Computer Implemented Method for Estimating Lung Perfusion from Thoracic Computed Tomography Images

A computer implemented method for estimating lung perfusion from thoracic computed tomography (CT) images, comprising the steps of:

• a. Receiving with at least one processor a CT image of at least a part of a lung, in particular a CT scan taken at inspiration, and more in particular a non-contrast CT scan taken at inspiration;
• and
• b. Processing the CT image with a trained computer implemented algorithm executed on at least one processor to estimate lung perfusion based on the CT image;
• c. Wherein the trained computer implemented algorithm is trained by
• i. Receiving with at least one processor a set of CT images from at least a part of a lung, in particular a set of CT scans taken at inspiration, and more in particular a set of non-contrast CT scans taken at inspiration;
• ii. Receiving with at least one processor reference perfusion information, such as a perfusion image, corresponding to the CT image, for instance provided by one or more image modalities like lung perfusion scintigraphy, Single Photon Computed Tomography (SPECT), dual-energy perfusion CT, contrast CT subtraction imaging, or perfusion based on multi temporal CT imaging;
• and
• iii. Processing the received set of CT images and reference perfusion information to train the computer implemented algorithm to estimate lung perfusion in a CT image based on the corresponding perfusion information, wherein preferably the training is a one-time training.