An automatic brain tumor segmentation tool.

Area of Science:

• Medical imaging analysis
• Computational neuroscience
• Artificial intelligence in medicine

Background:

• Accurate brain tumor segmentation is crucial for diagnosis and treatment planning.
• Manual segmentation is time-consuming and prone to inter-observer variability.
• Automated methods are needed to improve efficiency and consistency.

Purpose of the Study:

• To introduce an automatic brain tumor segmentation method (ABTS).
• To segment multiple components of brain tumors using MRI.
• To evaluate the effectiveness of ABTS compared to expert segmentation.

Main Methods:

• Utilized four magnetic resonance image modalities.
• Employed a four-stage process including automatic histogram multi-thresholding.
• Incorporated morphological operations such as geodesic dilation.

Main Results:

• Achieved 81% Dice accuracy for edema segmentation.
• Achieved 85% Dice accuracy for gross tumor volume (GTV) segmentation.
• Demonstrated high effectiveness on 16 real brain tumor cases.

Conclusions:

• The proposed ABTS method is effective for brain tumor segmentation.
• ABTS shows promising results comparable to expert segmentation.
• This automated approach can aid in clinical decision-making for brain tumors.