BE 645: Artificial Intelligence (AI) and Radiomics (Summer 2025) - Lecture 02 - Part A

BE 645: Artificial Intelligence (AI) and Radiomics (Summer 2025)

This second lecture focuses on first-order features used to analyze medical images, particularly liver images and their segmentation masks. It details the calculation of statistical features such as minimum, maximum, mean, variance, standard deviation, skewness, kurtosis, and excess kurtosis from pixel intensity histograms of Regions of Interest (ROIs). These calculations are demonstrated through practical examples and exercises, interpreting results in terms of central tendency, variability, symmetry, and peakedness relative to a normal distribution. Python scripts for both 2D and 3D volumes illustrate how normalization and background exclusion impact feature extraction. The lecture also highlights how these statistical measures inform texture and intensity patterns in radiomics, which are critical for distinguishing between healthy and pathological tissues. The workflow includes acquiring and segmenting images into ROIs, extracting first-order features, developing and training AI models, and ensuring explainability of results. Practical demonstrations using NIfTI files and Python scripts show image manipulation, ROI extraction, and radiomic feature calculation, emphasizing the importance of understanding the statistical properties of extracted features and considering background effects in analysis.

Codes can be accessed using: https://github.com/HossamBalaha/BE-64...

Summer 2025 Playlist:    • BE 645: Artificial Intelligence (AI) and R...  

Spring 2025 Playlist (Old):    • BE 645: Artificial Intelligence (AI) and R...  

Summer 2024 Playlist (Old):    • BE 645: Artificial Intelligence (AI) and R...