MetaSel

MetaSel automates selection of high-quality metaphase chromosome spreads for karyotyping to support diagnosis of genetic disorders.

Key Features:

• Gaussian-based classification: Implements a Gaussian-based classification technique to categorize image objects based on shape parameters.
• Chromosome shape classes: Classifies objects into four classes: chromosomes with straight shapes, chromosomes with skewed shapes, chromosomes exhibiting overlapping bodies, and non-chromosome objects.
• Rule-based selection criteria: Applies a rule-based system that prioritizes spreads containing predominantly straight and skewed chromosomes while minimizing overlapping chromosomes for effective karyotyping.
• Statistical thresholding: Uses statistical models with Gaussian distributions to determine threshold values for various image parameters.
• Accuracy: Reports classification performance of over 90% accuracy for identifying suitable metaphase spreads.
• Chromosome editing operations: Provides operations for splitting, merging, and fixing chromosome objects.
• Karyotyping editor operations: Supports moving, rotating, and pairing homologous chromosomes for downstream karyotyping.

Scientific Applications:

• Diagnostic karyotyping: Selection of suitable metaphase spreads to enable accurate karyotype analysis for diagnosis of genetic disorders.
• Automated metaphase spread pre-selection: Automated pre-selection of spreads to reduce the need for manual examination of metaphase images.

Methodology:

Uses a Gaussian-based classification technique and a rule-based system employing statistical models with Gaussian distributions to determine threshold values for image parameters and classify objects into four categories.