SAFARI

SAFARI performs shape analysis of AI- or manually-segmented regions of interest (ROIs) in medical images to extract quantitative shape descriptors for downstream signal processing, statistical analysis, modeling, and machine learning.

Key Features:

• ROI labeling: Converts segmented maps into labeled regions of interest for shape analysis.
• Shape feature extraction: Extracts quantitative shape descriptors and represents ROIs as analyzable shape features.
• Segmentation input: Accepts segmented maps produced by AI algorithms or by manual segmentation.
• Imaging modalities: Operates on images from X-ray, Computed Tomography (CT), Magnetic Resonance Imaging (MRI), and pathology imaging.
• Standardization: Provides standardized procedures for translating ROIs into consistent shape representations to improve reproducibility and comparability.
• Downstream compatibility: Produces features intended for signal processing, statistical analysis, modeling, and machine learning.
• Implementation: Implemented as an R package for computational analysis.

Scientific Applications:

• Lung cancer survival analysis: Extraction of shape features associated with survival outcomes in patients with stage I-IV lung cancer.
• Glioblastoma prognosis: Extraction of shape features associated with survival outcomes in glioblastoma patients.
• Medical image ROI analysis: Standardized shape-based analyses for medical image segmentation studies and shape-based research applications.

Methodology:

Translates ROIs from segmented maps (AI-generated or manual) into analyzable shape representations and extracts quantitative shape descriptors from segmented images derived from X-ray, CT, MRI, and pathology imaging.