Convolutional neural network with a hybrid loss function for fully automated segmentation of lymphoma lesions in FDG PET images.
Fereshteh YousefiriziNatalia DubljevicShadab AhamedIngrid BloiseClaire GowdyJoo Hyun OYoussef FaragRodrigue de SchaetzenPatrick MartineauDon WilsonCarlos F. UribeArman RahmimPublished in: Medical Imaging: Image Processing (2022)
Keyphrases
- fully automated
- fully automatic
- loss function
- pet images
- automated segmentation
- manual segmentation
- positron emission tomography
- image analysis
- segmentation method
- segmentation algorithm
- pet ct
- accurate segmentation
- image data
- ct images
- convolutional neural network
- edge detection
- mr images
- ground truth
- segmented images
- semi automatic
- computer aided
- imaging modalities
- automatic segmentation
- skin lesion
- mri images
- region of interest
- medical images
- low contrast
- image segmentation
- level set
- image registration
- region growing
- image reconstruction
- input image
- pairwise
- mri data
- visual inspection
- clinical applications
- medical imaging
- cell nuclei
- similarity measure