Explore 2D Imaging Properties Exercise

In the following exercise, you'll learn how to use the Python package pydicom to extract imaging data from a DICOM file. Similar to how you looked at images and their intensity distributions in Lesson 1, you'll do this again but also before and after normalizing the image's intensity values.

You will be provided with a dataframe that contains bounding box coordinates for masses that have been identified by a radiologist in three separate DICOM images. The dataframe gives the starting X & Y coordinates of the bounding box around each mass, and the width and height of the mass. These four values should allow you to identify the specific rectangular section of each image that contains a suspicious mass.

Your job here will be to extract the imaging data from each DICOM, visualize it, normalize it, and then visualize only the section of the image that contains the suspicious mass. Throughout this process, you should be drawing insights about whether or not intensity values are a particularly good characteristic of masses that might help a machine learning algorithm automatically identify them.