Scientific Disciplinary Sector (SSD)
INF/01 - INFORMATICS
The teaching is organized as follows:
The course aims at providing students with the applied and theoretical basis for processing biomedical images and extract useful information from them to support the diagnosis process. Knowledge and understanding. At the end of the course, the student shall demonstrate that he/she can apply the material discussed in the lectures to solve effectively the most common issues that may happen throughout a typical analysis pipeline, from the acquisition of the raw images to the correct interpretation of the information extracted from them. Applying knowledge and understanding. In particular, at the end of the course the student shall demonstrate to be able to: a) open, handle and manipulate the multidimensional data acquired with the major imaging modalities (X-rays, magnetic resonance imaging, nuclear medicine and ultrasounds); b) evaluate advantages, disadvantages and peculiarities of each modality; c) interpret correctly the content of such images and be able to link them to physical and biological features of the tissue/organ under exam. Making judgements. The student will be able to develop an analysis pipeline to extract useful information from such biomedical images and help the diagnostic process, applying at each step the most adequate processing choices for the specific data at hand. Communication. At the end of the course, the student shall demonstrate the ability to effectively interact with different collaborators having specific backgrounds typically required in a clinical study based on medical imaging, e.g. engineers, physicists, physicians etc. Lifelong learning skills. He/she will also have the required foundations to be able to elaborate further on any scientific, methodological and recent advances in the field beyond the content of the lectures to extend such basic techniques to diverse and more complex analysis scenarios.
Prerequisites and basic notions
Good familiarity with the main contents of basic courses such as "Signal and image processing" is strongly recommended (but not strictly necessary) for a proper and complete understanding of the course.
(1) Basic concepts
- Image properties: pixel vs voxel, spatial resolution, orientation, data type, etc
- File formats
- Signal-to-noise ratio, Contrast-to-noise ratio, etc
(2) Main imaging modalities (recall principles)
- Radiography: X-rays projection, fluoroscopy and computed tomography
- Nuclear medicine: SPECT and PET
- Magnetic Resonance Imaging
(3) Medical image registration
- Geometric transformations
- Features and similarity measures
- Transformations (linear vs non-linear)
(4) Morphometry analysis
- Region-of-interest analysis
- Voxel-based morphometry
- Surface-based morphometry
- Tract-based morphometry in white matter
(5) Structural connectivity estimation
- Diffusion MRI: principles and main applications
- Estimating microstructural features of the neuronal tissue
- Inferring fibers geometry and organization (a.k.a. tractography)
- Recent advances
(6) Functional connectivity estimation
- Physiology of neurons and how to record their activity
- Functional MRI: principles and main applications
- Elettroencefalography and magnetoencefalography: principles and main applications
- Static vs dynamic connectivity
(7) Network analysis (a.k.a. connectomics)
- A network representation of the brain: how and why?
- Studying brain networks with graph theory: concepts and measures
- Comparing brain networks in different groups of subjects
- Hands-on activities on the topics covered throughout the course
- Real neuroimaging data provided to be analyzed
Lectures for the theory part with various invited talks given by international experts; in the laboratory part, students will install and use the main software to analyze real images of clinical studies. All lessons will be recorded and made available.
Learning assessment procedures
The exam consists of a project, assigned at the end of the course, aimed at analyzing magnetic resonance images taken from a real clinical study. This final project is a fundamental part of the course, as it provides students with the opportunity to put into practice the concepts studied during the theory part, understand the peculiarities of each acquisition mode, touch some typical problems that can occur when biomedical images are processed and the most appropriate techniques applied to improve the quality of the images and extract useful information from them.
The evaluation will be based on (A) a short written report (max 10 pages) and (B) an oral presentation / discussion (which may also include questions on the theory part), in which both the exposure and the interpretation of the methodologies used and the results obtained will be assessed.
Criteria for the composition of the final grade
The final grade will consist of 50% of the written report and 50% based on the presentation.