Projects | Baoru Huang

Grasp-Anything

Tue, 30 Sep 2025 00:00:00 +0000

Welcome to Grasp-Anything project! We tackle grasp detection by utilizing foundation models. Our project represents a data centric approach for grasp detection.

Grasp-Anything offers universality, featuring a wide range of everyday objects in natural arrangements, unlike other benchmarks limited by object selection and controlled settings.

CathSim

Sat, 27 Sep 2025 00:00:00 +0000

Welcome to CathSim Project!

CathSim provides a versatile platform for both research and training in robot-assisted endovascular intervention. The simulator is designed to support autonomous catheterization, data generation, and medical training via AR/VR devices.

We develop endovascular robotic systems specifically designed to assist in robot-assisted catheterization tasks. The robotic platforms integrate seamlessly with the CathSim simulator, providing a comprehensive toolset for researchers and clinicians to explore endovascular techniques.

CathEase is an accessible and cost-effective endovascular robot that focuses on translating and rotating the guidewire. It uses a Nema 17 stepper motor for translation, an additional motor for rotation, and is controlled by an Arduino Uno Rev3 with a CNC shield and two A4899 drivers, powered by a 12V DC source. Teleoperation input is provided through a Google Stadia joystick.

CathBot is a versatile master-slave robotic system designed for use in MRI environments. CathBot employs pneumatic actuation, enabling safe operation within MR settings. The master robot closely mimicks natural human motion such as grasping, retracting, and rotating the instrument while providing users with haptic feedback. This motion is mapped directly to a 4DoF MRI-safe slave robot, offering precise control and enhanced user experience during procedures.

Endomapper

Wed, 20 Nov 2024 00:00:00 +0000

Endoscopes traversing body cavities such as the colon are routine in medical practice. However, they lack any autonomy. An endoscope operating autonomously inside a living body would require, in real-time, the cartography of the regions where it is navigating, and its localization within the map. The goal of EndoMapper is to develop the fundamentals for real-time localization and mapping inside the human body, using only the video stream supplied by a standard monocular endoscope.

In the short term, will bring to endoscopy live augmented reality, for example, to show to the surgeon the exact location of a tumour that was detected in a tomography, or to provide navigation instructions to reach the exact location where to perform a biopsy. In the longer term, deformable intracorporeal mapping and localization will become the basis for novel medical procedures that could include robotized autonomous interaction with the live tissue in minimally invasive surgery or automated drug delivery with millimetre accuracy.

Our objective is to research the fundamentals of non-rigid geometry methods to achieve, for the first time, mapping from GI endoscopies. We will combine three approaches to minimize the risk. Firstly, we will build a fully handcrafted EndoMapper approach based on existing state-of-the-art rigid pipelines. Overcoming the non-rigidity challenge will be achieved by the new non-rigid mathematical models for perspective cameras and tubular topology. Secondly, we will explore how to improve using machine learning. We propose to work on new deep learning models to compute matches along endoscopy sequences to feed them to a VSLAM algorithm where the non-rigid geometry is still hard-coded. We finally plan to attempt a more radical end-to-end deep learning approach, that incorporates the mathematical models for non-rigid geometry as part of the training of data-driven learning algorithms.