x

Computer Vision

Our research focuses on advancing healthcare through cutting-edge technologies. Our focus areas include innovative imaging, reconstruction, modelling, visualisation, and navigation technologies. We aim to assist doctors with a clear understanding of medical conditions and precise treatment plans using minimally invasive approaches.

Key Research Focus
  • Integration and Design of Mono, Stereo & Multiple view Optical Tracking Systems
  • Medical Image Registration
  • Surgical Navigation System Modelling and Optimization

Projects

Optical Tracking in Surgical Navigation

Optical tracking systems are indispensable for both image-guided and imageless navigation. By using fiducial markers, we can visualize and accurately track the patient's anatomy. Adapting to various project requirements, we have successfully worked with different cameras and developed in-house algorithms and modules to enhance the efficiency of our tracking systems.

Key Topics

  1. Data filtering
  2. Tool design and calibration
  3. Error modeling & System optimization

Automatic Fiducial-less Registration for Intraoperative Oarm & 3D C-arm Images

Intraoperative imaging is widely used for surgeries.Using navigation + intraoperative imaging we developed an automatic fiducial-less registration which aims to align intraoperative imaging data from Oarm and 3D C-arm systems without using physical jigs or fiducials. This technique offers improved accuracy, efficiency, and workflow integration, benefiting both surgeons and patients, especially in image-guided spine and neuro surgeries.

Key Topics

  1. Pre-Calibration
  2. Oarm / 3D-Carm tracker
  3. Intraoperative alignment
  4. Cone-beam CT calibration

Universal Automatic Registration for Intraoperative Imaging Modalities

Automatic intraoperative registration is employed across all imaging modalities, integrating a registration jig, patient-specific markers, and uniquely shaped surgical tools. This approach enhances precision, significantly decreases the need for postoperative CT scans to detect misalignment, and mitigates errors arising from imaging discrepancies, model inaccuracies, and registration issues.

Key Topics

  1. Design of Universal Registration Artefact
  2. Intraoperative alignment 

Non-Rigid Registration for Interventional Oncology Procedures (Biopsy, Ablation & Pain Management)

While targeting soft tissues, mainly in the thorax region, there will be a movement due to patient respiration. This movement may cause errors in rigid registration. To overcome this we are developing non-rigid registration techniques. Non-rigid registration techniques are crucial for accurately aligning images during interventional oncology procedures such as biopsy, ablation, and pain management. These methods account for patient breathing movements, ensuring precise targeting and treatment delivery.

Key Topics

  1. Non-rigid object detection
  2. Affine transformation
  3. Automated trajectory planning

Surgical Navigation Software Development

We are developing an easy-to-use surgical navigation software which is compatible with CT, MRI & X-ray images. It will support point-based , surface-based and automation patient to image registration. This software supports multiple optical tracking systems and has in-build functionalities to check and evaluate the tracking subsystem and navigated instruments. It has annotations and sphere detection modules.

Surgical Navigation Simulation

Simulating surgical navigation systems allows for the testing and validation of various components and algorithms in a controlled environment. This includes the integration of optical tracking systems, registration methods, and navigation workflows, ultimately reducing the prototyping and development time. 

Key Topics

  1. Modelling of registration errors
  2. Non-linear optimization
  3. 3D visualization

Our Team

Bala Siva Surya
Project Engineer - Computer Vision
Rahav Gowtham Venkateswaran
Project Scientist - Computer Vision
Khaleel Ur Rehman
Project Associate - Computer Vision
Sree Vignesh
Research Intern - Computer vision
Poojitha P
Research Intern
Elakkiya S P
Research Intern -Computer Vision
Haris S
Research Intern - Computer Vision