Chromosome Abnormality Detection Using Visual Geometric Transformer and Mantis Search Optimization.

Area of Science:

• Genetics and Molecular Biology
• Computational Biology
• Medical Imaging Analysis

Background:

• Karyotyping is essential for identifying genetic abnormalities but existing methods face challenges with processing time and feature extraction.
• Chromosomal analysis is critical for diagnosing genetic disorders, necessitating efficient and accurate detection techniques.

Purpose of the Study:

• To introduce a novel algorithm, Visual Geometric Transformer-based Mantis Search (VGT-MS), for enhanced detection of chromosomal anomalies.
• To address the limitations of current karyotyping methods, focusing on improved efficiency and accuracy in feature extraction.

Main Methods:

• A preprocessing step removes irrelevant background artifacts from chromosome images.
• Feature extraction is performed using the VGG-16 network, followed by classification with the Vision Transformer.
• The Mantis Search Algorithm optimizes the model's parameters for enhanced performance.

Main Results:

• The VGT-MS framework achieved high performance across key metrics: 98.0% accuracy, 97.2% precision, 96.2% recall, and 96.7% F1-score.
• The proposed model demonstrates significant improvements in detecting chromosomal anomalies while reducing computational overhead.
• Experimental results confirm the model's effectiveness in accurate and efficient chromosomal abnormality detection.

Conclusions:

• The VGT-MS framework offers a powerful and efficient solution for detecting chromosome abnormalities.
• This novel approach overcomes the drawbacks of conventional methods, providing a more effective tool for genetic analysis.
• The VGT-MS algorithm represents a significant advancement in automated karyotyping and genetic disorder diagnosis.