Research
I am interested in geometric mechanics, control theory and computer vision and their applications to mobile and medical robots. Below are the current projects I am involved with at Johns Hopkins.
Needle Steering
Needles are used in medical procedures such as biopsy, therapeutic drug delivery in cancer treatment, etc. Accurate placement of the needle tip is crucial for the success of these procedures. In this project, we use flexible bevel-tip needles. Due to the geometry of the needle tip, as the needle is inserted into tissue, the needle follows a circular path. This is advantageous in reaching hitherto unreachable (by rigid symmetrical-tip needles) while avoiding sensitive organs and bones.
I work on the image-guided control of needle steering. The goal of this part of the project is to develop feedback controllers that will automatically navigate the needle tip to the desired location or make the needle follow a desired path. Sensory feedback is through a vision system be it a stereo camera pair (for the transparent phantom tissue) or ultrasound or CT scans (for the live tissue). Needle steering is modeled as a 6 DOF nonholonomic generalized bicycle, which has a high degree of nonholonomy (4). We have designed and implemented controllers to stabilize the needle tip to a desired plane or a sphere. These controllers are developed to work alongside subspace planners to navigate the needle to a desired location. We are in the process of experimentally validating the controller-planner interaction. We are also working on the more generic trajectory following of the needle.
|
![]() |
- V. Kallem, D. E. Chang, and
N. J. Cowan.
Task-Induced Symmetry and Reduction in Kinematic Systems with
Application to Needle Steering.
In IEEE/RSJ Intelligent Robots and Systems (IROS), San Diego, USA, 2007.
Download: pdf
- V. Kallem, and
N. J. Cowan. Image-Guided Control of Flexible Bevel-Tip
Needles. In IEEE International Conference on Robotics and Automation (ICRA), Rome, Italy, 2007.
Download: pdf
Task-Induced Reduction
|
My work on the image-guided control of needle steering has led to a new framework for symmetry and reduction, which is induced by the control task. We have shown that for a class of kinematic systems on Lie groups, higher dimensional configuration spaces can be reduced to lower dimensional spaces, which simplifies the controller design to achieve a task. Examples of such systems: planar cart, needle steering, snake robots, etc. |
|
- V. Kallem, D. E. Chang, and
N. J. Cowan.
Task-Induced Symmetry and Reduction in Kinematic Systems with
Application to Needle Steering.
In IEEE/RSJ Intelligent Robots and Systems (IROS), San Diego, USA, 2007.
Download: pdf
Kernel-Based Visual Servoing
Traditionally in visual servoing, tracking and control are treated as two isolated problems. We are developing a technique called kernel-based visual servoing that eliminates the need to track features in the scene explicitly. We have developed this technique for 4 DOFs (3 translational directions and roll) and are currently working towards the complete 6 DOF case.
- V. Kallem, M. Dewan,
J. P. Swensen, G. D. Hager, and N. J. Cowan. Kernel-Based Visual
Servoing. In IEEE/RSJ Intelligent Robots and Systems
(IROS), San Diego, USA, 2007.
Download: pdf
For a complete list of my papers, go to my publications page.
© 2007-2008 Vinutha Kallem. Design Template by Andreas Viklund.
