Completed notes of the course

Pagina 1

1 Lesson 23/04/2018 – Medical Images Biomedical Images The first point is providing basic concept about biomedical images. We have to introduce image processing techniques (in particular the Tomographic Reconstruction), working in the space domain instead of the time domain we’ve seen. Even in space domain the object will be considered as a discrete object, so we have to extend the domain concept that we have already introduced: Fourier Transform. So, how to represent the special frequency, not in the frequency as intended in the previous part of the course in the 2D, passing through the space frequency? In this approach, we will use only the discrete FFT but applying it to the domain. There are some indicators, indexes, that contribute through the analysis of the images; among them there will be the resolution. The resolution can be associated to different objectives: Time resolution, spatial resolution, different kind of resolution. Resolution means which is the smallest object that we are able to recognize inside our image ( pixel in our PC). Object smaller than the dimension of the pixel are not distinguishable. The pixel is homogeneous and decide the pixel fixes the minimum dimension of the object we are able to recognize. We will speak about resolution and other characteristics of the images. One of them is for example “contrast”. What is “contrast”? Contrast is the ability to distinguish objects inside a certain picture. It increases or decreases the ability to recognize some object. As in time domain we have always the problem of noise that is superimposed to our images. Noise can come from other sources, different biological one that we are trying to analyze, but it can be also electronic noise or noise related to measure band. There are some different topics:…