Image Registration in Treatment Planning

A method for training a machine learning model to be used in image registration of images of a patient comprising:

• gathering training data to the machine learning model;
• estimating a loss function and/or its gradient for said machine learning model using the training data;
• and
• updating parameters of a parameterized model of said machine learning model based on the estimated loss function and/or its gradient.

A computer-readable medium having stored therein computer-readable instructions for a processor, wherein the instructions when read and implemented by the processor, cause the processor to:

• select one or more tracking regions in a set of image data of an image of a patient, each tracking region including a tracking target or set of tracking targets;
• identify corresponding tracking targets in sets of image data from images captured at different times of the patient;
• and
• estimate a displacement field describing displacement vectors for a number of points in said sets of image data including said tracking regions, each displacement vector specifying a position of a point in reference to a previous position.