For analyzing, sorting, and annotating large medical datasets it is of special interest to automatically determine the visible body part since this information is typically not sufficiently described in the image meta-data. In this talk, I will present you the python package Body Part Regression, which I developed in connection with my master thesis to automatically recognize the scanned body part in CT images. Through a live demonstration, you will obtain a good insight into how you can use the package yourself to analyze your medical datasets. We will go into 3 application examples in more depth:

1. Predicting the visible body part in a CT image.
2. Automatically cropping CT images to a required body region.
3. Automatically finding corrupted CT images.

The notebook to the presentation can be found here.