Modelling the movement of the liver during breathing has been studied with various models using Deep Learning and Machine Learning techniques.
These approaches include Active counter model, Mean Shift, Optical Flow. There are also Deep Learning approaches Unet with Densenet201, Unet with EfficientNetB7. Here, an approach was developed by applying Edge detection and Optical Flow on the masked images obtained.
As a result of this approach, the respirator motion can be obtained as a signal. This signal can estimate the instantaneous position of the liver or cancer in the liver in surgical operations.