RADDI - Radiology and Diagnostic Imaging

This course will discuss in detail the physics involved in the following imaging modalities: Radiography, Fluoroscopy, Conventional Tomography, Bone Densitometry, Mammography, Computed Tomography (CT), Nuclear Medicine, Ultrasound, and Magnetic Resonance Imaging (MRI). Prerequisites: Some fundamental physics of diagnostic imaging is required or consent of Department.

The course aims to cover medical image processing and analysis techniques, including de-noising, registration, segmentation, and 3D reconstruction, applicable in diagnostic imaging modalities such as ultrasound, computed tomography (CT), and magnetic resonance imaging (MRI). The course will also cover machine learning topics related to medical image analysis. Clinical examples in cardiovascular, musculoskeletal, and brain imaging will be discussed. Prerequisite: Linear algebra and knowledge in Python programming language or consent of the Department.

The course will cover applications of Machine Learning (ML) in medical imaging modalities like ultrasound, computed tomography (CT), and magnetic resonance imaging (MRI) that are commonly used in Radiology. Starting with a brief introduction to Artificial Intelligence and ML, this course will cover the perceptron model, multilayer perceptron (MLP), convolutional neural networks (CNN), recurrent neural networks (RNN), transformers, autoencoders, and generative models (like GANs). Image classification models, semantic segmentation models, and instance segmentation models used in medical image datasets will be discussed. This course is intended for graduate students in Radiology, Biomedical Engineering, and other relevant disciplines whose research interests are related to the use of Machine Learning (ML) techniques in medical imaging.

A seminar course for advanced students covering selected topics from the current literature in the fields of medical imaging, radiological physics, radiation biology and radiation biophysics.