Two-dimensional segmentation fusion tool: an extensible, free-to-use, user-friendly tool for combining different...

Area of Science:

• Medical Imaging
• Computational Geometry
• Bioinformatics

Background:

• Accurate tumor segmentation in medical imaging is complex and subjective, leading to diagnostic variability.
• Existing methods for fusing multiple 2D line segmentations lack validation and a defined ground truth.
• Subjectivity in contour identification affects the reliability of diagnoses and treatment planning.

Purpose of the Study:

• To develop a computational tool for objectively fusing multiple 2D closed line segmentations.
• To provide a validated solution for establishing a reliable contour line from subjective inputs.
• To support medical specialists and researchers in fields requiring 2D line fusion.

Main Methods:

• Development of the Two-Dimensional Segmentation Fusion Tool (TDSFT) as a standalone application.
• Implementation of various algorithms for computing the mean of multiple 2D lines.
• Design of a user-friendly interface for easy integration and parameter configuration.

Main Results:

• The TDSFT offers a user-friendly interface for combining multiple 2D lines using various algorithms.
• The tool is available as a free, standalone application for MAC, Linux, and Windows.
• The TDSFT is designed for extensibility, allowing the addition of new algorithms and parameters.

Conclusions:

• The TDSFT provides a valuable tool for medical specialists, particularly in oncology and histology, by reducing subjectivity in image segmentation.
• The application's design facilitates its use in diverse scientific fields requiring the fusion of 2D closed lines.
• The TDSFT's open-source nature and extensible architecture promote further research and development in image analysis and segmentation fusion.