Detection of Prostate Cancer in Whole-Slide Images Through End-to-End Training With Image-Level Labels.
Hans PinckaersWouter BultenJeroen van der LaakGeert LitjensPublished in: IEEE Trans. Medical Imaging (2021)
Keyphrases
- end to end
- input image
- labeled images
- image data
- image features
- text localization and recognition
- prostate cancer
- histological images
- image analysis
- segmented images
- segmentation method
- object detectors
- region of interest
- edge detection
- bounding box
- gray value
- segmentation algorithm
- multiscale
- computer aided
- deformable registration
- contrast enhancement
- prostate segmentation
- medical image analysis
- speckle noise
- training set
- feature vectors
- high resolution
- image guided
- superpixels
- magnetic resonance
- gray level
- image quality
- medical imaging
- multi label
- mr images