomplete 2D inputs—a critical challenge for minimally-invasive and robot-assisted surgeries where direct visual access is restricted. His most cited work introduces a tree-structured graph convolutional network for point cloud generation, enabling the fusion of medical images with 3D shape representations to provide complementary microstructural details that enhance surgical accuracy. Another notable achievement is his hierarchical shape-perception network, which reconstructs 3D brain geometry from a single incomplete image, directly addressing the practical limitations of intraoperative data acquisition. With his papers accumulating over a dozen citations, Hu’s research is establishing a vital bridge between advanced geometric deep learning and real-world clinical needs, promising to improve both the safety and precision of neurosurgical interventions.