[A brain tumor automatic assisted-diagnostic system based on medical image shape analysis].

Area of Science:

• Medical image analysis
• Artificial intelligence in healthcare
• Oncology

Background:

• Brain tumors require accurate and timely diagnosis.
• Current diagnostic methods can be time-consuming and may benefit from computational assistance.
• Integration of PACS functionalities with advanced analysis is crucial.

Purpose of the Study:

• To develop an automated brain tumor assisted diagnosis system.
• To enhance medical image analysis for tumor classification.
• To integrate image segmentation and feature extraction with machine learning for diagnosis.

Main Methods:

• Utilized a fuzzy region competition algorithm for real-time slice segmentation.
• Extracted shape features including contour label, compactness, moment, Fourier Descriptor, chord length, and radius.
• Employed a Bayesian network for automated brain tumor sorting and diagnosis.

Main Results:

• The system successfully segments brain tumor images.
• Key shape features are extracted for irregular tumor contours.
• The Bayesian network effectively sorts brain tumors for assisted diagnosis.

Conclusions:

• The developed system offers a viable approach for brain tumor assisted diagnosis.
• Medical image analysis combined with machine learning can improve diagnostic accuracy.
• This system has the potential to supplement existing PACS functionalities.