<aside> 💡 Summary :

This project focuses on developing methods for extracting and visualizing individual vessel branches from medical imaging data, particularly CT scans. The goal is to address the challenges posed by the variability in blood vessel morphology and position across individuals. By automating the process of distinguishing and isolating specific vascular branches, this research aims to reduce cognitive load on medical professionals and enhance efficiency in surgical planning and diagnostic procedures. The project explores various approaches, including mask-based and mesh-based methods, to create a robust tool for improved vascular analysis in clinical settings.

</aside>

Background

In surgical planning and diagnostic procedures, a profound understanding of individual blood vessels is crucial. However, the human body's inherent variability leads to significant variations in the position and morphology of blood vessels across individuals. This diversity poses challenges in accurately identifying and isolating specific vascular branches from medical imaging data such as CT scans.

The ability to automatically distinguish and visualize individual vessel branches within the entire vascular network from medical images is a critical technological advancement. Such a capability can substantially reduce the cognitive load on medical professionals, enhancing both efficiency and accuracy in clinical decision-making processes.

To address this need, we have undertaken a project focused on developing methods to generate masks for individual vessel branches. This project explores various approaches to effectively segment and label distinct vascular structures from medical imaging data, aiming to provide a robust tool for improved vascular analysis in clinical settings.

Method

Mask-based Method

graph TD
    A[Read NIFTI File] --> B[Obtain Skeleton]
    B --> C[Find End Points and Junctions]
    C --> D[Separate Branch Lines]
    
    B --> E[Skeletonization Algorithm]
    C --> F[Point Detection Algorithm]
    D --> G[Branch Segmentation Algorithm]
    
    E --> |Thinning| E1[Medial Axis Extraction]
    F --> |Feature Detection| F1[Endpoint Identification]
    F --> |Feature Detection| F2[Junction Identification]
    G --> |Path Tracing| G1[Individual Branch Extraction]
    
    E1 --> B
    F1 --> C
    F2 --> C
    G1 --> D

This Mermaid diagram illustrates the process of vessel branch extraction as described in the method. It shows the main steps:

  1. Reading the NIFTI file
  2. Obtaining the skeleton
  3. Finding end points and junctions
  4. Separating branch lines

The diagram also includes sub-processes for each main step, showing the algorithms and techniques that might be involved in each part of the process. This provides a clear visual representation of the workflow for the mask-based method of vessel branch extraction.

Results

bifucation.png

Reference